SlideShare ist ein Scribd-Unternehmen logo
1 von 86
Downloaden Sie, um offline zu lesen
IMPORTANT DISCLOSURE FOR U.S. INVESTORS: This document is prepared by Mediobanca Securities, the equity research department of Mediobanca S.p.A. (parent company of Mediobanca Securities USA LLC (“MBUSA”)) and it is distributed in the United States by MBUSA
which accepts responsibility for its content. The research analyst(s) named on this report are not registered / qualified as research analysts with Finra. Any US person receiving this document and wishing to effect transactions in any securities discussed herein should
do so with MBUSA, not Mediobanca S.p.A.. Please refer to the last pages of this document for important disclaimers.
RegObs Special Report
07 December 2022 Update
Bankosaurus jumping on the asteroid
RegObs Special Report - Digital euro: the
ECB saving Europe again
Sign up to our Digital Currency Editorial
Andrea Filtri
Equity Analyst
+44 203 0369 571
Andrea.Filtri@mediobanca.com
Alberto Nigro
Equity Analyst
+39 02 8829 9540
Alberto.Nigro@mediobanca.com
Jordan Bartlam
Equity Analyst
+44 203 0369 692
Jordan.Bartlam@mediobanca.com
Debunking the bank-dinosaur vs tech-asteroid myth
The rise of new technologies and of Big Tech/fintech companies based on them has
been juxtaposed to incumbent, slow-moving banks so far that the oversimplified
metaphor of the bank-dinosaur at risk of extinction from the tech-asteroid has been
widely utilised by mainstream media; chasing the evergreen popularity of the David
vs Goliath duel. This report provides elements confirming the threat is there, but
that the Bankosaurus could still evolve by embracing technology, gaining agility to
jump on the asteroid and buying the ticket for the perpetuation of the species.
Disruptive technologies: DLT, MLAI and quantum; embrace to evolve
We identified machine learning/artificial intelligence (MLAI), distributed ledger
technology (DLT, aka blockchain) and quantum computing (QC) as 3 innovations
which could disrupt the banking model (among many others) as we know it.
MB surveys: banks know they must change, they just do not know how to… yet
We engaged EU banks with 2 surveys to position them on the transformation
journey: 1) gauging where they stand on MLAI, DLT, QC, 2) showcasing of use cases.
With c.20 respondents, we draw some conclusions: a) the majority of investment is
going into low-key AI/chatbots, b) short inception-to-production, high external
staff reliance and efficiency-focus either reflect good project management or room
to move up along the innovation vector, c) low DLT penetration embodies aversion
to cooperation with peers and the need to fully control investments, d) absence of
payment projects echoes the exit from the space, a long-term strategic mistake to
us, while quasi-absence of QC is justified; it is just too early for it. Banks know they
must change at some point; they are just not daring/worried enough and mounting
external pressure could help them.
MB methodology: three steps to find the cure - Step 1: breaking the bank
We have gone into our lab to find a 3-step cure for the endangered species. EU
banks adopt a vertically integrated model mixing very different businesses and the
first step is to ringfence the affected organs. We separated standardised banking –
businesses more exposed to technological disruption – from customised banking -
which is more immune – of the 12 most representative EU banks (30% market share).
This shows 2/3 of operating income, 60% of costs and 50% of loans belong to
standardised banking, under threat from technological disruption.
Step 2: digital disruption eroding 30% of standardised revenues
We looked at sectors already disrupted by tech: film rental, photography, local
taxis and retail bookstores. Anecdotal evidence points to 30-50% margin erosion.
We simulate a central case made of c.30% compression of standardised banking
revenues, requiring c.60% cost cuts to hold profitability steady or C/I going from
c.50% to <30%, better than best-in-class Nordic banks.
Step 3: the great (cost) reset
We developed the DLT cost function, validated by Spunta’s case, featuring fixed
costs progressively handing over to variable transaction costs as volumes pickup.
Slimmed-up banks can jump on the asteroid
We simulated the migration of EU standardised banking onto a single, hypothetical
DLT platform maximising scale economies. Our simulation indicates a collapse in
aggregate cost, where MLAI and QC would also contribute, protecting sector
profitability. Our new, DLT based cost estimate shows a margin of error of c.400x,
still leaving banks on current profitability, even after revenue attrition from digital
disruption. Hence, cost miniaturisation is the evolution banks need to not only avoid
extinction, but potentially to thrive on higher profitability.
RegObs Special Report
07 December 2022 ◆ 2
Contents
Executive Summary ................................................................................................ 3
Where we left off .................................................................................................. 6
New technologies: evolution or extinction? .................................................................... 9
MB survey suggests A.I. and cost efficiency in focus… ....................................................... 11
…seconded by our “use case” almanac ........................................................................ 15
Good effort, but far from transformative innovation yet ................................................... 19
“breaking the bank”: isolating the species at risk ........................................................... 21
Digital disruption carries 30-50% margin compression… ..................................................... 26
…or standardised revenues -30%; costs -60% to offset ....................................................... 27
The great (cost) reset: scaling up DLT to do the trick....................................................... 36
Evolution: cost miniaturisation to avoid extinction .......................................................... 39
Digital Deep Dive.................................................................................................. 43
Digital disruption: case studies in-depth ...................................................................... 44
Distributed ledger technology (DLT), aka blockchain........................................................ 48
Machine Learning / Artificial Intelligence ..................................................................... 53
Quantum Computing.............................................................................................. 59
Mediobanca digital benchmark survey ......................................................................... 70
Showcasing banks’ use cases .................................................................................... 72
Annex ............................................................................................................... 78
List of Pictures .................................................................................................... 79
List of Tables ...................................................................................................... 80
References ......................................................................................................... 81
RegObs Special Report
07 December 2022 ◆ 3
Executive Summary
Debunking the bank-dinosaur, technology-asteroid myth
The rise of new technologies and Big Tech/fintech companies based on them has been often
journalistically contraposed to slow-moving, anachronistic commercial bank pachyderms. In the
oversimplified manner complex dynamics are often taken to the wider public, the metaphor of the
bank-dinosaur at risk of extinction from the technology-asteroid (fintech/Big Tech aggression) has
been widely disseminated by the mainstream media; chasing the evergreen popularity of the David vs
Goliath duel.
This report argues that the threat is there, but that the “dinosaur” could well decide to mutate by
embracing new technologies, gaining agility, adapting to the new ecosystem and ultimately jumping
on the “asteroid” to buy a survival ticket for the perpetuation of the species.
Banking Threat #1: in 2021, the digital euro
In Spring 2021 we explained digital currencies and how, more particularly, central bank digital
currency (CBDC) could be a threat to commercial banking deposits (see RegObs Special Report - Digital
euro: the ECB saving Europe again - by A.Filtri & Team). The conclusion was that the digital euro could
both defend monetary sovereignty from the thread of euro displacement by foreign (private or public)
digital currencies and – a European peculiarity – take the European integration process to the next
level by mutualising banking risk via the ECB as, with the digital euro, the central bank could become
the primary source of funding for commercial banks.
Banking Threat #2: new technologies (MLAI, DLT, QC) today
After much studying and learning about new technologies, we have identified Machine
Learning/Artificial Intelligence (MLAI), Distributed Ledger Technology (DLT, aka blockchain) and
Quantum Computing (QC) as three powerful innovations which could disrupt the banking model (among
many others) as we know it. (see DLT, MLAI and QC for in-depth background material).
MLAI – Machine Learning/Artificial Intelligence can be thought of as an increasingly less supervised
spectrum (“Symbolic” up to “Strong”) of processes to be operated and executed by machines.
However, while thoughts immediately fly to Terminator, we are still scrambling on the ground. Banking
is currently positioned at the weaker end of the spectrum, such as “ChatBots”.
DLT - Distributed ledger technology (DLT) - more commonly known as blockchain – is a tool to provide
trust between parties relying on the same data records, without the need for a centralised, third-
party authority. The immense data management potential introduced by the technology contrasts with
the very limited adoption by banks thus far due to their adversity to cooperation and shared control.
QC - Quantum computing is a new technology taking computation power to a new dimension,
leveraging on quantum physics. We are still in the development stage of the technology, so that it is
just too early to see its adoption by the industry.
Banks’ psychology: rational reaction or amygdala hijack? Evolve or go extinct
Over the millennia, human beings have developed autopilot response mechanisms to escape immediate
threats. When sudden life-threatening situations arise, we react by instinct, without thinking. We
naturally experience the amygdala hijack, i.e. when rational thought is inefficient or too slow to get
us out of trouble, we behave in patterns, usually attack or flight, to increase our chances of survival.
Whereas this could be life-saving when suddenly we find ourselves in front of an incoming bus, the
amygdala hijack can work against us in more complex, real life situations, when we usually have the
time to evaluate the option and rationally choose the action with the highest chance of success.
Another instinctive reaction human beings tend to have before a threat is to freeze, put the head
under the sand and do nothing about it, hoping that it will go away.
RegObs Special Report
07 December 2022 ◆ 4
Our banking-psychology expertise imposes us to counsel banks: a) to accept that there is a clear threat
and that it is coming, b) that there is enough time to avoid an instinctive reaction and instead adopt
a rational behaviour, c) not to be complacent with the time allowed and get on with it.
Our work should help fast-tracking banks through the investigation phase and getting them working
towards a solution, with the comforting conclusion that – if they act in a timely fashion and in an
adequate response – their existence is not only under question, but it could even prosper.
Mediobanca surveys: banks know they must change, but they do not know how to… yet
To take a snapshot of the current location of European banks along their transformation journey, we
send them two surveys: 1) a multiple choice questionnaire to gauge where each bank is on MLAI, DLT,
QC, 2) a showcasing of current success projects. We thank the c.20 respondents, whose responses we
aggregated to describe the overall state of the sector on innovation. The two surveys offer coherent
pictures, allowing us to take the following conclusions: a) banks know they need to change at some
point and they are progressively increasing their engagement, b) the majority of their efforts are going
in a linear evolution of the current operating model, with heavy investment in chatbots as the “new
automation”, not quite a radical rethink, yet, c) short inception-to-production timing, high usage of
external staff and focus on efficiency gains either reflect good project management and cost discipline
or room to increase ambition along the innovation vector, with innovation leaders often incentivised
by short-term, more quantifiable results rather than by transformative, multi-year projects, d) low
penetration of DLT reflects the additional complexity of the technology requiring cooperation with
peers/competitors and the desire to fully control each investment, e) the absence of payment projects
reflects the exit by banks from the space, potentially a strategic mistake in the long-term given the
scale, the innovation know-how and the key role of the business within banks’ product offering to
clients, f) the quasi-absence of QC is justified in our view. The technology is just not mature enough
to invest into it now; it will come, it is just too early for the time being.
We conclude that banks have realised they must change. They are just not daring enough or in enough
of a hurry to accelerate their more radical, self-initiated mutation. External pressure (market,
regulator, competitors) could improve self-awareness going forward, we suspect.
Mediobanca Lab: three steps to finding the cure
Concerned with the potentially endangered species, we have gone into our laboratory to diagnose the
patient by analysing the threats and working out a possible treatment through a three-step
methodology. We see the current banking model at risk, hence requiring a deep genotype mutation to
evolve and survive. We are afraid there is no plan B.
Step 1 - “breaking the bank”: isolating the species at risk of extinction
European banks adopt a vertically integrated model, bundling together retail banking services
(deposits, current accounts, mortgages and personal loans/credit cards, payments) with corporate &
investment banking services (SME and corporate lending, markets, advisory) and wealth management
(private banking, asset management, insurance, asset gathering) and specialised financial services
(car financing). Technology is not threatening such a vast variety of businesses at once. While QC still
needs some time to hit the ground with its superior computational power, MLAI and DLT offer vast
enhancements in the interpretation and in the handling of data, transforming processes/products
which can be digitised and standardised into platforms able to support enormous scale economies.
We attempted a surgical procedure separating standardised banking – more exposed to technological
disruption – from customised banking - where human capital and personal advisory will remain key –
of the twelve most representative European banks, attracting 30% market share. This shows 2/3 of
operating income, 60% of costs and 50% of loans belong to standardised banking.
RegObs Special Report
07 December 2022 ◆ 5
Step 2 – Digital disruption: 30% revenue attrition from margin compression
We look at prior sectors disrupted by technological transformation such as the film rental,
photography, local taxis and retail bookstores sectors. Anecdotal evidence points towards 30-50%
margin compression (you can read our in-depth case studies here). We simulate three scenarios (base,
pessimistic, optimistic) of revenue attrition to standardised banking. Our central case indicates c.30%
erosion in standardised revenues, requiring c.60% cost reduction to hold profitability steady, implying
standardised banking cost/income should shrink from c.50% today to <30%, tomorrow, ahead of the
currently most efficient Nordic banks.
Step 3 – The great (cost) reset: scaling up DLT to do the trick
Despite being in vogue, DLT technology is still vastly used merely in small, experimental projects. This
makes it hard to extrapolate what it could/would do, if applied to a scale business, either assuming
the trilemma (i.e. the right balance of decentralisation, scalability, security) will be solved at some
point or that permissioned-DLT can do so already. Our intense dialogue with industry experts helps us
to move from theory to practice in formulating the cost function of a (permissioned) DLT network,
featuring fixed costs which progressively handover to variable transaction costs as volumes pickup.
The Spunta DLT case validates our function, which embeds a margin of conservatism. Our conclusion
is that DLT technology could provide banks with astronomic cost efficiencies in some of their
businesses.
The treatment: evolution in cost miniaturisation to avert extinction and revive banking
We simulate the migration of European standardised banking onto a single DLT network, maximising
scale economies. Our simulation indicates the change would see a collapse vs current aggregate cost,
where MLAI and QC would also clearly contribute.
System-wide DLT costs could be c.400x larger than we estimated to still leave banks on current
profitability level, even after accounting for revenue attrition from digital disruption.
We conclude that potential for cost miniaturisation from the industrialisation of standardised banking
products/processes represents a thick buffer to absorb revenues pressures under very adverse
scenarios.
Hence – in the wake of the digital disruption the banking industry is about to experience – instead of
feeling threatened by extinction from fintechs/Big Tech, banks should embrace new technologies to
revive their standardised businesses. This mutation would not only grant their survival, but it could
possibly see them even thriving on low-cost structures applied to large-volume markets.
2 notes in 1 report: the 1st
for the experts, the 2nd
providing background knowledge
This report has been conceived for a very heterogenous reader, ranging from private market and public
market investors, industry experts, innovators, regulators, supervisors, authorities and people
passionate about innovation. We have therefore composed the report of two separate notes: 1) the
first eleven chapters for more digital-savvy readers, 2) the additional chapters are for those needing
background and deeper technical knowledge. In such case, it is probably better reading in reverse
order.
Credits and acknowledgements
Our deepest gratitude goes to the many friends, colleagues, experts and professionals who have
supported the investigation and knowledge development necessary to produce this report and to the
many banks which have contributed to our surveys. We hope your efforts have been worthwhile, we
have certainly enjoyed the interaction and learned plenty from it.
Further thanks go to the patience and passion of our readers and clients whose support is crucial to
allow us to continue developing our expertise and accompany along their investment journey.
RegObs Special Report
07 December 2022 ◆ 6
Where we left off
Our debut work on digital currency and CBDC outlined the types of digital currency and how it
came about. It explained the mounting pressure on western central banks to develop their digital
currency (CBDC) starting with the initially ignored pressure from the development of bitcoin, to
the much cleared threats of the Facebook-sponsored Libra/Diem to the advanced stage of the
Chinese CBDC, DC/EP. We elaborated around the intricacies of currency and geopolitics, outlining
how the US currently benefit from a hegemonic USD which – combined to SWIFT – allows them
oversight and influence over third countries via their sanction regime. Digital currency offers a
technological innovation making current systems potentially obsolete. For Europe, the ECB is
seeing the development of a digital euro as key to defend monetary sovereignty. Yet, depending
on the different depths of implementation, the digital euro could potentially destabilise the
current banking model, putting the ECB at the heart of the financial system, with the (desirable
for Brussels/Frankfurt) side effect of making a leap in European integration. Since our publication,
we have seen the successive events supporting our view: the ECB is proceeding steadily towards
the development of a digital euro. While US-China tensions continue, China has launched the
digital Yuan and the US are looking to recover the lost ground in the space, after having sunk
Libra/Diem, embracing the development of the d$.
CBDC, a digital currency with vast implications for geopolitics and financial stability…
Our report on the Digital euro (Digital euro: the ECB saving Europe again, 8 March 2021) explored the
digital currency world, the drivers behind the development of Central Bank Digital Currency (CBDC),
and the connection between currency and geopolitics. We identified in Facebook’s Diem/Libra initial
stablecoin project and in the advanced Chinese CBDC project the main reasons why most western
central banks are currently studying the development of CBDC. By more easily creating new, parallel
networks for payments, we argued that digital currency can hamper the current domain of the US and
of the US dollar on global trade and reserves - granted by the SWIFT messaging payment system and
by the Patriot Act – providing the US with oversight and coercion on third countries via their sanctions
regime. In this context, we argued that within the growing tensions between the US and China, Europe
is not well positioned on the future strategic assets: data, data networks, semiconductors, rare earths
and currency. Hence, the development of a digital euro is the only European strategic card at hand.
…possibly having positive side effects on the European integration process
In our digital euro report, we elaborated on the different forms the project could take, from a
“constrained” (i.e. with a cap), retail CBDC - more aligned to the initial ECB goals - which we deemed
compatible with financial stability, to an “unconstrained” one (i.e. no cap) - more aligned with
European Commission ambitions - potentially requiring a deeper rethink of how the financial system
operates. In the latter case, we concluded a large sum of banking deposits could migrate to the ECB,
which would in turn redeposit the sum into the banks, de facto mutualising risk across EU countries as
a side effect of equipping Europe with its own strategic asset.
The latest events seem to support our thesis
A lot has changed since March 2021, we would argue in support of our original thesis. We summarise
the main events:
• GEOPOLITICAL DOMINANCE: the Ukrainian war. The Russian invasion of Ukraine has
magnified the geopolitical considerations we made then, flagging Europe’s weak strategic
independence.
• EUROPE: integration speedup. We have had a string of elements showing Europe is
proceeding towards more integration, post Brexit, post pandemic and post Ukraine’s invasion:
 European army – The disorderly abandonment of Afghanistan by European forces,
dictated by the American ally without a discussion with EU partners, the recent EU-Russia
RegObs Special Report
07 December 2022 ◆ 7
controversy on natural gas supply and the AUKUS agreement with the US forcing France
out of a large submarine supply to Australia revived the debate over European military
autonomy for common defence and to safeguard European interest on the global map.
Italy’s President Mattarella, France’s President Macron, France’s Finance Minister Le
Maire and EU Commission President Von der Leyen spoke at the unison about the need
for Europe to think about gaining military autonomy (see Download 2021_09_16_EU policy
- EU agenda = strategic independence from theory to implementation.pdf).
 Energy Union – Germany is emerging from the ongoing Russia-Ukraine conflict as the
weakest link: depending on Russia for energy and on China for exports. Economic
sanctions to Russia have triggered Russian choking of the gas flow to Europe, testing
Europe’s resilience. Ongoing talks of energy solidarity entail European partners sharing
the pain on the risk of energy shortage, while progressive steps towards energy
diversification and independence are made over time. Finding an agreement on common
gas purchases, interconnection of networks and mutual assistance would be a massive
step towards a more integrated Europe, in our view.
 EU microchips – the lack of availability of semiconductors from Asia is eroding c.1/4 of
the European car production, with a normalisation expected not before 2023, confirming
the dependency of Europe on its largest sector by employment. The EU is working towards
boosting European manufacturing of microchips by 2030, doubling the market share at
global level from 10% to 20%.
 The recovery fund – in the end this project gravitates around digital investment and the
green transition, i.e. digital and energy independence.
EUROPE: advanced stage of the investigation phase of the digital euro. The ECB has spent the past
fourteen months investigating how the digital euro could look like. We see the following key features
having emerged from the process: privacy (no anonymity, no big brother), integration of supervised
intermediaries in the d€ distribution, inhibition of use of d€ as reserve of value and ECB-centred
settlement model. We still see the definition of the following features pending: transfer mechanism
and offline, architecture (DLT vs centralized), form factor (ECB app vs integrated within banks’ apps).
Politicians have recently embraced the project, materially raising its chances of launch, in our view.
CHINA: relentlessly progressing on its strategic path. Unsurprisingly, China is pursuing its objectives:
• Mounting pressure on Taiwan – China is pursuing its quest to gain more military relevance
by boosting investments and showing its muscles off. Hypersonic missiles, a larger navy fleet,
and the mounting pressure on asserting its sovereignty over Taiwan.
• Taking back control over digital payments and fighting crypto… – The overhaul of Chinese
digital payments is complete. From the failed IPO of Ant Financial to date, China has taken
back control over digital payments, previously firmly in the hands of the Wechat-Alipay
duopoly. Meanwhile, China banned cryptocurrency, confirming the tough stance on private
digital currencies. (Download 2021_09_24_Digital currency - The Regulators are marching in
(part 4) - China bans crypto.pdf).
• …replacing them with the digital Renminbi – China was fast in proposing the alternative to
the Alipay-We Chat digital payments duopoly by launching the digital Renminbi in Feb 2022.
(see article). This confirms China’s head start on the west on this front.
US: finally biting the bullet. Meanwhile, the US have seen the Biden administration take decisive
steps on geopolitics and digital currency:
• Shift to Indo-Pacific focus – The US retreat from Afghanistan has attracted the attention of
the global press and has been interpreted as the need to concentrate efforts on China. The
AUKUS pact shows the firm intentions of the US to form an alliance in the region to foster
RegObs Special Report
07 December 2022 ◆ 8
oversight over China. Taiwan has been at the centre of the US-China confrontation, reaching
the apex on the visit to Taipei of US House Speaker, Nancy Pelosi.
• Support to Ukraine – the US has stood firmly behind Ukraine from even before the Russian
invasion, warning about its imminency beforehand and through the supply of support,
weapons, intelligence and funds.
• Decisive action in digital currency regulation – The Biden administration removed the prior
ambiguity in the regulation of cryptocurrencies. On one hand, the US stopped for good
Facebook-sponsored Libra/Diem and the SEC stepped in on Coinbase’s Lend initiative and
asked all players in the space to consider application to SEC regulation. On the other, and
more importantly, the Administration put the development of a digital dollar as a strategic
priority.
LIBRA/DIEM: game over, not so for private stablecoins. We have been arguing for a long time the
initial Libra/Diem project was indigestible for regulators but that the Facebook initiative had merits
on its technical advancements (see Download 2021_05_27_Digital currency - DIEM turnaround from the
worst enemy to the central banks' best friend.pdf). Diem attempted a sudden u-turn, in vain:
1. Abandoning Switzerland to return home – Diem left the Swiss venture returning to the US.
2. Teaming up with a Fed-regulated Californian bank to issue digital currency – It teamed up
with a Fed-regulated Californian bank (Silverbank) to issue a USD stablecoin to back the USD-
Diem (see Download 2021_05_13_Digital currency - Carpe USDIEM - Uncle Sam calling
home.pdf).
3. Opened up to abandoning the stablecoin in favour of CBDC – At a livestreaming which we
hosted back in July-21, Christian Catalini, co-founder of Diem, confirmed that they would be
prepared to abandon the stablecoin in favour of CBDC if and when central banks would be
ready for it, offering DIEM platform as a partner for the CBDC project. (see Download
2021_07_08_Digital currency - Co-founder of Facebook sponsored DIEM confirmed
repositioning.pdf, Youtube link, Download 2021_06_24_Digital currency - ECB proceeding
steady vs Diem's regulatory flirting__1.pdf).
DIEM was sold to Silvergate bank in January 2022, ending Facebook’s journey in digital currency. This
unhappy ending does not mean that all private stablecoins should suffer the same fate. Big Tech is
working to find the most convenient solution to benefit from this new instrument.
CRYPTO: US rates spoiling the party. The digital currency world is in a crunch, started by the new
USD rates cycle and the FED’s balance sheet shrinkage which is drying the ample liquidity and trigger-
happy investment into DeFi. Over the past period we have gone from “limitless partying” to “crypto
winter”, i.e. from a phase of proliferation to one of harsh rationalisation in crypto land, even stained
by cases of fraud.
RegObs Special Report
07 December 2022 ◆ 9
New technologies: evolution or extinction?
This report flips the approach to technological implications for banks upside-down vs its digital
euro predecessor, i.e. from a top-down view to a bottom-up perspective. Here, we focus on the
disruptive potential of three new technologies: 1) artificial intelligence/machine learning (MLAI),
2) distributed ledger technology (DLT) and 3) quantum computing (QC). We imagine how these
could redefine the way banks are and operate, or if instead they could disrupt banks to the point
of ridding of them in full. Borrowing from the often-used bank/dinosaur metaphor, are new
technologies the asteroid bringing banks to extinction or could banks not just embrace them and
jump on the asteroid, i.e. evolving rather than succumbing to the innovation tsunami?
The report bases on three pillars: 1) two surveys to take a snapshot of where banks are on their
technological transformation, 2) our diagnosis of how technology could transform parts of the
banking businesses and our take on the final outcome for survival (profitability), 3) in-depth
background knowledge on the prior cases of technological disruption and on the three
technologies.
The first step in solving a problem is recognising it. Banks are progressively becoming more
conscious of it but are yet scrambling both what to do about it and when. Before an existential
threat, there should be no taboos. At the same time, every threat hides an opportunity: banks
need to identify it and pursue it. Getting there first usually pays dividends…
From external threat…
Our digital currency work identified this new tool as a serious potential threat for commercial banks,
particularly in the form of central bank digital currency (CBDC), where central banks can potentially
become competitors not only on the funding side. This is a case where technological innovation can
disrupt existing, stable, regulated sectors and markets.
…to internal rethink
From time to time, the advent of new technologies allows the redefinition, re-engineering and re-
thinking of companies, sectors, business models and processes. We have several evidences of this in
airlines, local taxis, music distribution, retail, etc. For banking, the advent of the internet was first
predicted to extinguish banks, which then embraced change and are on course to rebalance their
distribution channels towards remote/digital ones vs traditional branches. Technological
transformation represents another challenge as it has the potential to be more of a revolution than an
evolution, testing the survival of the species, meaning that processes, products – and potentially even
banks as a whole - could require deep rethinking rather than adaptation.
MLAI, DLT, quantum computing in focus
We focussed on the main technological innovations, namely artificial intelligence/machine learning
(MLAI), distributed ledger technology (DLT) and quantum computing (QC), which we describe in depth
later in the report. They have very different stages of development, business implications and
requirements to be implemented (see DLT, MLAI and QC for in-depth background material).
MLAI - MLAI can be thought of as an increasingly less supervised spectrum (“Symbolic” up to “Strong”)
of processes to be operated and executed by machines. However, while thoughts immediately fly to
Terminator, we are still scrambling on the ground. Banking is currently positioned at the weaker end
of the spectrum, such as “ChatBots”.
DLT - Distributed ledger technology (DLT) - more commonly known as blockchain – is a tool to provide
trust between parties relying on the same data records, without the need for a centralised, third-
party authority. The immense data management potential introduced by the technology contrasts with
the very limited adoption by banks thus far due to their adversity to cooperation and shared control.
RegObs Special Report
07 December 2022 ◆ 10
QC - Quantum computing is a new technology taking computational power to a new dimension,
leveraging on quantum physics. We are still in the development stage of the technology, so that it is
just too early to see its adoption by the industry.
Banks need to perceive the risk
The first step to solving a problem is to recognise it. Banks are more and more becoming conscious of
the threat, but they have not quite worked out what to do about it and to what degree of urgency, in
our view.
Preservation instinct means there is no room for taboos
Preservation instinct kicks in when a creature feels its life is in danger. This puts everything else
behind in terms of priority. In this parallel, banks should consider making radical changes to how they
look like, how they work and what they do, to adapt to the new environment. If technological
innovations are bringing a change of ice age, studying them should be urgent, so to have time to adapt
to the new ecosystem.
Every threat hides an opportunity
We believe technological disruption represents both a threat and an opportunity for banks. The novelty
new technologies introduced is to be able to manage vast amounts of activity at much higher efficiency
levels. On one hand, they allow for a material increase in competition – both within the industry and
from newcomers – likely eroding margins and leading to consolidation. On the other, the increased
transparency which comes with them could trigger a positive volume effect and a radical cost
reduction. Getting there first usually represents a material advantage…
RegObs Special Report
07 December 2022 ◆ 11
MB survey suggests A.I. and cost efficiency in focus…
To better understand the current state of play when it comes to adoption of new technologies in
the banking sphere, we enquired European banks on two fronts. The first – our “benchmark
survey” – is more simplistic and standardised, helping us to map out an industry standard. We
thank the c.20 participating banks which, for the sake of anonymity, we won’t name. The second
- our case study almanac – is described in the next chapter. We summarise our findings in a
synthetic aggregate benchmark embodied by our “Banca Benchmark”, a large, fictional European
bank: 1) Banca Benchmark spends a somewhat modest €0.5bn per year on developing applications
that use DLT/MLAI/QC technologies, although the overwhelming majority (>80%) is going into AI
(mostly chatbots). 2) Banca Benchmark mostly aims at better cost efficiency, with revenue
enhancement and customer experience also important, but to a lesser extent. 3) Banca
Benchmark conceives Fintechs as partners in the digital journey. 4) Banca Benchmark believes
that adoption of these technologies will not just cut costs but change the entire role that banks
play within the industry, emblematic perhaps of a transformation of the whole banking business
model, with Risk Management & Control, Digital Channels and Back Office & IT most exposed to
this transformation. 5) Banca Benchmark is more focused on innovating back office functions and
cutting costs rather than transforming products and the way they are conceived and offered.
MB-designed benchmark and a use-case surveys to map EU banks tech activity
Looking primarily at DLT/MLAI/QC technologies we see as defining the new banking landscape, we
wanted to understand better what banks had already been doing (or were planning on doing). As such,
we queried banks along two directions. The first was through a survey designed to build a benchmark
across our respondent banks, we refer to this as our “Benchmark Survey”. Secondly, we asked banks
to showcase their successes which we aggregated in our almanac in the next chapter.
The Mediobanca Innovative Technology Benchmark Survey
Our benchmark survey consisted of six questions focused primarily on gauging the size and allocation
of the investment budget into the three technology groups, as well as what part of the bank was likely
to be most impacted by the adoption of the technologies. Whilst not mentioning any bank by name,
we share the results on an aggregated basis across the c.20 respondent banks.
Banca Benchmark
We aggregate the responses received into a fictional Banca Benchmark, representing EU banks. This is
strong of c.€250bn market capitalisation, investing a modest €0.5bn a year on A.I./machine learning,
quantum computing and DLT, with an overwhelming dominance of A.I. and mainly for risk
management, digital channels and back office purposes. The main goal is usually cost saving, with
projects more and more in partnerships with fintechs.
RegObs Special Report
07 December 2022 ◆ 12
Picture 1: Benchmark – Typical response from the benchmark survey (embodied by “Banca
Benchmark”)
Source: Mediobanca Securities
€0.5bn annual budget appears small
Based on size of the amount of budgeted investment into the three technologies and investment
horizon, we detected an aggregate €0.5bn annual budget. On one hand, this hopefully reflects more
carefully investments in the three technologies (DLT/MLAI/QC) as opposed to more generic IT budgets.
On the other, this suggests that on average banks are not yet investing massive amounts in this
direction.
Cost efficiency and machine learning the focus
We asked respondents how they were splitting their investments across the different technologies and
the ultimate purpose of the investment. On average, nearly one-third of the investment was directed
at cost efficiency. Whilst, around 20% was aimed at boosting revenues, enhancing customer experience
and improving operational risk. How this was split across the different technologies was much more
skewed, with c.80% going into MLAI. The remainder was split 15% into DLT and just 3% into Quantum.
Figure 1: Investment split (purpose, %)
Source: Mediobanca Securities
Figure 2: Investment split (technology, %)
Source: Mediobanca Securities
Market Cap: €252bn
Investment (€mn): €0.5bn/year
Investment (split):
AI/ML 82%
DLT 15%
Quantum 3%
Main part of bank: Risk management/control
Digital Channels
Back office/IT
Main goal: Cost sav ings
View on FinTechs: Partners
Banca Benchmark
30%
22%
21%
17%
10%
Cost efficiency Revenue Boost Customer Experience
Fraud, Operational Risk Other Risk Management
56%
26%
15%
3%
Machine Learning Automation DLT Quantum
RegObs Special Report
07 December 2022 ◆ 13
Yet, there are still considerable differences between individual banks
Although there is little doubt most banks see cost efficiency savings and AI (automation and machine
learning) as the foremost purpose/technology where they are investing this budget, there is still a fair
bit of heterogeneity across the different banks.
For example, Bank 2, 3, 12 and 16 are pretty much all bang on average in terms of purpose of their
investments. Contrast this with Bank 13, 9, 4 or 14, for example, where customer experience, cost
efficiency, fraud/operational risk or risk management are much more important features than for the
rest of our sample.
Picture 2: Investment split (purpose, %) - distribution from the mean (=0) for each bank in the sample
Source: Mediobanca Securities
In terms of the split across the different technologies Bank 2 and Bank 3 are again pretty much in-line
with the consensus. Bank 4 meanwhile has put a much higher weighting on Quantum investments than
for the rest of the sample and bank 14 is investing primarily in automation.
Picture 3: Investment split (technology, %) – distribution from the mean (=0) for each bank in the sample
Source: Mediobanca Securities
FinTechs seen as partners. New role for banks post-adoption of these technologies
Our survey suggests banks are not scared about FinTechs sending them to extinction. As we transition
to this new banking landscape, banks were overwhelmingly positive about the role of FinTechs in that
journey. Nearly 90% view them as partners, as opposed to either rivals or just temporary players. Once
widespread adoption of the technologies is realised, banks unsurprisingly see better cost efficiency as
a major outcome (in-line with what they had said previously about the main goal being to cut costs).
Perhaps more interesting, is that even more of our sample of banks recognise that these technologies
are likely to fundamentally change the business model and the overarching role banks play within the
-2
-1
0
1
2
3
Bank 1 Bank 2 Bank 3 Bank 4 Bank 5 Bank 6 Bank 7 Bank 8 Bank 9 Bank 10 Bank 11 Bank 12 Bank 13 Bank 14 Bank 15 Bank 16
Distance
from
mean
(Z-Score)
Cost efficiency Revenue Boost Customer Experience Fraud, Operational Risk Other Risk Management
-2
-1
0
1
2
3
4
Bank 1 Bank 2 Bank 3 Bank 4 Bank 5 Bank 6 Bank 7 Bank 8 Bank 9 Bank 10 Bank 11 Bank 12 Bank 13 Bank 14 Bank 15 Bank 16
Distance
from
mean
(Z-Score)
Automation Quantum DLT Machine Learning
RegObs Special Report
07 December 2022 ◆ 14
industry. It goes beyond just cutting costs, or better risk management. Fears over the possibility of
revenue erosion are further down the pecking order.
Figure 3: View on FinTechs
Source: Mediobanca Securities
Figure 4: Largest impact of these technologies for banks
Source: Mediobanca Securities
Risk management, digital channels & back office most susceptible to tech disruption
Finally, we asked our recipients what part of the bank they would see as most impacted by the
adoption of our three technologies. There were three areas that were clearly identified as the most
susceptible: risk management & control; digital channels; and back office and IT. With each being
included in the “Top 3” by 11 of the sample of 18 banks that responded to this question.
Picture 4: Part of the bank most impacted by technological disruption
Source: Mediobanca Securities; *18 banks (rather than 16) responded to this question.
89%
5%
6%
Partner Temporary Player Contender
50%
39%
11%
0% 0%
0%
10%
20%
30%
40%
50%
60%
Role of banks in
the financial
industry
Cut costs Improve
fraud/operational
risk
management
Erode revenues Improve other
aspects of risk
management
11x 11x 11x
7x
6x
3x
2x 2x
0x 0x
0x
2x
4x
6x
8x
10x
12x
Risk
management
and control
Digital channels Back office and
IT
Payments and
transactions
Securities and
capital markets
Credit granting Physical
network
Financial
advisory
Deposit
gathering
Treasury
No. of respondent banks that included within their "Top 3"
RegObs Special Report
07 December 2022 ◆ 15
…seconded by our “use case” almanac
The second survey we sent out was to map the “use-cases” banks have been working on as
anecdotal evidence of their interest and successes. We give full details on the individual cases
here (where we have permission to disclose). We aggregate the use cases in a fictional “Banca
Case”, composed of 26 cases across 14 respondent banks. The majority (c.70%) of Banca Case
relates to MLAI (mostly “Chatbot”/Virtual Assistants), ¼ in DLT and just a couple in QC. In general,
they were built for internal use rather than to be used alongside other peers. Most cases were
developed in Partnership with Fintechs/technology providers, although internal resources were
also (at least partially) employed in ½ of cases: usually around 4 FTEs. The journey from inception
to production was short (279 days), and on average each use case had now been in active operation
for a similar time. Consistent with Banca Benchmark, Banca Case’s goal tended to be cost
efficiency where, for a budget of around €1.3m, a return of €1.9m was expected. Digital Channels
and Back Office & IT were the departments most impacted, followed by Securities & Capital
Markets and Risk Management & Control, with far fewer use-cases related to Payments than
indicated in the Benchmark survey.
Use Case Survey
Our use case survey was filled out by 14 respondent banks, which gave us a total of 26 existing use
cases for the three investigated technologies. It was designed to be less restrictive than the benchmark
survey, with more freedom for respondents to go into greater detail on one (or more) use cases. Yet,
it is still possible to make some general observations about the types of use cases that our banks have
implemented.
Banca Case: our aggregate benchmark
We aggregate the responses from the use cases survey into a fictional Banca Case, representing EU
banks. This is strong of €205bn market capitalisation and showcasing 26 use cases. 2/3 of these apply
MLAI, ¼ DLT and c.1/20 QC, with an overwhelming cost cutting scope mainly for digital channels and
Backoffice. ¾ of use cases have involved a partner, with 50% internal development and 50%
outsourcing. c.2/3 of cases involve standalone adoption and 1/3 adopt a cooperative approach. The
26 cases have had an average time-to-production of 280 days, €1.3m investment and €1.9m return.
Picture 5: Use cases - snapshot of the findings from the use cases sent back by banks
Source: Mediobanca Securities
Use-cases dominated by MLAI applications, but many interesting DLT applications also in pipeline…
Given the overwhelming majority under our benchmark survey stated they were investing in AI
technologies, it is of little surprise that most use cases were also based on this technology (70% of
total use cases). Although, it was also clear that there was a bias towards classifying the use-case as
Technology split: AI/ML (66%); DLT (28%); Quantum (6%)
Partnership: Partner (77%); No partner (23%)
Development: Internal (50%); Outsourced (50%)
Main user(s): Stand-alone adoption (65%); Co-operativ e (35%)
Time to production: c.280 days
Budget: €1.3mn
Return: €1.9mn
Expected ROI: 40-50%
Main purpose: Cost efficiency
Main part of bank: Digital Channels
Back Office/IT
Use Cases
RegObs Special Report
07 December 2022 ◆ 16
machine learning, even when it sounded much more aligned with the technical definition of
automation. Within AI, the largest proportion of focus areas were for some form of ChatBot application
(CallBot, Virtual Assistant, VoiceBot) with a number of banks providing use cases in this area.
DLT-based applications made up around a quarter of the overall use-cases. Here we saw quite varied
applications, with very little cross-over between banks. As we illustrate below, this includes post-
trade services, Bitcoin trading/custody and digital surety amongst others.
There weren’t many use cases for Quantum technology, with this making up just 6% of the overall use-
case examples we received. As we stated previously, the technology is still very much in its infancy
with the focus more on partnering with FinTechs and BigTech organisations in advance of the eventual
Quantum revolution. What has been done so far is targeted at more efficient, quicker and more
comprehensive dynamic portfolio optimisation.
Picture 6: Area of focus for use-cases, split by type of technology employed
Source: Mediobanca Securities
…whilst typically built with a partner (sometimes supported by internal staff) and designed for
standalone use by developing the firm rather than in co-operation with peers
In 3/4 of our sample (77%), these use cases were built either entirely or partially with the support of
a partner, with around half of the use cases completely outsourced to external organisations. Where
the use case was built internally, it was implemented on average by the equivalent of 4 FTEs. Finally,
we highlight that whilst the majority were built to be used solely by the respondent bank, about 1/3
were designed to be used co-operatively with other organisations.
Picture 7: Use cases mostly built with a partner and designed for stand-alone use
Source: Mediobanca Securities
280-day time to production
On average, it took each use case around 280 days to move to the production phase. This works out to
a little over a year given c.250 trading days in a year. As we demonstrate below, it consists of around
3-4 months (based on 20 trading days in a month) for each of the initial exploratory analysis and proof-
of-concept phases. The development phase then takes a little longer at around 4-5 months (although
likely slightly exaggerated by a couple banks consolidating the analysis and POC phases into the
development phase). The pre-production phase then takes a further 3 months, on average.
Automation Machine Learning DLT Quantum
Back-office processing Foreclosure Post-trade serv ices Dynamic portfolio optimisation
Employee Q&A MIFID compliance Bitcoin trading/custody Asset allocation
Loan decision making Chatbot Trav el insurance
Financial health analysis Callbot Digital surety
Structured product issuance DCM book building Cash-for-collateral lending
Legal email automation Salary-backed loans
Checking reciprocal accounts
Record of carbon credits
RegObs Special Report
07 December 2022 ◆ 17
At the time of receiving survey responses, use cases had been operational for a little over a year on
average. As such, we would point out that whilst most of the use cases we have been provided are
fairly newly implemented, banks are very likely also working on other projects using these technologies
that are in pre-production phase and not disclosed, yet.
Picture 8: Breakdown of days for each stage of production
Source: Mediobanca Securities; *POC = Proof of Concept
Cost efficiency the dominant purpose of use-cases, customer experience also important
In our use case survey, we also asked respondents to tick which of the below five options were a
targeted benefit of the use case. They could tick as many as deemed pertinent. Once again, we saw
cost efficiency come out on top, where around two-thirds of use cases flagged this as a purpose.
Customer experience was also considered an important feature in the use cases, with half mentioning
this as a purpose. Revenue boost and other risk management were both ticked in nearly a third of the
total use cases. Fraud and operational risk management was relevant to c.20% of all use cases.
Picture 9: Percentage of use cases that ticked this purpose as one of their options
Source: Mediobanca Securities
Estimated return of around 40-50%, but some doubts on consistency of data inputs
One aspect of the survey where we had initial doubts about the quality of responses was with regards
to asking for absolute numbers on the budget and projected return of the projects. The outcome of
our survey was that on average, around €1.3mn was budgeted for each of the use cases. Meanwhile,
the anticipated return was on the order of €1.9mn. The way these numbers were reported makes
identifying whether they were on an annualised or overall basis ambiguous. Assuming simply that both
65%
50%
31% 31%
19%
0%
10%
20%
30%
40%
50%
60%
70%
Cost efficiency Customer experience Revenue boost Other risk mgmt. Fraud/Op. Risk
% of total use cases that ticked this as an option
RegObs Special Report
07 December 2022 ◆ 18
were provided on an overall basis, we get a basic implied return of 40-50%. However, we would suggest
taking this with a big “pinch of salt”.
Digital channels & back office again most important, but shake-up in rest of pecking order…
Lastly, we asked banks what part of the bank they thought would be most revolutionised by their use
case. Given that we had previously asked the same question to banks in relation to the three
technologies during our benchmark survey, we found it interesting to compare the two. Digital
Channels and Back Office & IT were consonant with what we had found with our benchmark, where
both were considered the most transformed (10 use-cases out of 26 mentioning them). Interestingly,
Risk Management & Control was significantly less relevant to our use cases vis-à-vis with the
benchmark. Where Securities & Capital Markets, in particular, took its place for the use-cases. As a
final point here, we stress the more marginal proportion of banks referring to Payments & Transactions
in their Top 3 for the use-case survey.
We’d suggest that a logical possibility here is that Risk Management & Control or Payments innovations
are further away from the production stage for most banks yet are both areas where they are investing
considerable portions of the overall transformation budget.
Picture 10: Part of the bank most revolutionised by each of the use cases (versus benchmark survey)
Source: Mediobanca Securities; *Two axis are not comparable: lbs is out of possible 26x and rhs is out of possible 18x.
0x
2x
4x
6x
8x
10x
12x
0x
2x
4x
6x
8x
10x
12x
Digital channels Back office and
IT
Securities and
capital markets
Risk
management
and control
Financial
advisory
Credit granting Physical
network
Payments and
transactions
Treasury Deposit
gathering
No.
banks
that
included
within
their
"Top
3"
No.
use
cases
that
included
within
their
"Top
3"
No. of use cases that included within their "Top 3" (lhs)
No. of banks in benchmark that included within their "Top 3" (rhs)
RegObs Special Report
07 December 2022 ◆ 19
Good effort, but far from transformative innovation yet
Our two surveys show coherent results, broad enough to depict the current location of banks along
their innovation journey. While banks are trying hard to innovate, most of their efforts are going
in a linear evolution of the current operating model, with heavy investment in chatbots as the
“new automation”, not quite a technological revolution yet, in our view. Short inception-to-
production timing, the high usage of external staff and the focus on efficiency-oriented projects
can both reflect good project management and cost discipline and room to increase ambition along
the innovation vector. There is something to be said here about the current career incentives for
innovation leaders: showing short term results plays against truly transforming multi-year
projects. Governance likely also plays a role in the low penetration of DLT, in our view, reflecting
the low propensity to share and coordinate with peers/competitors and the need to control the
project/investment. The absence of payments reflects the exit of this lucrative scale business by
banks, potentially a strategic mistake in the long term and a drawback for the overall innovation
know-how, in our view. Finally, it is just too early for QC to take room, but it will come. The
overarching conclusion is that banks know they must change, but they are doing so too slowly and
linearly, in our view.
Coherence between Mediobanca’ s surveys
Our two surveys provided coherent results, comforting us on the overall picture these describe and
hence allowing us to draw some preliminary conclusions on where European banks are on their
innovation journey now.
Trying hard…
We see banks are trying hard to innovate, with a multitude of different initiatives, dedicated
structures, more and more experiments and a growing attention to the innovation world.
…good inception to production timing…
The surveys confirm banks have short inception-to-production timing, showing project management
efficiency, but also, possibly, suggesting they could increase their ambition and aim for more complex
targets.
…but not daring enough, yet
70% focus on chatbot development suggests to us that the current focus is on moving further on
efficiency and the evolution of the automation process of the past decades. More ambitious projects,
potentially rethinking the current business model and production processes, only represent the fringe
of innovation investments, at best, for now.
Low DLT penetration: cooperation is the real hurdle
We find a particularly low penetration of DLT initiatives in the innovation pipeline. In our view, the
main reason for this is intrinsic to DLT, i.e. the necessity to open and cooperate with competitors. We
find a focus on projects which can be fully controlled and owned by the individual banks. This approach
is a hurdle to the development of effective DLT-based initiatives, which instead require the creation
of shared networks with win-win characteristics for networks participants.
More internalisation to own and develop
Banks seem to rely heavily on external providers/consultants for their innovation projects. The more
the internalisation of the staff involved, the more the ownership and the ability to give birth to
concatenated developments which can transform products/processes over time. A cultural shift based
on internal staff is needed for change to take root and flourish.
RegObs Special Report
07 December 2022 ◆ 20
Efficiency and digital channel/IT focus rather than rethinking products/processes
Our surveys highlighted the banks’ focus on efficiency and on digital channels/IT. These projects tend
to be easier to justify and provide a faster payback time, to the advantage of the sponsors who need
to show results. Yet, transformations take time and are hardly plausible through small, cost-cutting-
focused projects. The issue with this is that multi-year programs tend to work against career progress.
More innovative proposals require a change in the incentive structure of innovation leaders, we would
argue.
Payments are out of scope
Payments seem to fall out of the scope of banks. More and more banks have abandoned this very large
and lucrative space, to the advantage of specialised payment companies. Yet, payments remain a core
product provided by banks and one which is at the forefront of the digital revolution.
It is just too early for Quantum Computing… but it will come
If we had to rank the technologies by stage of implementation/maturity, we would place MLAI up top,
DLT runner-up and QC in third place, at a distance. The supreme computational power this technology
could provide would come in handy for the transition of banks into data companies, but we are just
not there yet, and we understand why banks place it last in the priority list. It will come, but it is just
too early now.
Pass mark for the effort, but much more is required to drive a mutation in the species
Our surveys show banks are experimenting, they are active, they are getting their hands dirty to learn.
Yet, the horizon is still too narrow and the ambition too low for us to see the intention to radically
change course, mutate, change skin. We are still in a slow, risk averse and mistake averse attitude.
The perception of threat is likely still very low, and the survival instinct is yet to kick in for banks to
understand faster, more radical change (a mutation) is due for the perpetuation of the species.
A vision of how to get more ambitions in the coming chapters
We will argue what the mutation could be and how what the path to survival is over the next chapters
of the note.
RegObs Special Report
07 December 2022 ◆ 21
“breaking the bank”: isolating the species at risk
New technologies are likely to disrupt businesses/products which are highly standardised. This
chapter executes a surgical procedure - never attempted before – to separate the standardised
businesses from the customised businesses of the twelve most representative European banks.
This predicates on the assumption that standardised businesses are more porous to re-engineering
on new technology than customised businesses. We identify 2/3 of operating income, 60% of costs
and 50% of loans belong to standardised banking.
New technologies (such as DLT, AI and quantum computing) are usually seen as a threat to the status-
quo, as a rapid adoption in various sectors of the economy could have the potential to change current
rules of the game. However, this would not necessarily mean that they are threats for current players
if these are quick enough in adapting and improving their own business models. In this context, new
technologies can be seen as (inevitable) enablers of future growth. Banks are not exempt from this,
and emerging technologies could change how things are currently made. Clear examples could be:
 Industrialise/streamline credit granting - A common DLT platform could facilitate, accelerate
and optimize the process to grant credit to customers. Real estate transactions involve a large
number of legal, financial and real-estate intermediaries acting for the buyer, seller and
lender, each adding fees and time to the transaction. A shared/permissioned blockchain for
mortgage loans could be a powerful application for DLT technology.
 Revolutionise the liability structure - The adoption of a specific type of CBDC could change
banks’ funding structure (see our report on Digital Euro);
 Democratise the exchange of securities - The tokenization of financial assets could change
the way they are distributed to clients and how they are managed by banks (i.e. banks could
need less FTE to manage the same products/clients through the implementation of smart
contracts). They could theoretically even disintermediate banks in full via direct P2P
transactions or perhaps leaving banks as custodians.
 Take efficiency to another planet - Could further increase the efficiency of banks by reducing
costs and increasing volumes.
Standardisation vs customisation
Yet, the transformation would require banks to invest massively in new technologies to adjust their
business model for all these potential changes. Moreover, we believe that the more a financial product
is standardised, the more it is exposed to technological change, while customised products, i.e. tailor-
made products with a large component of human advisory/interaction, are less exposed to
technological disruption.
RegObs Special Report
07 December 2022 ◆ 22
Picture 11: banks’ product matrix based on high/low impact from new technologies
Source: Mediobanca Securities
In the following section, we try to identify which businesses are more exposed to this, and which could
be the impacts to volumes, margins, and costs.
Standardised products account for 2/3 of operating income
Creating the European megabank to dissect its business model
We dissected the business models of the 12 most representative banks in Europe, which together
represent >€480bn in market cap or c€15trn in total assets, across eight European markets: France,
Spain, Italy, UK, Germany, Switzerland, Nordics & Benelux. In other words, our sample is composed of
the major European investment, commercial and diversified banks. From here, we generally refer to
this as the “aggregate”.
Net interest income & fees were split into “standardised” and “customised” buckets…
Our main goal is to understand which business of our banks’ sample can be challenged, improved, or
even substituted by new emerging technologies and players. In the last decades, banks have continued
to grow and diversify their business models by adding new sources of revenues and shrinking their cost
base to increase profitability. They have expanded in asset management, insurance and strengthened
their services to corporates by offering more sophisticated and tailor-made products. We believe new
emerging technologies could represent a threat for some core businesses of banks. To identify which
business is more exposed to this trend, we split the banks’ main sources of revenues (NII and fees) in
two categories: standardised and customised businesses as a proxy of exposure to technological
disruption:
i) standardised, i.e. businesses based on a standardised model across the banking sector,
in which there is little customisation of the services provided to clients (clear example
are mortgages, consumer loans, payment and FX services, mass market of asset
management and insurance products);
ii) customised, i.e. all businesses where services are generated on an ad hoc basis, tailored
on client needs (clear example are corporate and SME loans, private banking and wealth
management services, corporate finance and advisory services).
For the NII decomposition we used the data from the H121 EBA transparency exercise. By taking the
geographical exposures of each bank for each category (i.e. mortgages, consumer, corporate and SME
loans), and applying the average interest rates of each country for each exposure, we estimate a
Mortgages
Consumer loans
Payments
FX transactions
Brokerage
Underwriting fees
Distribution of insurance products
Distribution of saving products
Corporate and SME loans
IB services
Private banking fees
Wealth management fees
Insurance products
Standardised
Customised
Advisory services
Potential
impact
from
new
technologies
Low
High
RegObs Special Report
07 December 2022 ◆ 23
theoretical interest income generated by standardised and customised businesses. Then we applied
these proportions to the 4Y average interest income from loans of each bank in our sample. All other
sources of income in the NII have been equally split among the four categories.
For the funding side, we assume that mortgage exposures are financed by covered bonds (taking the
average yield of each country’s covered bond curve), consumer loans by securitizations or senior non-
preferred securities, while the cost of funding for corporate and SME loans is based on the difference
between total interest expenses and the cost of funding to finance mortgages and consumer loans.
While the loan book is evenly split, NII is skewed more towards standardised products
We classified mortgages and other personal loans as “standardised” products, whereas corporate and
SME lending is considered “customised”, needing a more personalised approach making it more
difficult to industrialise. Based on this classification, the loan book of our aggregate bank was pretty
much split 50/50. Yet, NII was much more skewed towards the standardised products (c.2/3) versus
1/3 from customised businesses.
Picture 12: Composition of aggregated loan book
(standardized vs customized, %)
Source: Mediobanca Securities
Picture 13: Composition of net interest income
(standardized vs customized, %)
Source: Mediobanca Securities
Fee income also tilted more towards standardisation (c.60%)
As for fees, we have assumed that asset management fees, insurance distribution, custody, brokerage,
payments, underwriting, and FX fees are all standardised revenues. Meanwhile, any fees related to
wealth management, corporate finance, advisory and in-house insurance factories are customised
revenue streams. On this basis, we found that roughly 60% of banking fees relate to standardised
products/processes, with the other 40% being customised.
Table 1: Revenues decomposition: standardised vs customised
Revenue
decomposition (€bn)
%
NII [standardised] 94 66%
NII [customised] 48 34%
NII [total] 142 100%
Fees [standardised] 57 59%
Fees [customised] 40 41%
Fees [total] 97 100%
Revenues [standardised] 151 63%
Revenues [customised] 88 37%
Revenues [total] 240 100%
Source: Mediobanca Securities
Mortgages
38%
Consumer
credit
10%
Corporate
39%
SME
13%
Mortgages
36%
Consumer
credit
30%
Corporate
22%
SME
12%
RegObs Special Report
07 December 2022 ◆ 24
…with associated costs backed out from the C/I ratio of specialised firms/subsidiaries
Now that revenues have been split between standardised and customised pools, we needed a way to
identify the associated costs. To do this, we constructed a representative sample of either highly
specialised firms/pure players or business units that were focused on each of the various revenue
sources. For instance, for Asset Management fees we took Amundi, DWS, UBS Asset Management,
Schroders and HSBC Asset Management. Where we then took the average cost-income ratio to estimate
the costs associated with each of the revenue streams, giving us a way to obtain the standardised and
customised costs (and ultimately how the pre-provision profit was split across standardised and
customised).
Suggestive of a roughly 60%/40% split on costs between standardised/customised
Our assumed cost-income ratio ended up at around 53% for both standardised and customised products.
Given most revenues came from standardised sources, naturally we also saw this was the main driver
for costs at around 60% of the total.
Picture 14: Assumed cost-income ratios per activity (derived from samples of pure plays)
Source: Mediobanca Securities, Benchmark companies: SAN UK – Retail, LLOY – Retail, ABN – Retail, NWG – Retail, SAN US, BARC
CCP, ACA Consumer Credit, BKT CB, ISP CB, DBK CB, ABN CB, LLOY CB, AMUN, DWS, UBS AM, Schroders AM, HSBC AM, ACA
Insurance, BNP Insurance, ISP Insurance, CABK Insurance, KBC Insurance, HSBC Insurance, BNP Security Services, ACA Asset
Servicing, BNY Mellon Securities Services, NEXI, ACA Financing Activities, BNP GM, GLE GMIS, GS GM, ACA Wealth, ABN PB, BNP
Global Banking, Rothschild, ACA Leasing & Factoring.
c.2/3 of operating income is coming from standardised streams
Putting our P&L back together, we compute that 65% of operating income relates to standardised
products and services, whilst only around 35% are customised. Across our sample we also see a fair bit
of deviation, ranging from as much as 85% of operating profits coming from standardised, down to 43%.
81%
69%
61%
49%
48%
44%
43%
29%
80%
61%
58%
57%
55%
54%
51%
51%
44%
23%
0% 10% 20% 30% 40% 50% 60% 70% 80% 90%
Custody/Trust
Markets
Asset Management
Consumer banks
Mortgage banks
Underwriting
Payments/FX
Insurance
Wealth Management
Other Asset Management
Commercial banks
Corp. Fin/ Advisory
Factoring
Other
Guarantees/ Commitments
Loans
Corporate banks
Insurance/Other brokerage
RegObs Special Report
07 December 2022 ◆ 25
Picture 15: Operating profit split into standardized and customized buckets for our sample
Source: Mediobanca Securities
Table 2: Aggregate bank PBT decomposition - standardised vs customised (€bn)
Simulation %
NII [standardised] 94 66%
NII [customised] 48 34%
NII [total] 142 100%
Fees [standardised] 57 59%
Fees [customised] 40 41%
Fees [total] 97 100%
Revenues [standardised] 151 63%
Revenues [customised] 88 37%
Revenues [total] 240 100%
Expenses [standardised] -74 61%
Expenses [customised] -47 39%
Expenses [total] -121 100%
Pre-provision profit [standardised] 77 65%
Pre-provision profit [customised] 42 35%
Pre-provision profit [total] 119 100%
Source: Mediobanca Securities, company data, EBA
85%
80% 77%
71%
65% 64% 63% 58%
49% 49% 46% 43%
65%
15%
20% 23%
29%
35% 36% 37% 42%
51% 51% 54% 57%
35%
0%
20%
40%
60%
80%
100%
Bank
6
Bank
1
Bank
12
Bank
11
Bank
5
Bank
4
Bank
7
Bank
10
Bank
3
Bank
9
Bank
2
Bank
8
Aggregate
Standardised Customised
RegObs Special Report
07 December 2022 ◆ 26
Digital disruption carries 30-50% margin compression…
We look at prior sectors disrupted by technological transformation such as the film rental,
photography, local taxi and retail bookstores sectors. Anecdotal evidence points towards 30-50%
margin compression. You can read our in-depth case studies here.
Netflix, Amazon, Uber: 30-50% margin compression from digital disruption
To understand how margins on standardised products might be squeezed, we look to other sectors that
have undergone sweeping technological change for inspiration. Across the four case studies we looked
at, we estimate roughly 30-50% margin compression. At the top of the range is the c.50% compression
in the movie rental sector where digital streaming services (think Netflix/Amazon Prime) crushed the
high margins enjoyed by brick & mortar incumbents (Blockbuster). Whilst the shift from physical
bookstores (Barnes & Noble) to the digitalised Amazon model where books were instead purchased
online and delivered via a sprawling distribution network saw a compression of c.30%.
We also looked at the advent of Uber for the local public transport sector and the launch of digital
cameras for Kodak. Our detailed work on case studies can be found here.
Picture 16: Margin transformation of photography sector
Source: Mediobanca Securities, Factset
Picture 17: Margin transformation of movie rental sector
Source: Mediobanca Securities, Factset
Picture 18: Margin transformation of taxi industry
Source: Mediobanca Securities, Factset, Gov.UK (link) – Data on taxis/PHVs
given on bi-annual basis pre-2017 so have interpolated interceding years.
Picture 19: Margin transformation of book retail sector
Source: Mediobanca Securities, Factset
0%
10%
20%
30%
40%
50%
60%
70%
80%
1980
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
Kodak Fuji Film
First commercially
marketed digital camera in
1990.
20%
25%
30%
35%
40%
45%
50%
55%
60%
65%
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
Blockbuster Netflix
40
60
80
100
120
140
160
20%
25%
30%
35%
40%
45%
50%
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
Uber Founded Taxis/PHVs [London, rhs] Addison Lee
15%
20%
25%
30%
35%
40%
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
Barnes & Noble Amazon
RegObs Special Report
07 December 2022 ◆ 27
…or standardised revenues -30%; costs -60% to offset
We simulate three possible scenarios for the tech-driven disruption of the standardised banking
sector. These are informed by the degree of margin compression we have witnessed during the
digital transformation of several case studies. Our base case uses the mid-point of 40% margin
compression to NII and fees, resulting in roughly -30% standardised revenue decline (not all fees –
insurance, AM, custody, placement and underwriting – suffer the same margin erosion with e.g.
payments/FX taking the full haircut) that would require a c.60% decrease in standardised costs to
be fully offset. This would imply that standardised banking would have to collapse the cost-income
ratio from c50% today to <30%, something even the most efficient of the famously lean Nordic
banks do not reach. Our optimistic and pessimistic scenarios instead suggest -22% and -36% of
revenue compression, necessitating c.-45% and c.-75% cost contraction, respectively, to maintain
profitability unchanged.
20-40% revenue attrition from digital disruption…
Picture 20 shows the revenue attrition of the standardised business of the twelve banks in our sample
under the base, optimistic and pessimistic scenarios from digital disruption. On aggregate, this
indicates c.20-40% revenue compression of standardised revenues.
Picture 20: Summary of standardised revenue compression under pessimistic, base & optimistic scenarios (% revenue)
Source: Mediobanca Securities
…calling for c.45-75% cost cuts to keep profitability of standardised banking unchanged
Picture 21 shows the equivalent cost cuts of the standardised business of the twelve banks in our
sample required to keep profitability unchanged under the base, optimistic and pessimistic scenarios
of revenue attrition from digital disruption. On aggregate, this indicates c.45-75% cost cuts, a
magnitude so great to require a rethink of the cost structure rather than linear cuts to the current
business model, in our view.
Picture 21: cost cuts to offset standardised revenue compression under pessimistic, base & optimistic scenarios, %
Source: Mediobanca Securities
-43%
-41%
-40%
-40%
-37%
-35%
-34%
-32%
-30%
-29%
-28%
-25%
-36%
-35%
-34%
-33%
-32%
-29%
-28%
-28%
-26%
-24%
-24%
-22%
-21%
-29%
-27%
-27%
-25%
-24%
-22%
-22%
-21%
-20%
-19%
-18%
-17%
-16%
-22%
-50%
-45%
-40%
-35%
-30%
-25%
-20%
-15%
-10%
-5%
0%
Bank
6
Bank
8
Bank
12
Bank
1
Bank
7
Bank
9
Bank
5
Bank
3
Bank
4
Bank
11
Bank
10
Bank
2
Aggregate
Pessimistic Base Optimistic
-85%
-73%
-73%
-69%
-69%
-62%
-52%
-48%
-45%
-41%
-39%
-25%
-73%
-60%
-53%
-52%
-51%
-49%
-46%
-39%
-36%
-34%
-31%
-29%
-19%
-59%
-41%
-37%
-36%
-35%
-34%
-32%
-28%
-26%
-24%
-22%
-20%
-14%
-45%
-90%
-80%
-70%
-60%
-50%
-40%
-30%
-20%
-10%
0%
Bank
6
Bank
12
Bank
7
Bank
8
Bank
1
Bank
9
Bank
5
Bank
11
Bank
4
Bank
10
Bank
3
Bank
2
Aggregate
Pessimistic Base Optimistic
RegObs Special Report
07 December 2022 ◆ 28
Scenario analysis: From the Amazon to Netflix experience
In this chapter we illustrate in more detail our methodology and the results of our scenario analysis
(base, optimistic and pessimistic) and its impacts on banks’ standardised business.
We trim NII by haircutting the credit spread on standardised loan products
For NII we isolated how much of the net interest margin is represented by the risk-free rate and how
much is coming from credit spread. To compute the overall risk-free rate for each of the companies
in our sample we first took the average duration for mortgages, consumer, corporate and SMEs for all
the relevant countries. Allowing us to get the relevant risk-free rate for the different types of loans
in each country. Again, using the EBA database, we were able to calculate the implied Group level
risk-free rates based on their how exposures were split between the different types of loans and
geographies. The credit spread was then simply assumed to be the delta between this risk-free rate
and the net interest margin. For our 12-bank sample the overall credit spread was found to be c.200bps
with consumer credit offering the widest margin (c.700bps) and mortgages the lowest (c.135bps).
Under each of the different scenarios we then hit the credit spread for the standardised products (i.e.
mortgages/consumer) based on what we had learnt from our historic case study examples:
• Optimistic scenario (-30%) = the Amazon experience
• Base scenario (-40%) = the Fuji experience
• Pessimistic scenario (-50%) = the Netflix experience
Not all fees are equal: standardisation score given to determine margin pressure
Our approach for fees was a little different given the breadth of different fee-earning activities
conducted by our banks. As such, we tried to adjust the magnitude of the margin cuts in each scenario
to a sliding scale. In this instance, we chose a scale from 0 to 5, where 0 represented a customised
product such as Wealth Management. We assigned this to each of our “customised” fee sources. At
the other end of the spectrum we had completely standardised products or services such as FX fees.
In terms of margin compression for each, we assumed that the most standardised products would see
compression akin to the top of the range from our case study analysis (i.e. -50%), whilst customised
products would be untouched. We then allowed for a 10% step between each of the sub-levels and
assumed that the optimistic scenario would be 5% less compression and the pessimistic 5% more
compression than this base case.
Picture 22: MB assumed sliding scale from customized to standardized sources of fee income
Source: Mediobanca Securities
RegObs Special Report
07 December 2022 ◆ 29
We summarise both the NII and fee compression in the below table…
Picture 23: Margin compression under our different scenarios
Source: Mediobanca Securities
#1: Base Case – The “Fuji Experience”
Standardised revenues decline roughly -29%, 2/3 explained by NII compression
We begin by looking at our base case. For this we use the mid-point of our 30-50% margin compression
range under the case study examples we gave earlier. A scenario which most closely follows what
happened to Fuji film in the photography industry during the shift away from film. Hence, we coin this
the Fuji scenario. Here, we find that on aggregate there is a -29% revenue when applying our margin
compression assumptions. This is split into roughly -19% coming from standardised NII and a further -
10% from standardised fee income. However, there is also a wide disparity around this mean across
the individual banks in our sample (standard deviation = 4%). At the extremes we see Bank 6 facing
nearly -35% revenue decline, with almost all of that coming from the NII side. The opposite is true for
Bank 2, where the revenue impact is more inconsequential at -21% and most of that is from lower fees.
Picture 24: Base Case – Impact on NII & fee income (% total revenues)
Source: Mediobanca Securities
Mortgages and consumer both equal drivers of the NII compression
Going into a little more detail on what is driving the changes in standardised revenues, we split out
the two standardised elements (mortgages and consumer). The picture here is very mixed at the
Current Optimistic Base Pessimistic
100% 70% 60% 50%
0 100% 100% 100% 100%
1 100% 95% 90% 85%
2 100% 85% 80% 75%
3 100% 75% 70% 65%
4 100% 65% 60% 55%
5 100% 55% 50% 45%
Margin Cut (% of initial revenue)
Fee rank
Net Interest Income
Standardised
Customised
-35% -34%
-33% -32%
-29% -28% -28% -26%
-24% -24%
-22%
-21%
-29%
-40%
-35%
-30%
-25%
-20%
-15%
-10%
-5%
0%
Bank
6
Bank
8
Bank
12
Bank
1
Bank
7
Bank
9
Bank
5
Bank
3
Bank
4
Bank
11
Bank
10
Bank
2
Aggregate
NII Fees
RegObs Special Report
07 December 2022 ◆ 30
individual bank level, which sort of balances out at the aggregate level to be roughly equal
contributions from mortgages and consumer loans. This is because while mortgages tend to make up
more of the overall loan book, credit spreads on consumer loans are much wider than for mortgages.
As such, the two effects more or less offset each other on aggregate.
Picture 25: Base Case – standardised NII impact split by product (% total revenues)
Source: Mediobanca Securities
Highly standardised ‘payments’ element dominates the story for fee compression
On the fee side, we also illustrate the contribution from the various fee buckets towards the total. We
do caution here that the way that different banks disclose their fee breakdown is not entirely
consonant with one another. Clearly, the main driver here is payments which makes up a hefty chunk
of the total (half of the -10%), although this is likely to also be picking up some of the FX fees as well
for example. This reflects the fact that payments are a major contributor to the fee income line and
at least in our view, are a highly standardised product.
Picture 26: Base Case – Fee impact split by product/service (% total revenues)
Source: Mediobanca Securities
-28% -28%
-22%
-20% -19%
-17%
-16% -16%
-14%
-13% -13%
-4%
-19%
-30%
-25%
-20%
-15%
-10%
-5%
0%
Bank
6
Bank
1
Bank
7
Bank
12
Bank
8
Bank
11
Bank
5
Bank
10
Bank
3
Bank
4
Bank
9
Bank
2
Aggregate
Mortgages Consumer
-16%
-15%
-15%
-13%
-12%
-11% -11%
-7% -7% -7% -6%
-4%
-10%
-18%
-16%
-14%
-12%
-10%
-8%
-6%
-4%
-2%
0%
Bank
2
Bank
9
Bank
8
Bank
12
Bank
3
Bank
5
Bank
4
Bank
7
Bank
10
Bank
11
Bank
6
Bank
1
Aggregate
Custody/Trust Underwriting/ Placement Insurance
FX Fees Asset Management Securities brokerage
Payments
RegObs Special Report
07 December 2022 ◆ 31
2/3 reduction in costs required to hold profitability steady…
New technologies should lead to some cost synergies as banks would be more efficient in offering the
same product. Yet, it is hard at this stage to estimate the amount of cost synergies new technologies
could free up. In our base case scenario, we estimate that core standardised revenues could decrease
by c30%, meaning that if costs remain the same, the operating income could decrease by c60% on
aggregate level. Hence, we estimate that cost should decrease by c60% to offset the margin pressure
and leaving unchanged the aggregated profitability of our aggregate standardised bank.
Picture 27: Base Case – potential impact to the standardized operating income
Source: Mediobanca Securities
…implying <30% C/I ratio post cuts in the standardised bank…
We show what the cost cuts required to neutralise revenue erosion would do to the cost-income ratio.
On aggregate, the C/I for the standardized business would need to drop from c.50% to <30%. This does
vary quite a bit across our sample of banks though. At the top of the range, Bank 6 would be required
to slash costs by c.74% to get to a C/I ratio of around 12%. With this being one fourth of the cost-
income of the most efficient banks in Europe today, getting here will be far from easy (and realistically
require a lot of work also on the revenue side). Bank 2 on the other hand would only require costs to
come down around 36%, i.e. some are genetically fitter for survival than others.
Picture 28: Base Case – cost cutting required to offset margin compression
Source: Mediobanca Securities
-19%
-10%
-29%
-57%
-59%
0
20,000
40,000
60,000
80,000
100,000
120,000
140,000
160,000
180,000
200,000
Revenues NII margin
pressure
Fee margin
pressure
Revenues post
margin
pressure
Costs New
Operating
income
Cost cutting Initial
operating
income
% change in
revenues
% change in
operating income
% fall in costs required
to offset revenue
margin compression
-60%
-53% -52% -51% -49% -46%
-39% -36% -34% -31% -29%
-19%
-42%
-80%
-60%
-40%
-20%
0%
20%
40%
60%
Bank
6
Bank
12
Bank
7
Bank
8
Bank
1
Bank
9
Bank
5
Bank
11
Bank
4
Bank
10
Bank
3
Bank
2
Aggregate
C/I (pre) C/I (post) Change in C/I (%)
C/I
ratio
(%)
C/I
ratio
(%)
Change
in
C/I
(%)
RegObs Special Report
07 December 2022 ◆ 32
…implying <40% C/I ratio of standardised+customised banking, at Nordic level
Finally, considering that banks are a mix of standardised and customised businesses, usually sharing
costs among businesses, we illustrate the impact of our analysis to the aggregate banks’ sample taking
into consideration both standardised and customised businesses. Total combined revenues would
decline by 18% (assuming no revenue erosion in the customised bank), while costs should drop by 36%
to perpetuate the initial profitability. This would imply aggregate C/I ratio dropping to <40% from
c.50%, for reference, in line with best in class Nordic banks.
Picture 29: Base Case – potential impact to the aggregate operating income (standardized +
custumised business)
Source: Mediobanca Securities
#2: Bookshop scenario - Optimistic case
Standardised revenues decline by 22%
If the banking sector faced margin compression more akin in magnitude to that of book retailers post-
Amazon (i.e. 30% margin compression), revenue compression would be more benign at -22% (vs -29%
in our base case). In this more optimistic scenario, the range of revenue compression corridor for
individual banks would narrow to -27% to -16%.
Picture 30: Optimistic Case – Impact on standardized revenues
Source: Mediobanca Securities
​
-37%
-18%
-36%
0
50,000
100,000
150,000
200,000
250,000
Revenues Margin
pressure
Revenues
post margin
pressure
Costs New
Operating
income
Cost cutting Initial
operating
income
% change
in
revenues
% change in
operating
% fall in costs
required to offset
revenue margin
compression
-27% -27%
-25%
-24%
-22% -22% -21%
-20%
-19% -18% -17%
-16%
-22%
-30%
-25%
-20%
-15%
-10%
-5%
0%
Bank
8
Bank
6
Bank
12
Bank
1
Bank
9
Bank
7
Bank
5
Bank
3
Bank
4
Bank
11
Bank
10
Bank
2
Aggregate
NII Fees
RegObs Special Report
07 December 2022 ◆ 33
c.1/2 reduction in costs required to hold profitability steady
In our optimistic case scenario, we estimate that core standardised revenues could decrease by c22%,
meaning that if costs remain the same, the operating income of the standardised bank could decrease
by c44%. Hence, we estimate that cost should decrease by c45% to offset the margin pressure and
leaving the aggregated profitability of our banks’ sample unchanged. In this case the C/I ratio for the
standardised bank would need to fall to c.35% from c.50%.
Picture 31: Optimistic Case – potential impact to the standardized operating income
Source: Mediobanca Securities
#3: Movie rental scenario - Pessimistic case
Standardised revenues decline by 36%
The final scenario we show is for a more “pessimistic case”. This more closely traces the very high
margin compressions we have seen in industries like movie & TV rental/streaming of around -50%.
Here we might observe standardised revenues coming down around -36% on aggregate. In this instance,
we could even see some banks have around 40% of their total standardised revenues eroded, with a
minimum of 25% for the least impacted names.
Picture 32: Pessimistic Case – Impact on standardized revenues
Source: Mediobanca Securities
-14%
-8%
-22%
-44%
-45%
0
20,000
40,000
60,000
80,000
100,000
120,000
140,000
160,000
180,000
200,000
Revenues NII margin
pressure
Fee margin
pressure
Revenues post
margin
pressure
Costs New
Operating
income
Cost cutting Initial
operating
income
% change in
revenues
% change in
operating income
% fall in costs required
to offset revenue
margin compression
-43%
-41% -40% -40%
-37%
-35% -34% -32%
-30% -29%
-28%
-25%
-36%
-50%
-45%
-40%
-35%
-30%
-25%
-20%
-15%
-10%
-5%
0%
Bank
6
Bank
8
Bank
12
Bank
1
Bank
7
Bank
9
Bank
5
Bank
3
Bank
4
Bank
11
Bank
10
Bank
2
Aggregate
NII Fees
RegObs Special Report
07 December 2022 ◆ 34
Cutting cost base by c73% to retain the same profitability…
In our pessimistic case scenario, we estimate that core revenues could decrease by c36% (vs 29% and
22% in our base and optimistic scenarios, respectively), leading to 70% cut of the operating income,
without taking into consideration any cost optimization. Hence, we estimate that cost should decrease
by c73% to offset the margin pressure, to perpetuate the profitability pre-disruption.
Picture 33: Pessimistic Case – potential impact to the aggregated operating income
Source: Mediobanca Securities
…implying C/I c.20% in the standardised bank; c.35% at standardised+customised level
With 73% cost cuts required to offset the margin compression in the pessimistic scenario, our sample
of standardised banks would be required to bring their C/I ratio down significantly to c.20%. Leaving
the customised perimeter unchanged, this would imply c.35% C/I at standardised+customised level,
well below the current best-in-class Nordic banks.
The transformation of standardised banking
The current aggregated core revenues (NII+fees) for the 12 banks in our sample works out to be around
€240bn (split c.€150bn standardised and c.€90bn customised). Whereas the relevant expense base to
support these revenues is c.€120bn (split c.€75bn standardised and c.€45bn customised). This gives us
a pre-provision profit of roughly €120bn, again split 60-65% standardised and 35-40% customised.
The table shows the deep transformation of standardised banking versus the stability in customised
banking.
-24%
-12%
-36%
-70%
-73%
0
20,000
40,000
60,000
80,000
100,000
120,000
140,000
160,000
180,000
200,000
Revenues NII margin
pressure
Fee margin
pressure
Revenues post
margin
pressure
Costs New
Operating
income
Cost cutting Initial
operating
income
% change in
revenues
% change in
operating income
% fall in costs required
to offset revenue
margin compression
RegObs Special Report
07 December 2022 ◆ 35
Table 3: Summary of aggregate P&L under adverse, base and optimistic scenarios
Source: Mediobanca Securities, EBA
Current Pessimistic Base Optimistic
NII [standardised] 94 57 65 72
NII [customised] 48 48 48 48
NII [total] 142 106 113 120
Fees [standardised] 57 40 43 46
Fees [customised] 40 40 40 40
Fees [total] 97 80 83 85
Rev enues [standardised] 151 97 107 118
Rev enues [customised] 88 88 88 88
Revenues [total] 240 185 196 206
Revenue deterioration [% current] -23% -18% -14%
Expenses [standardised] (74) (20) (30) (41)
Expenses [customised] (47) (47) (47) (47)
Expenses [total] (121) (67) (77) (87)
Cost cutting [% current] 0% -45% -36% -28%
C/I [standardised] 49% 21% 28% 35%
C/I [customised] 53% 53% 53% 53%
C/I [total] 51% 36% 39% 42%
RegObs Special Report
07 December 2022 ◆ 36
The great (cost) reset: scaling up DLT to do the trick
Despite being in vogue, DLT technology is still disproportionately used merely in small,
experimental projects. This makes it hard to extrapolate what it could/would do, if applied to a
scale business, either assuming the trilemma will be solved at some point or that permissioned-
DLT can do so already. Our intense dialogue with industry experts helps us to move from theory
to practice in formulating the cost function of a (permissioned) DLT network, featuring fixed costs
which progressively handover to variable transaction costs as volumes pick up. The Spunta DLT
case validates our function which lands c.2-3x above its costs (a spread of €2-4m), i.e. embedding
a margin of conservatism. Our conclusion is that DLT technology could provide banks with
astronomic cost efficiencies in some of their businesses.
Lack of evidence on blockchain’s running costs
An often quoted paper (link) estimates DLT technology “could reduce banks’ infrastructure costs
attributed to cross-border payments, securities trading and regulatory compliance by between $15-
20bn”. While there are oceans of narrative about the immense cost-saving opportunity of digitisation,
there is very little practical evidence about the quantum involved. For example, there is no easy
evidence of what the running costs of a DLT-based platform look like.
Sunk development costs, linear opex until scale economies kick in to flatten the curve
Logically, the cost function of DLT should involve sunk development costs. These will vary with the
complexity of the applications that the DLT network is intended to operate and to the number of nodes
required. Counterintuitively, all things are considered, a permissionless network is likely to cost more
than a permissioned one. In fact, while the former may require lower sunk investment to setup the
network, it will also require higher operating costs to encompass transparency with privacy, entailing
the archive and accessibility of ever larger amounts of data for an overall significant overhead.
Instead, permissioned DLT networks will be leaner on operating costs once the higher sunk setup
investment on machinery is satisfied, i.e. resulting in lower individual transaction costs when
implementing the tool on large volumes.
We have spoken to industry experts to try to gauge the shape of the cost function of a permissioned
DLT network, which we summarise here:
1. Hardware per node – blockchain ledger and part dedicated to support the application - €30-
35k per annum
2. Service per node – staff, hardware maintenance, helpdesk – 0.1x FTE in case of fragmentation
in network managers, 0.025-0.05x when optimising
3. Private network connection per node – running the DLT network on a private network incurs
into costs depending on the type of network, the throughput etc - €12-15k per annum
4. Hardware - One single hardware component such as transaction validation / notary functions
- €150k per annum regardless of the number of nodes in the DLT network
5. Finally, the cost of DLT software licencing. This can vary dramatically from the menu price
(€0.3/transaction even reaching levels 100x smaller)
In essence, what we have gathered is a cost function looking like:
𝐴𝑛𝑛𝑢𝑎𝑙 𝐶𝑜𝑠𝑡 = €150,000 + (€35,000 + 0.05 ∗ 𝐹𝑇𝐸 𝑐𝑜𝑠𝑡 + €15,000) ∗ 𝑛 + 𝑡 (𝑘) ∗ 𝑘
Where n = number of DLT nodes, t = the transaction cost and k = the number of transactions, with t
progressively decreasing as k expands.
RegObs Special Report
07 December 2022 ◆ 37
Spunta, the largest DLT network in operation
We have dug far and wide to find a living example of DLT network to gauge the costs related to
operating it. Spunta, is a DLT-based network comprised of 100 Italian banks, dedicated to interbank
account reconciliation governed by ABI Lab, part of the Italian Banking Association.
Spunta’s application “verifies the matching of correspondent accounts that involve two different
banks. The interbank reconciliation procedure in Italy is linked to processes traditionally carried out
by the back office and aimed at reconciling the transaction flows that generate accounting entries in
the mutual accounts in Italy and at managing pending transactions. Up to now [then for today’s
reader], reconciliation was based on bilateral registers with a low level of standardisation and
operating processes that were not very advanced. The implementation of a blockchain-based process
using Distributed Ledger Technology (DLT) for interbank reconciliations in Italy makes it possible to
automatically detect non-matching transactions using a shared algorithm that standardises both the
process and the single communication channel, and provides a comprehensive view of the transactions
among the interested parties. As a consequence, the principles of the new Spunta envisage full
visibility of the transactions and those of the counterparty; rapid management of the flows with
daily, rather than monthly, reconciliations; shared rules for the symmetrical reconciliation of
transactions between counterparty banks; and the integrated management of communications and
processes in the event of an imbalance”. See https://www.abilab.it/en/aree-ricerca/blockchain-
dlt/spunta-banca-dlt.
Spunta only operates one hour a day, carrying out c.350m transactions a year. “If this application will
be applied to more complex cases, working at its full capacity, it was estimated that the
infrastructure would be able to manage a total volume of 8.4 billion transactions. As term of
comparison, Bitcoin’s blockchain in the entire 2020 managed 113 million transactions”.
Spunta’s costs: €2-3m a year
So how much does the largest DLT network cost? Going through the annual reports of ABI Lab, the
entity operating Spunta for ABI, we were able to detect annual revenues budgeted from Blockchain &
DLT of €2-3m per annum, growing slightly in 2021 (so for 2022). As we would expect ABI Lab to operate
Spunta at a breakeven regime for the ABI affiliate banks, we consider this a good proxy for the
comprehensive costs of the Spunta DLT application.
Picture 34: blockchain cost evolution, €
Source: Mediobanca Securities, ABI LAB annual report
Fixed costs handover to variable costs beyond 10bn transactions
Picture 35 plots the cost evolution from our cost function calibrated for DLT networks of 100, 320 and
640 nodes, ramping up the volume of transactions from 10m to 10bn. This illustrative example shows
the high incidence on total cost of the network size and of other fixed cost when the network is
relatively new and transaction volumes are low. Instead, when transactions start to grow in the
-
500,000
1,000,000
1,500,000
2,000,000
2,500,000
3,000,000
2019 2020 2021
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf
RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf

Weitere ähnliche Inhalte

Ähnlich wie RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf

Cryptocurrencies, Blockchain and Initial Coin Offerings
Cryptocurrencies, Blockchain and Initial Coin OfferingsCryptocurrencies, Blockchain and Initial Coin Offerings
Cryptocurrencies, Blockchain and Initial Coin Offeringssarah marville
 
Blockchain, Investment Banking's Soteria
Blockchain, Investment Banking's SoteriaBlockchain, Investment Banking's Soteria
Blockchain, Investment Banking's SoteriaThéo Tortorici
 
Blockchain: Investment Banking's Soteria
Blockchain: Investment Banking's SoteriaBlockchain: Investment Banking's Soteria
Blockchain: Investment Banking's SoteriaJulius Kühn
 
It's the to take a fresh look at the crypto market
It's the to take a fresh look at the crypto marketIt's the to take a fresh look at the crypto market
It's the to take a fresh look at the crypto marketSeamus Donoghue
 
The Finance, The Digital & The Society - Smart Cities Summit 2018 - Algiers
The Finance, The Digital & The Society - Smart Cities Summit 2018 - AlgiersThe Finance, The Digital & The Society - Smart Cities Summit 2018 - Algiers
The Finance, The Digital & The Society - Smart Cities Summit 2018 - AlgiersSmart Algiers
 
COVID-19: Sustaining Liquidity/Funding Management and Treasury Operations in ...
COVID-19: Sustaining Liquidity/Funding Management and Treasury Operations in ...COVID-19: Sustaining Liquidity/Funding Management and Treasury Operations in ...
COVID-19: Sustaining Liquidity/Funding Management and Treasury Operations in ...Boston Consulting Group
 
Libra and the Others The Future of Digital Money
Libra and the Others The Future of Digital MoneyLibra and the Others The Future of Digital Money
Libra and the Others The Future of Digital Moneycanadaustaxplanning
 
Global Payments 2020: Transformation
Global Payments  2020: TransformationGlobal Payments  2020: Transformation
Global Payments 2020: TransformationAndrew Gumenniy
 
The Future of Cryptocurrency
The Future of Cryptocurrency The Future of Cryptocurrency
The Future of Cryptocurrency SHarriman1
 
WP-Digital-Banking-EN-HD
WP-Digital-Banking-EN-HDWP-Digital-Banking-EN-HD
WP-Digital-Banking-EN-HDStephen PERIN
 
Disruptive Innovations 2014 / CITI
Disruptive Innovations 2014 / CITIDisruptive Innovations 2014 / CITI
Disruptive Innovations 2014 / CITIRana Babaç
 
Fabian Cesarini - Language Lesson - Future Banking Magazine i1 2010
Fabian Cesarini - Language Lesson - Future Banking Magazine i1 2010Fabian Cesarini - Language Lesson - Future Banking Magazine i1 2010
Fabian Cesarini - Language Lesson - Future Banking Magazine i1 2010Fabian M. Cesarini, CITP
 
Blockchain beyond fintech by ridgelift.io
Blockchain beyond fintech by ridgelift.ioBlockchain beyond fintech by ridgelift.io
Blockchain beyond fintech by ridgelift.ioUdayan Modhe
 
FIXatdl and the 2010 Flash Crash presented at Princeton Qwafafew
FIXatdl and the 2010 Flash Crash presented at Princeton QwafafewFIXatdl and the 2010 Flash Crash presented at Princeton Qwafafew
FIXatdl and the 2010 Flash Crash presented at Princeton QwafafewRobert Golan
 
A Financial Tech Tsunami Driven by Blockchain AI Crypto Economics
A Financial Tech Tsunami Driven by Blockchain AI Crypto EconomicsA Financial Tech Tsunami Driven by Blockchain AI Crypto Economics
A Financial Tech Tsunami Driven by Blockchain AI Crypto EconomicsDinis Guarda
 
Mobile Money Transfer & Remittances:
Mobile Money Transfer & Remittances:Mobile Money Transfer & Remittances:
Mobile Money Transfer & Remittances:ReportLinker.com
 
The european union payments landscape in perspective
The european union payments landscape in perspectiveThe european union payments landscape in perspective
The european union payments landscape in perspectivePaperjam_redaction
 
Bcg media-what-happens-when-if-turns-to-when-in-quantum-computing-jul-2021
Bcg media-what-happens-when-if-turns-to-when-in-quantum-computing-jul-2021Bcg media-what-happens-when-if-turns-to-when-in-quantum-computing-jul-2021
Bcg media-what-happens-when-if-turns-to-when-in-quantum-computing-jul-2021Javier López Valbuena
 

Ähnlich wie RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf (20)

Cryptocurrencies, Blockchain and Initial Coin Offerings
Cryptocurrencies, Blockchain and Initial Coin OfferingsCryptocurrencies, Blockchain and Initial Coin Offerings
Cryptocurrencies, Blockchain and Initial Coin Offerings
 
Blockchain, Investment Banking's Soteria
Blockchain, Investment Banking's SoteriaBlockchain, Investment Banking's Soteria
Blockchain, Investment Banking's Soteria
 
Blockchain: Investment Banking's Soteria
Blockchain: Investment Banking's SoteriaBlockchain: Investment Banking's Soteria
Blockchain: Investment Banking's Soteria
 
It's the to take a fresh look at the crypto market
It's the to take a fresh look at the crypto marketIt's the to take a fresh look at the crypto market
It's the to take a fresh look at the crypto market
 
The Finance, The Digital & The Society - Smart Cities Summit 2018 - Algiers
The Finance, The Digital & The Society - Smart Cities Summit 2018 - AlgiersThe Finance, The Digital & The Society - Smart Cities Summit 2018 - Algiers
The Finance, The Digital & The Society - Smart Cities Summit 2018 - Algiers
 
COVID-19: Sustaining Liquidity/Funding Management and Treasury Operations in ...
COVID-19: Sustaining Liquidity/Funding Management and Treasury Operations in ...COVID-19: Sustaining Liquidity/Funding Management and Treasury Operations in ...
COVID-19: Sustaining Liquidity/Funding Management and Treasury Operations in ...
 
Libra and the Others The Future of Digital Money
Libra and the Others The Future of Digital MoneyLibra and the Others The Future of Digital Money
Libra and the Others The Future of Digital Money
 
Retail Banking: In tech we trust
Retail Banking: In tech we trustRetail Banking: In tech we trust
Retail Banking: In tech we trust
 
Global Payments 2020: Transformation
Global Payments  2020: TransformationGlobal Payments  2020: Transformation
Global Payments 2020: Transformation
 
The Future of Cryptocurrency
The Future of Cryptocurrency The Future of Cryptocurrency
The Future of Cryptocurrency
 
WP-Digital-Banking-EN-HD
WP-Digital-Banking-EN-HDWP-Digital-Banking-EN-HD
WP-Digital-Banking-EN-HD
 
Disruptive Innovations 2014 / CITI
Disruptive Innovations 2014 / CITIDisruptive Innovations 2014 / CITI
Disruptive Innovations 2014 / CITI
 
IBM Report Final
IBM Report FinalIBM Report Final
IBM Report Final
 
Fabian Cesarini - Language Lesson - Future Banking Magazine i1 2010
Fabian Cesarini - Language Lesson - Future Banking Magazine i1 2010Fabian Cesarini - Language Lesson - Future Banking Magazine i1 2010
Fabian Cesarini - Language Lesson - Future Banking Magazine i1 2010
 
Blockchain beyond fintech by ridgelift.io
Blockchain beyond fintech by ridgelift.ioBlockchain beyond fintech by ridgelift.io
Blockchain beyond fintech by ridgelift.io
 
FIXatdl and the 2010 Flash Crash presented at Princeton Qwafafew
FIXatdl and the 2010 Flash Crash presented at Princeton QwafafewFIXatdl and the 2010 Flash Crash presented at Princeton Qwafafew
FIXatdl and the 2010 Flash Crash presented at Princeton Qwafafew
 
A Financial Tech Tsunami Driven by Blockchain AI Crypto Economics
A Financial Tech Tsunami Driven by Blockchain AI Crypto EconomicsA Financial Tech Tsunami Driven by Blockchain AI Crypto Economics
A Financial Tech Tsunami Driven by Blockchain AI Crypto Economics
 
Mobile Money Transfer & Remittances:
Mobile Money Transfer & Remittances:Mobile Money Transfer & Remittances:
Mobile Money Transfer & Remittances:
 
The european union payments landscape in perspective
The european union payments landscape in perspectiveThe european union payments landscape in perspective
The european union payments landscape in perspective
 
Bcg media-what-happens-when-if-turns-to-when-in-quantum-computing-jul-2021
Bcg media-what-happens-when-if-turns-to-when-in-quantum-computing-jul-2021Bcg media-what-happens-when-if-turns-to-when-in-quantum-computing-jul-2021
Bcg media-what-happens-when-if-turns-to-when-in-quantum-computing-jul-2021
 

Mehr von Chris Skinner

20200128 Ethical by Design - Whitepaper.pdf
20200128 Ethical by Design - Whitepaper.pdf20200128 Ethical by Design - Whitepaper.pdf
20200128 Ethical by Design - Whitepaper.pdfChris Skinner
 
20220816-EthicsGrade_Scorecard-JP_Morgan_Chase-Q2-63_57.pdf
20220816-EthicsGrade_Scorecard-JP_Morgan_Chase-Q2-63_57.pdf20220816-EthicsGrade_Scorecard-JP_Morgan_Chase-Q2-63_57.pdf
20220816-EthicsGrade_Scorecard-JP_Morgan_Chase-Q2-63_57.pdfChris Skinner
 
WEF_Navigating_the_Industrial_Metaverse_A_Blueprint_2024.pdf
WEF_Navigating_the_Industrial_Metaverse_A_Blueprint_2024.pdfWEF_Navigating_the_Industrial_Metaverse_A_Blueprint_2024.pdf
WEF_Navigating_the_Industrial_Metaverse_A_Blueprint_2024.pdfChris Skinner
 
Temenos_Global_Report_Final_Sep22.pdf
Temenos_Global_Report_Final_Sep22.pdfTemenos_Global_Report_Final_Sep22.pdf
Temenos_Global_Report_Final_Sep22.pdfChris Skinner
 
the-second-wave-resilient-inclusive-exponential-fintechs.pdf
the-second-wave-resilient-inclusive-exponential-fintechs.pdfthe-second-wave-resilient-inclusive-exponential-fintechs.pdf
the-second-wave-resilient-inclusive-exponential-fintechs.pdfChris Skinner
 
Evident_AI_Innovation_Report_.pdf
Evident_AI_Innovation_Report_.pdfEvident_AI_Innovation_Report_.pdf
Evident_AI_Innovation_Report_.pdfChris Skinner
 
global-pulse-of-fintech-h123-report-web (1).pdf
global-pulse-of-fintech-h123-report-web (1).pdfglobal-pulse-of-fintech-h123-report-web (1).pdf
global-pulse-of-fintech-h123-report-web (1).pdfChris Skinner
 
roland_berger_trend_compendium_2050_compact_version.pdf
roland_berger_trend_compendium_2050_compact_version.pdfroland_berger_trend_compendium_2050_compact_version.pdf
roland_berger_trend_compendium_2050_compact_version.pdfChris Skinner
 
US core banking share
US core banking shareUS core banking share
US core banking shareChris Skinner
 
DIFC Summit, May 2023.pptx
DIFC Summit, May 2023.pptxDIFC Summit, May 2023.pptx
DIFC Summit, May 2023.pptxChris Skinner
 
JPMorgan_Chase_Presentation.pdf
JPMorgan_Chase_Presentation.pdfJPMorgan_Chase_Presentation.pdf
JPMorgan_Chase_Presentation.pdfChris Skinner
 
Finacle-Innovation-in-Retail-Banking-2023-Report.pdf
Finacle-Innovation-in-Retail-Banking-2023-Report.pdfFinacle-Innovation-in-Retail-Banking-2023-Report.pdf
Finacle-Innovation-in-Retail-Banking-2023-Report.pdfChris Skinner
 
European Union survey of digital banking progress
European Union survey of digital banking progressEuropean Union survey of digital banking progress
European Union survey of digital banking progressChris Skinner
 
Climate Impact Report
Climate Impact ReportClimate Impact Report
Climate Impact ReportChris Skinner
 
FT Partners Research - FinTech in Africa.pdf
FT Partners Research - FinTech in Africa.pdfFT Partners Research - FinTech in Africa.pdf
FT Partners Research - FinTech in Africa.pdfChris Skinner
 
Accenture-Banking-Top-10-Trends-2023.pdf
Accenture-Banking-Top-10-Trends-2023.pdfAccenture-Banking-Top-10-Trends-2023.pdf
Accenture-Banking-Top-10-Trends-2023.pdfChris Skinner
 
Global Cryptasset Standards
Global Cryptasset StandardsGlobal Cryptasset Standards
Global Cryptasset StandardsChris Skinner
 
European Banks Outlook 2023_09.12.22.pdf
European Banks Outlook 2023_09.12.22.pdfEuropean Banks Outlook 2023_09.12.22.pdf
European Banks Outlook 2023_09.12.22.pdfChris Skinner
 

Mehr von Chris Skinner (20)

20200128 Ethical by Design - Whitepaper.pdf
20200128 Ethical by Design - Whitepaper.pdf20200128 Ethical by Design - Whitepaper.pdf
20200128 Ethical by Design - Whitepaper.pdf
 
20220816-EthicsGrade_Scorecard-JP_Morgan_Chase-Q2-63_57.pdf
20220816-EthicsGrade_Scorecard-JP_Morgan_Chase-Q2-63_57.pdf20220816-EthicsGrade_Scorecard-JP_Morgan_Chase-Q2-63_57.pdf
20220816-EthicsGrade_Scorecard-JP_Morgan_Chase-Q2-63_57.pdf
 
WEF_Navigating_the_Industrial_Metaverse_A_Blueprint_2024.pdf
WEF_Navigating_the_Industrial_Metaverse_A_Blueprint_2024.pdfWEF_Navigating_the_Industrial_Metaverse_A_Blueprint_2024.pdf
WEF_Navigating_the_Industrial_Metaverse_A_Blueprint_2024.pdf
 
Temenos_Global_Report_Final_Sep22.pdf
Temenos_Global_Report_Final_Sep22.pdfTemenos_Global_Report_Final_Sep22.pdf
Temenos_Global_Report_Final_Sep22.pdf
 
the-second-wave-resilient-inclusive-exponential-fintechs.pdf
the-second-wave-resilient-inclusive-exponential-fintechs.pdfthe-second-wave-resilient-inclusive-exponential-fintechs.pdf
the-second-wave-resilient-inclusive-exponential-fintechs.pdf
 
Evident_AI_Innovation_Report_.pdf
Evident_AI_Innovation_Report_.pdfEvident_AI_Innovation_Report_.pdf
Evident_AI_Innovation_Report_.pdf
 
global-pulse-of-fintech-h123-report-web (1).pdf
global-pulse-of-fintech-h123-report-web (1).pdfglobal-pulse-of-fintech-h123-report-web (1).pdf
global-pulse-of-fintech-h123-report-web (1).pdf
 
roland_berger_trend_compendium_2050_compact_version.pdf
roland_berger_trend_compendium_2050_compact_version.pdfroland_berger_trend_compendium_2050_compact_version.pdf
roland_berger_trend_compendium_2050_compact_version.pdf
 
S&P on DeFi
S&P on DeFiS&P on DeFi
S&P on DeFi
 
US core banking share
US core banking shareUS core banking share
US core banking share
 
DIFC Summit, May 2023.pptx
DIFC Summit, May 2023.pptxDIFC Summit, May 2023.pptx
DIFC Summit, May 2023.pptx
 
JPMorgan_Chase_Presentation.pdf
JPMorgan_Chase_Presentation.pdfJPMorgan_Chase_Presentation.pdf
JPMorgan_Chase_Presentation.pdf
 
Finacle-Innovation-in-Retail-Banking-2023-Report.pdf
Finacle-Innovation-in-Retail-Banking-2023-Report.pdfFinacle-Innovation-in-Retail-Banking-2023-Report.pdf
Finacle-Innovation-in-Retail-Banking-2023-Report.pdf
 
European Union survey of digital banking progress
European Union survey of digital banking progressEuropean Union survey of digital banking progress
European Union survey of digital banking progress
 
Climate Impact Report
Climate Impact ReportClimate Impact Report
Climate Impact Report
 
FT Partners Research - FinTech in Africa.pdf
FT Partners Research - FinTech in Africa.pdfFT Partners Research - FinTech in Africa.pdf
FT Partners Research - FinTech in Africa.pdf
 
Accenture-Banking-Top-10-Trends-2023.pdf
Accenture-Banking-Top-10-Trends-2023.pdfAccenture-Banking-Top-10-Trends-2023.pdf
Accenture-Banking-Top-10-Trends-2023.pdf
 
Global Cryptasset Standards
Global Cryptasset StandardsGlobal Cryptasset Standards
Global Cryptasset Standards
 
European Banks Outlook 2023_09.12.22.pdf
European Banks Outlook 2023_09.12.22.pdfEuropean Banks Outlook 2023_09.12.22.pdf
European Banks Outlook 2023_09.12.22.pdf
 
2023 Outlook
2023 Outlook2023 Outlook
2023 Outlook
 

Kürzlich hochgeladen

Kharghar Blowjob Housewife Call Girls NUmber-9833754194-CBD Belapur Internati...
Kharghar Blowjob Housewife Call Girls NUmber-9833754194-CBD Belapur Internati...Kharghar Blowjob Housewife Call Girls NUmber-9833754194-CBD Belapur Internati...
Kharghar Blowjob Housewife Call Girls NUmber-9833754194-CBD Belapur Internati...priyasharma62062
 
VIP Call Girl in Mumbai 💧 9920725232 ( Call Me ) Get A New Crush Everyday Wit...
VIP Call Girl in Mumbai 💧 9920725232 ( Call Me ) Get A New Crush Everyday Wit...VIP Call Girl in Mumbai 💧 9920725232 ( Call Me ) Get A New Crush Everyday Wit...
VIP Call Girl in Mumbai 💧 9920725232 ( Call Me ) Get A New Crush Everyday Wit...dipikadinghjn ( Why You Choose Us? ) Escorts
 
Call Girls Banaswadi Just Call 👗 7737669865 👗 Top Class Call Girl Service Ban...
Call Girls Banaswadi Just Call 👗 7737669865 👗 Top Class Call Girl Service Ban...Call Girls Banaswadi Just Call 👗 7737669865 👗 Top Class Call Girl Service Ban...
Call Girls Banaswadi Just Call 👗 7737669865 👗 Top Class Call Girl Service Ban...amitlee9823
 
Diva-Thane European Call Girls Number-9833754194-Diva Busty Professional Call...
Diva-Thane European Call Girls Number-9833754194-Diva Busty Professional Call...Diva-Thane European Call Girls Number-9833754194-Diva Busty Professional Call...
Diva-Thane European Call Girls Number-9833754194-Diva Busty Professional Call...priyasharma62062
 
Booking open Available Pune Call Girls Shivane 6297143586 Call Hot Indian Gi...
Booking open Available Pune Call Girls Shivane  6297143586 Call Hot Indian Gi...Booking open Available Pune Call Girls Shivane  6297143586 Call Hot Indian Gi...
Booking open Available Pune Call Girls Shivane 6297143586 Call Hot Indian Gi...Call Girls in Nagpur High Profile
 
VIP Kalyan Call Girls 🌐 9920725232 🌐 Make Your Dreams Come True With Mumbai E...
VIP Kalyan Call Girls 🌐 9920725232 🌐 Make Your Dreams Come True With Mumbai E...VIP Kalyan Call Girls 🌐 9920725232 🌐 Make Your Dreams Come True With Mumbai E...
VIP Kalyan Call Girls 🌐 9920725232 🌐 Make Your Dreams Come True With Mumbai E...roshnidevijkn ( Why You Choose Us? ) Escorts
 
Vip Call US 📞 7738631006 ✅Call Girls In Sakinaka ( Mumbai )
Vip Call US 📞 7738631006 ✅Call Girls In Sakinaka ( Mumbai )Vip Call US 📞 7738631006 ✅Call Girls In Sakinaka ( Mumbai )
Vip Call US 📞 7738631006 ✅Call Girls In Sakinaka ( Mumbai )Pooja Nehwal
 
Best VIP Call Girls Morni Hills Just Click Me 6367492432
Best VIP Call Girls Morni Hills Just Click Me 6367492432Best VIP Call Girls Morni Hills Just Click Me 6367492432
Best VIP Call Girls Morni Hills Just Click Me 6367492432motiram463
 
Call Girls in New Friends Colony Delhi 💯 Call Us 🔝9205541914 🔝( Delhi) Escort...
Call Girls in New Friends Colony Delhi 💯 Call Us 🔝9205541914 🔝( Delhi) Escort...Call Girls in New Friends Colony Delhi 💯 Call Us 🔝9205541914 🔝( Delhi) Escort...
Call Girls in New Friends Colony Delhi 💯 Call Us 🔝9205541914 🔝( Delhi) Escort...Delhi Call girls
 
falcon-invoice-discounting-unlocking-prime-investment-opportunities
falcon-invoice-discounting-unlocking-prime-investment-opportunitiesfalcon-invoice-discounting-unlocking-prime-investment-opportunities
falcon-invoice-discounting-unlocking-prime-investment-opportunitiesFalcon Invoice Discounting
 
Call Girls Koregaon Park Call Me 7737669865 Budget Friendly No Advance Booking
Call Girls Koregaon Park Call Me 7737669865 Budget Friendly No Advance BookingCall Girls Koregaon Park Call Me 7737669865 Budget Friendly No Advance Booking
Call Girls Koregaon Park Call Me 7737669865 Budget Friendly No Advance Bookingroncy bisnoi
 
Top Rated Pune Call Girls Dighi ⟟ 6297143586 ⟟ Call Me For Genuine Sex Servi...
Top Rated  Pune Call Girls Dighi ⟟ 6297143586 ⟟ Call Me For Genuine Sex Servi...Top Rated  Pune Call Girls Dighi ⟟ 6297143586 ⟟ Call Me For Genuine Sex Servi...
Top Rated Pune Call Girls Dighi ⟟ 6297143586 ⟟ Call Me For Genuine Sex Servi...Call Girls in Nagpur High Profile
 
Business Principles, Tools, and Techniques in Participating in Various Types...
Business Principles, Tools, and Techniques  in Participating in Various Types...Business Principles, Tools, and Techniques  in Participating in Various Types...
Business Principles, Tools, and Techniques in Participating in Various Types...jeffreytingson
 
VIP Call Girl Service Andheri West ⚡ 9920725232 What It Takes To Be The Best ...
VIP Call Girl Service Andheri West ⚡ 9920725232 What It Takes To Be The Best ...VIP Call Girl Service Andheri West ⚡ 9920725232 What It Takes To Be The Best ...
VIP Call Girl Service Andheri West ⚡ 9920725232 What It Takes To Be The Best ...dipikadinghjn ( Why You Choose Us? ) Escorts
 
Mira Road Awesome 100% Independent Call Girls NUmber-9833754194-Dahisar Inter...
Mira Road Awesome 100% Independent Call Girls NUmber-9833754194-Dahisar Inter...Mira Road Awesome 100% Independent Call Girls NUmber-9833754194-Dahisar Inter...
Mira Road Awesome 100% Independent Call Girls NUmber-9833754194-Dahisar Inter...priyasharma62062
 
VIP Independent Call Girls in Mumbai 🌹 9920725232 ( Call Me ) Mumbai Escorts ...
VIP Independent Call Girls in Mumbai 🌹 9920725232 ( Call Me ) Mumbai Escorts ...VIP Independent Call Girls in Mumbai 🌹 9920725232 ( Call Me ) Mumbai Escorts ...
VIP Independent Call Girls in Mumbai 🌹 9920725232 ( Call Me ) Mumbai Escorts ...dipikadinghjn ( Why You Choose Us? ) Escorts
 
Top Rated Pune Call Girls Viman Nagar ⟟ 6297143586 ⟟ Call Me For Genuine Sex...
Top Rated  Pune Call Girls Viman Nagar ⟟ 6297143586 ⟟ Call Me For Genuine Sex...Top Rated  Pune Call Girls Viman Nagar ⟟ 6297143586 ⟟ Call Me For Genuine Sex...
Top Rated Pune Call Girls Viman Nagar ⟟ 6297143586 ⟟ Call Me For Genuine Sex...Call Girls in Nagpur High Profile
 
Call Girls Rajgurunagar Call Me 7737669865 Budget Friendly No Advance Booking
Call Girls Rajgurunagar Call Me 7737669865 Budget Friendly No Advance BookingCall Girls Rajgurunagar Call Me 7737669865 Budget Friendly No Advance Booking
Call Girls Rajgurunagar Call Me 7737669865 Budget Friendly No Advance Bookingroncy bisnoi
 

Kürzlich hochgeladen (20)

Kharghar Blowjob Housewife Call Girls NUmber-9833754194-CBD Belapur Internati...
Kharghar Blowjob Housewife Call Girls NUmber-9833754194-CBD Belapur Internati...Kharghar Blowjob Housewife Call Girls NUmber-9833754194-CBD Belapur Internati...
Kharghar Blowjob Housewife Call Girls NUmber-9833754194-CBD Belapur Internati...
 
VIP Call Girl in Mumbai 💧 9920725232 ( Call Me ) Get A New Crush Everyday Wit...
VIP Call Girl in Mumbai 💧 9920725232 ( Call Me ) Get A New Crush Everyday Wit...VIP Call Girl in Mumbai 💧 9920725232 ( Call Me ) Get A New Crush Everyday Wit...
VIP Call Girl in Mumbai 💧 9920725232 ( Call Me ) Get A New Crush Everyday Wit...
 
Call Girls Banaswadi Just Call 👗 7737669865 👗 Top Class Call Girl Service Ban...
Call Girls Banaswadi Just Call 👗 7737669865 👗 Top Class Call Girl Service Ban...Call Girls Banaswadi Just Call 👗 7737669865 👗 Top Class Call Girl Service Ban...
Call Girls Banaswadi Just Call 👗 7737669865 👗 Top Class Call Girl Service Ban...
 
Diva-Thane European Call Girls Number-9833754194-Diva Busty Professional Call...
Diva-Thane European Call Girls Number-9833754194-Diva Busty Professional Call...Diva-Thane European Call Girls Number-9833754194-Diva Busty Professional Call...
Diva-Thane European Call Girls Number-9833754194-Diva Busty Professional Call...
 
Booking open Available Pune Call Girls Shivane 6297143586 Call Hot Indian Gi...
Booking open Available Pune Call Girls Shivane  6297143586 Call Hot Indian Gi...Booking open Available Pune Call Girls Shivane  6297143586 Call Hot Indian Gi...
Booking open Available Pune Call Girls Shivane 6297143586 Call Hot Indian Gi...
 
VIP Kalyan Call Girls 🌐 9920725232 🌐 Make Your Dreams Come True With Mumbai E...
VIP Kalyan Call Girls 🌐 9920725232 🌐 Make Your Dreams Come True With Mumbai E...VIP Kalyan Call Girls 🌐 9920725232 🌐 Make Your Dreams Come True With Mumbai E...
VIP Kalyan Call Girls 🌐 9920725232 🌐 Make Your Dreams Come True With Mumbai E...
 
Vip Call US 📞 7738631006 ✅Call Girls In Sakinaka ( Mumbai )
Vip Call US 📞 7738631006 ✅Call Girls In Sakinaka ( Mumbai )Vip Call US 📞 7738631006 ✅Call Girls In Sakinaka ( Mumbai )
Vip Call US 📞 7738631006 ✅Call Girls In Sakinaka ( Mumbai )
 
Best VIP Call Girls Morni Hills Just Click Me 6367492432
Best VIP Call Girls Morni Hills Just Click Me 6367492432Best VIP Call Girls Morni Hills Just Click Me 6367492432
Best VIP Call Girls Morni Hills Just Click Me 6367492432
 
Call Girls in New Friends Colony Delhi 💯 Call Us 🔝9205541914 🔝( Delhi) Escort...
Call Girls in New Friends Colony Delhi 💯 Call Us 🔝9205541914 🔝( Delhi) Escort...Call Girls in New Friends Colony Delhi 💯 Call Us 🔝9205541914 🔝( Delhi) Escort...
Call Girls in New Friends Colony Delhi 💯 Call Us 🔝9205541914 🔝( Delhi) Escort...
 
falcon-invoice-discounting-unlocking-prime-investment-opportunities
falcon-invoice-discounting-unlocking-prime-investment-opportunitiesfalcon-invoice-discounting-unlocking-prime-investment-opportunities
falcon-invoice-discounting-unlocking-prime-investment-opportunities
 
Call Girls Koregaon Park Call Me 7737669865 Budget Friendly No Advance Booking
Call Girls Koregaon Park Call Me 7737669865 Budget Friendly No Advance BookingCall Girls Koregaon Park Call Me 7737669865 Budget Friendly No Advance Booking
Call Girls Koregaon Park Call Me 7737669865 Budget Friendly No Advance Booking
 
Top Rated Pune Call Girls Dighi ⟟ 6297143586 ⟟ Call Me For Genuine Sex Servi...
Top Rated  Pune Call Girls Dighi ⟟ 6297143586 ⟟ Call Me For Genuine Sex Servi...Top Rated  Pune Call Girls Dighi ⟟ 6297143586 ⟟ Call Me For Genuine Sex Servi...
Top Rated Pune Call Girls Dighi ⟟ 6297143586 ⟟ Call Me For Genuine Sex Servi...
 
(INDIRA) Call Girl Mumbai Call Now 8250077686 Mumbai Escorts 24x7
(INDIRA) Call Girl Mumbai Call Now 8250077686 Mumbai Escorts 24x7(INDIRA) Call Girl Mumbai Call Now 8250077686 Mumbai Escorts 24x7
(INDIRA) Call Girl Mumbai Call Now 8250077686 Mumbai Escorts 24x7
 
W.D. Gann Theory Complete Information.pdf
W.D. Gann Theory Complete Information.pdfW.D. Gann Theory Complete Information.pdf
W.D. Gann Theory Complete Information.pdf
 
Business Principles, Tools, and Techniques in Participating in Various Types...
Business Principles, Tools, and Techniques  in Participating in Various Types...Business Principles, Tools, and Techniques  in Participating in Various Types...
Business Principles, Tools, and Techniques in Participating in Various Types...
 
VIP Call Girl Service Andheri West ⚡ 9920725232 What It Takes To Be The Best ...
VIP Call Girl Service Andheri West ⚡ 9920725232 What It Takes To Be The Best ...VIP Call Girl Service Andheri West ⚡ 9920725232 What It Takes To Be The Best ...
VIP Call Girl Service Andheri West ⚡ 9920725232 What It Takes To Be The Best ...
 
Mira Road Awesome 100% Independent Call Girls NUmber-9833754194-Dahisar Inter...
Mira Road Awesome 100% Independent Call Girls NUmber-9833754194-Dahisar Inter...Mira Road Awesome 100% Independent Call Girls NUmber-9833754194-Dahisar Inter...
Mira Road Awesome 100% Independent Call Girls NUmber-9833754194-Dahisar Inter...
 
VIP Independent Call Girls in Mumbai 🌹 9920725232 ( Call Me ) Mumbai Escorts ...
VIP Independent Call Girls in Mumbai 🌹 9920725232 ( Call Me ) Mumbai Escorts ...VIP Independent Call Girls in Mumbai 🌹 9920725232 ( Call Me ) Mumbai Escorts ...
VIP Independent Call Girls in Mumbai 🌹 9920725232 ( Call Me ) Mumbai Escorts ...
 
Top Rated Pune Call Girls Viman Nagar ⟟ 6297143586 ⟟ Call Me For Genuine Sex...
Top Rated  Pune Call Girls Viman Nagar ⟟ 6297143586 ⟟ Call Me For Genuine Sex...Top Rated  Pune Call Girls Viman Nagar ⟟ 6297143586 ⟟ Call Me For Genuine Sex...
Top Rated Pune Call Girls Viman Nagar ⟟ 6297143586 ⟟ Call Me For Genuine Sex...
 
Call Girls Rajgurunagar Call Me 7737669865 Budget Friendly No Advance Booking
Call Girls Rajgurunagar Call Me 7737669865 Budget Friendly No Advance BookingCall Girls Rajgurunagar Call Me 7737669865 Budget Friendly No Advance Booking
Call Girls Rajgurunagar Call Me 7737669865 Budget Friendly No Advance Booking
 

RegObs Special Report - Bankosaurus jumping on the asteroid - by A_Filtri & Team - pp 86.pdf

  • 1. IMPORTANT DISCLOSURE FOR U.S. INVESTORS: This document is prepared by Mediobanca Securities, the equity research department of Mediobanca S.p.A. (parent company of Mediobanca Securities USA LLC (“MBUSA”)) and it is distributed in the United States by MBUSA which accepts responsibility for its content. The research analyst(s) named on this report are not registered / qualified as research analysts with Finra. Any US person receiving this document and wishing to effect transactions in any securities discussed herein should do so with MBUSA, not Mediobanca S.p.A.. Please refer to the last pages of this document for important disclaimers. RegObs Special Report 07 December 2022 Update Bankosaurus jumping on the asteroid RegObs Special Report - Digital euro: the ECB saving Europe again Sign up to our Digital Currency Editorial Andrea Filtri Equity Analyst +44 203 0369 571 Andrea.Filtri@mediobanca.com Alberto Nigro Equity Analyst +39 02 8829 9540 Alberto.Nigro@mediobanca.com Jordan Bartlam Equity Analyst +44 203 0369 692 Jordan.Bartlam@mediobanca.com Debunking the bank-dinosaur vs tech-asteroid myth The rise of new technologies and of Big Tech/fintech companies based on them has been juxtaposed to incumbent, slow-moving banks so far that the oversimplified metaphor of the bank-dinosaur at risk of extinction from the tech-asteroid has been widely utilised by mainstream media; chasing the evergreen popularity of the David vs Goliath duel. This report provides elements confirming the threat is there, but that the Bankosaurus could still evolve by embracing technology, gaining agility to jump on the asteroid and buying the ticket for the perpetuation of the species. Disruptive technologies: DLT, MLAI and quantum; embrace to evolve We identified machine learning/artificial intelligence (MLAI), distributed ledger technology (DLT, aka blockchain) and quantum computing (QC) as 3 innovations which could disrupt the banking model (among many others) as we know it. MB surveys: banks know they must change, they just do not know how to… yet We engaged EU banks with 2 surveys to position them on the transformation journey: 1) gauging where they stand on MLAI, DLT, QC, 2) showcasing of use cases. With c.20 respondents, we draw some conclusions: a) the majority of investment is going into low-key AI/chatbots, b) short inception-to-production, high external staff reliance and efficiency-focus either reflect good project management or room to move up along the innovation vector, c) low DLT penetration embodies aversion to cooperation with peers and the need to fully control investments, d) absence of payment projects echoes the exit from the space, a long-term strategic mistake to us, while quasi-absence of QC is justified; it is just too early for it. Banks know they must change at some point; they are just not daring/worried enough and mounting external pressure could help them. MB methodology: three steps to find the cure - Step 1: breaking the bank We have gone into our lab to find a 3-step cure for the endangered species. EU banks adopt a vertically integrated model mixing very different businesses and the first step is to ringfence the affected organs. We separated standardised banking – businesses more exposed to technological disruption – from customised banking - which is more immune – of the 12 most representative EU banks (30% market share). This shows 2/3 of operating income, 60% of costs and 50% of loans belong to standardised banking, under threat from technological disruption. Step 2: digital disruption eroding 30% of standardised revenues We looked at sectors already disrupted by tech: film rental, photography, local taxis and retail bookstores. Anecdotal evidence points to 30-50% margin erosion. We simulate a central case made of c.30% compression of standardised banking revenues, requiring c.60% cost cuts to hold profitability steady or C/I going from c.50% to <30%, better than best-in-class Nordic banks. Step 3: the great (cost) reset We developed the DLT cost function, validated by Spunta’s case, featuring fixed costs progressively handing over to variable transaction costs as volumes pickup. Slimmed-up banks can jump on the asteroid We simulated the migration of EU standardised banking onto a single, hypothetical DLT platform maximising scale economies. Our simulation indicates a collapse in aggregate cost, where MLAI and QC would also contribute, protecting sector profitability. Our new, DLT based cost estimate shows a margin of error of c.400x, still leaving banks on current profitability, even after revenue attrition from digital disruption. Hence, cost miniaturisation is the evolution banks need to not only avoid extinction, but potentially to thrive on higher profitability.
  • 2. RegObs Special Report 07 December 2022 ◆ 2 Contents Executive Summary ................................................................................................ 3 Where we left off .................................................................................................. 6 New technologies: evolution or extinction? .................................................................... 9 MB survey suggests A.I. and cost efficiency in focus… ....................................................... 11 …seconded by our “use case” almanac ........................................................................ 15 Good effort, but far from transformative innovation yet ................................................... 19 “breaking the bank”: isolating the species at risk ........................................................... 21 Digital disruption carries 30-50% margin compression… ..................................................... 26 …or standardised revenues -30%; costs -60% to offset ....................................................... 27 The great (cost) reset: scaling up DLT to do the trick....................................................... 36 Evolution: cost miniaturisation to avoid extinction .......................................................... 39 Digital Deep Dive.................................................................................................. 43 Digital disruption: case studies in-depth ...................................................................... 44 Distributed ledger technology (DLT), aka blockchain........................................................ 48 Machine Learning / Artificial Intelligence ..................................................................... 53 Quantum Computing.............................................................................................. 59 Mediobanca digital benchmark survey ......................................................................... 70 Showcasing banks’ use cases .................................................................................... 72 Annex ............................................................................................................... 78 List of Pictures .................................................................................................... 79 List of Tables ...................................................................................................... 80 References ......................................................................................................... 81
  • 3. RegObs Special Report 07 December 2022 ◆ 3 Executive Summary Debunking the bank-dinosaur, technology-asteroid myth The rise of new technologies and Big Tech/fintech companies based on them has been often journalistically contraposed to slow-moving, anachronistic commercial bank pachyderms. In the oversimplified manner complex dynamics are often taken to the wider public, the metaphor of the bank-dinosaur at risk of extinction from the technology-asteroid (fintech/Big Tech aggression) has been widely disseminated by the mainstream media; chasing the evergreen popularity of the David vs Goliath duel. This report argues that the threat is there, but that the “dinosaur” could well decide to mutate by embracing new technologies, gaining agility, adapting to the new ecosystem and ultimately jumping on the “asteroid” to buy a survival ticket for the perpetuation of the species. Banking Threat #1: in 2021, the digital euro In Spring 2021 we explained digital currencies and how, more particularly, central bank digital currency (CBDC) could be a threat to commercial banking deposits (see RegObs Special Report - Digital euro: the ECB saving Europe again - by A.Filtri & Team). The conclusion was that the digital euro could both defend monetary sovereignty from the thread of euro displacement by foreign (private or public) digital currencies and – a European peculiarity – take the European integration process to the next level by mutualising banking risk via the ECB as, with the digital euro, the central bank could become the primary source of funding for commercial banks. Banking Threat #2: new technologies (MLAI, DLT, QC) today After much studying and learning about new technologies, we have identified Machine Learning/Artificial Intelligence (MLAI), Distributed Ledger Technology (DLT, aka blockchain) and Quantum Computing (QC) as three powerful innovations which could disrupt the banking model (among many others) as we know it. (see DLT, MLAI and QC for in-depth background material). MLAI – Machine Learning/Artificial Intelligence can be thought of as an increasingly less supervised spectrum (“Symbolic” up to “Strong”) of processes to be operated and executed by machines. However, while thoughts immediately fly to Terminator, we are still scrambling on the ground. Banking is currently positioned at the weaker end of the spectrum, such as “ChatBots”. DLT - Distributed ledger technology (DLT) - more commonly known as blockchain – is a tool to provide trust between parties relying on the same data records, without the need for a centralised, third- party authority. The immense data management potential introduced by the technology contrasts with the very limited adoption by banks thus far due to their adversity to cooperation and shared control. QC - Quantum computing is a new technology taking computation power to a new dimension, leveraging on quantum physics. We are still in the development stage of the technology, so that it is just too early to see its adoption by the industry. Banks’ psychology: rational reaction or amygdala hijack? Evolve or go extinct Over the millennia, human beings have developed autopilot response mechanisms to escape immediate threats. When sudden life-threatening situations arise, we react by instinct, without thinking. We naturally experience the amygdala hijack, i.e. when rational thought is inefficient or too slow to get us out of trouble, we behave in patterns, usually attack or flight, to increase our chances of survival. Whereas this could be life-saving when suddenly we find ourselves in front of an incoming bus, the amygdala hijack can work against us in more complex, real life situations, when we usually have the time to evaluate the option and rationally choose the action with the highest chance of success. Another instinctive reaction human beings tend to have before a threat is to freeze, put the head under the sand and do nothing about it, hoping that it will go away.
  • 4. RegObs Special Report 07 December 2022 ◆ 4 Our banking-psychology expertise imposes us to counsel banks: a) to accept that there is a clear threat and that it is coming, b) that there is enough time to avoid an instinctive reaction and instead adopt a rational behaviour, c) not to be complacent with the time allowed and get on with it. Our work should help fast-tracking banks through the investigation phase and getting them working towards a solution, with the comforting conclusion that – if they act in a timely fashion and in an adequate response – their existence is not only under question, but it could even prosper. Mediobanca surveys: banks know they must change, but they do not know how to… yet To take a snapshot of the current location of European banks along their transformation journey, we send them two surveys: 1) a multiple choice questionnaire to gauge where each bank is on MLAI, DLT, QC, 2) a showcasing of current success projects. We thank the c.20 respondents, whose responses we aggregated to describe the overall state of the sector on innovation. The two surveys offer coherent pictures, allowing us to take the following conclusions: a) banks know they need to change at some point and they are progressively increasing their engagement, b) the majority of their efforts are going in a linear evolution of the current operating model, with heavy investment in chatbots as the “new automation”, not quite a radical rethink, yet, c) short inception-to-production timing, high usage of external staff and focus on efficiency gains either reflect good project management and cost discipline or room to increase ambition along the innovation vector, with innovation leaders often incentivised by short-term, more quantifiable results rather than by transformative, multi-year projects, d) low penetration of DLT reflects the additional complexity of the technology requiring cooperation with peers/competitors and the desire to fully control each investment, e) the absence of payment projects reflects the exit by banks from the space, potentially a strategic mistake in the long-term given the scale, the innovation know-how and the key role of the business within banks’ product offering to clients, f) the quasi-absence of QC is justified in our view. The technology is just not mature enough to invest into it now; it will come, it is just too early for the time being. We conclude that banks have realised they must change. They are just not daring enough or in enough of a hurry to accelerate their more radical, self-initiated mutation. External pressure (market, regulator, competitors) could improve self-awareness going forward, we suspect. Mediobanca Lab: three steps to finding the cure Concerned with the potentially endangered species, we have gone into our laboratory to diagnose the patient by analysing the threats and working out a possible treatment through a three-step methodology. We see the current banking model at risk, hence requiring a deep genotype mutation to evolve and survive. We are afraid there is no plan B. Step 1 - “breaking the bank”: isolating the species at risk of extinction European banks adopt a vertically integrated model, bundling together retail banking services (deposits, current accounts, mortgages and personal loans/credit cards, payments) with corporate & investment banking services (SME and corporate lending, markets, advisory) and wealth management (private banking, asset management, insurance, asset gathering) and specialised financial services (car financing). Technology is not threatening such a vast variety of businesses at once. While QC still needs some time to hit the ground with its superior computational power, MLAI and DLT offer vast enhancements in the interpretation and in the handling of data, transforming processes/products which can be digitised and standardised into platforms able to support enormous scale economies. We attempted a surgical procedure separating standardised banking – more exposed to technological disruption – from customised banking - where human capital and personal advisory will remain key – of the twelve most representative European banks, attracting 30% market share. This shows 2/3 of operating income, 60% of costs and 50% of loans belong to standardised banking.
  • 5. RegObs Special Report 07 December 2022 ◆ 5 Step 2 – Digital disruption: 30% revenue attrition from margin compression We look at prior sectors disrupted by technological transformation such as the film rental, photography, local taxis and retail bookstores sectors. Anecdotal evidence points towards 30-50% margin compression (you can read our in-depth case studies here). We simulate three scenarios (base, pessimistic, optimistic) of revenue attrition to standardised banking. Our central case indicates c.30% erosion in standardised revenues, requiring c.60% cost reduction to hold profitability steady, implying standardised banking cost/income should shrink from c.50% today to <30%, tomorrow, ahead of the currently most efficient Nordic banks. Step 3 – The great (cost) reset: scaling up DLT to do the trick Despite being in vogue, DLT technology is still vastly used merely in small, experimental projects. This makes it hard to extrapolate what it could/would do, if applied to a scale business, either assuming the trilemma (i.e. the right balance of decentralisation, scalability, security) will be solved at some point or that permissioned-DLT can do so already. Our intense dialogue with industry experts helps us to move from theory to practice in formulating the cost function of a (permissioned) DLT network, featuring fixed costs which progressively handover to variable transaction costs as volumes pickup. The Spunta DLT case validates our function, which embeds a margin of conservatism. Our conclusion is that DLT technology could provide banks with astronomic cost efficiencies in some of their businesses. The treatment: evolution in cost miniaturisation to avert extinction and revive banking We simulate the migration of European standardised banking onto a single DLT network, maximising scale economies. Our simulation indicates the change would see a collapse vs current aggregate cost, where MLAI and QC would also clearly contribute. System-wide DLT costs could be c.400x larger than we estimated to still leave banks on current profitability level, even after accounting for revenue attrition from digital disruption. We conclude that potential for cost miniaturisation from the industrialisation of standardised banking products/processes represents a thick buffer to absorb revenues pressures under very adverse scenarios. Hence – in the wake of the digital disruption the banking industry is about to experience – instead of feeling threatened by extinction from fintechs/Big Tech, banks should embrace new technologies to revive their standardised businesses. This mutation would not only grant their survival, but it could possibly see them even thriving on low-cost structures applied to large-volume markets. 2 notes in 1 report: the 1st for the experts, the 2nd providing background knowledge This report has been conceived for a very heterogenous reader, ranging from private market and public market investors, industry experts, innovators, regulators, supervisors, authorities and people passionate about innovation. We have therefore composed the report of two separate notes: 1) the first eleven chapters for more digital-savvy readers, 2) the additional chapters are for those needing background and deeper technical knowledge. In such case, it is probably better reading in reverse order. Credits and acknowledgements Our deepest gratitude goes to the many friends, colleagues, experts and professionals who have supported the investigation and knowledge development necessary to produce this report and to the many banks which have contributed to our surveys. We hope your efforts have been worthwhile, we have certainly enjoyed the interaction and learned plenty from it. Further thanks go to the patience and passion of our readers and clients whose support is crucial to allow us to continue developing our expertise and accompany along their investment journey.
  • 6. RegObs Special Report 07 December 2022 ◆ 6 Where we left off Our debut work on digital currency and CBDC outlined the types of digital currency and how it came about. It explained the mounting pressure on western central banks to develop their digital currency (CBDC) starting with the initially ignored pressure from the development of bitcoin, to the much cleared threats of the Facebook-sponsored Libra/Diem to the advanced stage of the Chinese CBDC, DC/EP. We elaborated around the intricacies of currency and geopolitics, outlining how the US currently benefit from a hegemonic USD which – combined to SWIFT – allows them oversight and influence over third countries via their sanction regime. Digital currency offers a technological innovation making current systems potentially obsolete. For Europe, the ECB is seeing the development of a digital euro as key to defend monetary sovereignty. Yet, depending on the different depths of implementation, the digital euro could potentially destabilise the current banking model, putting the ECB at the heart of the financial system, with the (desirable for Brussels/Frankfurt) side effect of making a leap in European integration. Since our publication, we have seen the successive events supporting our view: the ECB is proceeding steadily towards the development of a digital euro. While US-China tensions continue, China has launched the digital Yuan and the US are looking to recover the lost ground in the space, after having sunk Libra/Diem, embracing the development of the d$. CBDC, a digital currency with vast implications for geopolitics and financial stability… Our report on the Digital euro (Digital euro: the ECB saving Europe again, 8 March 2021) explored the digital currency world, the drivers behind the development of Central Bank Digital Currency (CBDC), and the connection between currency and geopolitics. We identified in Facebook’s Diem/Libra initial stablecoin project and in the advanced Chinese CBDC project the main reasons why most western central banks are currently studying the development of CBDC. By more easily creating new, parallel networks for payments, we argued that digital currency can hamper the current domain of the US and of the US dollar on global trade and reserves - granted by the SWIFT messaging payment system and by the Patriot Act – providing the US with oversight and coercion on third countries via their sanctions regime. In this context, we argued that within the growing tensions between the US and China, Europe is not well positioned on the future strategic assets: data, data networks, semiconductors, rare earths and currency. Hence, the development of a digital euro is the only European strategic card at hand. …possibly having positive side effects on the European integration process In our digital euro report, we elaborated on the different forms the project could take, from a “constrained” (i.e. with a cap), retail CBDC - more aligned to the initial ECB goals - which we deemed compatible with financial stability, to an “unconstrained” one (i.e. no cap) - more aligned with European Commission ambitions - potentially requiring a deeper rethink of how the financial system operates. In the latter case, we concluded a large sum of banking deposits could migrate to the ECB, which would in turn redeposit the sum into the banks, de facto mutualising risk across EU countries as a side effect of equipping Europe with its own strategic asset. The latest events seem to support our thesis A lot has changed since March 2021, we would argue in support of our original thesis. We summarise the main events: • GEOPOLITICAL DOMINANCE: the Ukrainian war. The Russian invasion of Ukraine has magnified the geopolitical considerations we made then, flagging Europe’s weak strategic independence. • EUROPE: integration speedup. We have had a string of elements showing Europe is proceeding towards more integration, post Brexit, post pandemic and post Ukraine’s invasion:  European army – The disorderly abandonment of Afghanistan by European forces, dictated by the American ally without a discussion with EU partners, the recent EU-Russia
  • 7. RegObs Special Report 07 December 2022 ◆ 7 controversy on natural gas supply and the AUKUS agreement with the US forcing France out of a large submarine supply to Australia revived the debate over European military autonomy for common defence and to safeguard European interest on the global map. Italy’s President Mattarella, France’s President Macron, France’s Finance Minister Le Maire and EU Commission President Von der Leyen spoke at the unison about the need for Europe to think about gaining military autonomy (see Download 2021_09_16_EU policy - EU agenda = strategic independence from theory to implementation.pdf).  Energy Union – Germany is emerging from the ongoing Russia-Ukraine conflict as the weakest link: depending on Russia for energy and on China for exports. Economic sanctions to Russia have triggered Russian choking of the gas flow to Europe, testing Europe’s resilience. Ongoing talks of energy solidarity entail European partners sharing the pain on the risk of energy shortage, while progressive steps towards energy diversification and independence are made over time. Finding an agreement on common gas purchases, interconnection of networks and mutual assistance would be a massive step towards a more integrated Europe, in our view.  EU microchips – the lack of availability of semiconductors from Asia is eroding c.1/4 of the European car production, with a normalisation expected not before 2023, confirming the dependency of Europe on its largest sector by employment. The EU is working towards boosting European manufacturing of microchips by 2030, doubling the market share at global level from 10% to 20%.  The recovery fund – in the end this project gravitates around digital investment and the green transition, i.e. digital and energy independence. EUROPE: advanced stage of the investigation phase of the digital euro. The ECB has spent the past fourteen months investigating how the digital euro could look like. We see the following key features having emerged from the process: privacy (no anonymity, no big brother), integration of supervised intermediaries in the d€ distribution, inhibition of use of d€ as reserve of value and ECB-centred settlement model. We still see the definition of the following features pending: transfer mechanism and offline, architecture (DLT vs centralized), form factor (ECB app vs integrated within banks’ apps). Politicians have recently embraced the project, materially raising its chances of launch, in our view. CHINA: relentlessly progressing on its strategic path. Unsurprisingly, China is pursuing its objectives: • Mounting pressure on Taiwan – China is pursuing its quest to gain more military relevance by boosting investments and showing its muscles off. Hypersonic missiles, a larger navy fleet, and the mounting pressure on asserting its sovereignty over Taiwan. • Taking back control over digital payments and fighting crypto… – The overhaul of Chinese digital payments is complete. From the failed IPO of Ant Financial to date, China has taken back control over digital payments, previously firmly in the hands of the Wechat-Alipay duopoly. Meanwhile, China banned cryptocurrency, confirming the tough stance on private digital currencies. (Download 2021_09_24_Digital currency - The Regulators are marching in (part 4) - China bans crypto.pdf). • …replacing them with the digital Renminbi – China was fast in proposing the alternative to the Alipay-We Chat digital payments duopoly by launching the digital Renminbi in Feb 2022. (see article). This confirms China’s head start on the west on this front. US: finally biting the bullet. Meanwhile, the US have seen the Biden administration take decisive steps on geopolitics and digital currency: • Shift to Indo-Pacific focus – The US retreat from Afghanistan has attracted the attention of the global press and has been interpreted as the need to concentrate efforts on China. The AUKUS pact shows the firm intentions of the US to form an alliance in the region to foster
  • 8. RegObs Special Report 07 December 2022 ◆ 8 oversight over China. Taiwan has been at the centre of the US-China confrontation, reaching the apex on the visit to Taipei of US House Speaker, Nancy Pelosi. • Support to Ukraine – the US has stood firmly behind Ukraine from even before the Russian invasion, warning about its imminency beforehand and through the supply of support, weapons, intelligence and funds. • Decisive action in digital currency regulation – The Biden administration removed the prior ambiguity in the regulation of cryptocurrencies. On one hand, the US stopped for good Facebook-sponsored Libra/Diem and the SEC stepped in on Coinbase’s Lend initiative and asked all players in the space to consider application to SEC regulation. On the other, and more importantly, the Administration put the development of a digital dollar as a strategic priority. LIBRA/DIEM: game over, not so for private stablecoins. We have been arguing for a long time the initial Libra/Diem project was indigestible for regulators but that the Facebook initiative had merits on its technical advancements (see Download 2021_05_27_Digital currency - DIEM turnaround from the worst enemy to the central banks' best friend.pdf). Diem attempted a sudden u-turn, in vain: 1. Abandoning Switzerland to return home – Diem left the Swiss venture returning to the US. 2. Teaming up with a Fed-regulated Californian bank to issue digital currency – It teamed up with a Fed-regulated Californian bank (Silverbank) to issue a USD stablecoin to back the USD- Diem (see Download 2021_05_13_Digital currency - Carpe USDIEM - Uncle Sam calling home.pdf). 3. Opened up to abandoning the stablecoin in favour of CBDC – At a livestreaming which we hosted back in July-21, Christian Catalini, co-founder of Diem, confirmed that they would be prepared to abandon the stablecoin in favour of CBDC if and when central banks would be ready for it, offering DIEM platform as a partner for the CBDC project. (see Download 2021_07_08_Digital currency - Co-founder of Facebook sponsored DIEM confirmed repositioning.pdf, Youtube link, Download 2021_06_24_Digital currency - ECB proceeding steady vs Diem's regulatory flirting__1.pdf). DIEM was sold to Silvergate bank in January 2022, ending Facebook’s journey in digital currency. This unhappy ending does not mean that all private stablecoins should suffer the same fate. Big Tech is working to find the most convenient solution to benefit from this new instrument. CRYPTO: US rates spoiling the party. The digital currency world is in a crunch, started by the new USD rates cycle and the FED’s balance sheet shrinkage which is drying the ample liquidity and trigger- happy investment into DeFi. Over the past period we have gone from “limitless partying” to “crypto winter”, i.e. from a phase of proliferation to one of harsh rationalisation in crypto land, even stained by cases of fraud.
  • 9. RegObs Special Report 07 December 2022 ◆ 9 New technologies: evolution or extinction? This report flips the approach to technological implications for banks upside-down vs its digital euro predecessor, i.e. from a top-down view to a bottom-up perspective. Here, we focus on the disruptive potential of three new technologies: 1) artificial intelligence/machine learning (MLAI), 2) distributed ledger technology (DLT) and 3) quantum computing (QC). We imagine how these could redefine the way banks are and operate, or if instead they could disrupt banks to the point of ridding of them in full. Borrowing from the often-used bank/dinosaur metaphor, are new technologies the asteroid bringing banks to extinction or could banks not just embrace them and jump on the asteroid, i.e. evolving rather than succumbing to the innovation tsunami? The report bases on three pillars: 1) two surveys to take a snapshot of where banks are on their technological transformation, 2) our diagnosis of how technology could transform parts of the banking businesses and our take on the final outcome for survival (profitability), 3) in-depth background knowledge on the prior cases of technological disruption and on the three technologies. The first step in solving a problem is recognising it. Banks are progressively becoming more conscious of it but are yet scrambling both what to do about it and when. Before an existential threat, there should be no taboos. At the same time, every threat hides an opportunity: banks need to identify it and pursue it. Getting there first usually pays dividends… From external threat… Our digital currency work identified this new tool as a serious potential threat for commercial banks, particularly in the form of central bank digital currency (CBDC), where central banks can potentially become competitors not only on the funding side. This is a case where technological innovation can disrupt existing, stable, regulated sectors and markets. …to internal rethink From time to time, the advent of new technologies allows the redefinition, re-engineering and re- thinking of companies, sectors, business models and processes. We have several evidences of this in airlines, local taxis, music distribution, retail, etc. For banking, the advent of the internet was first predicted to extinguish banks, which then embraced change and are on course to rebalance their distribution channels towards remote/digital ones vs traditional branches. Technological transformation represents another challenge as it has the potential to be more of a revolution than an evolution, testing the survival of the species, meaning that processes, products – and potentially even banks as a whole - could require deep rethinking rather than adaptation. MLAI, DLT, quantum computing in focus We focussed on the main technological innovations, namely artificial intelligence/machine learning (MLAI), distributed ledger technology (DLT) and quantum computing (QC), which we describe in depth later in the report. They have very different stages of development, business implications and requirements to be implemented (see DLT, MLAI and QC for in-depth background material). MLAI - MLAI can be thought of as an increasingly less supervised spectrum (“Symbolic” up to “Strong”) of processes to be operated and executed by machines. However, while thoughts immediately fly to Terminator, we are still scrambling on the ground. Banking is currently positioned at the weaker end of the spectrum, such as “ChatBots”. DLT - Distributed ledger technology (DLT) - more commonly known as blockchain – is a tool to provide trust between parties relying on the same data records, without the need for a centralised, third- party authority. The immense data management potential introduced by the technology contrasts with the very limited adoption by banks thus far due to their adversity to cooperation and shared control.
  • 10. RegObs Special Report 07 December 2022 ◆ 10 QC - Quantum computing is a new technology taking computational power to a new dimension, leveraging on quantum physics. We are still in the development stage of the technology, so that it is just too early to see its adoption by the industry. Banks need to perceive the risk The first step to solving a problem is to recognise it. Banks are more and more becoming conscious of the threat, but they have not quite worked out what to do about it and to what degree of urgency, in our view. Preservation instinct means there is no room for taboos Preservation instinct kicks in when a creature feels its life is in danger. This puts everything else behind in terms of priority. In this parallel, banks should consider making radical changes to how they look like, how they work and what they do, to adapt to the new environment. If technological innovations are bringing a change of ice age, studying them should be urgent, so to have time to adapt to the new ecosystem. Every threat hides an opportunity We believe technological disruption represents both a threat and an opportunity for banks. The novelty new technologies introduced is to be able to manage vast amounts of activity at much higher efficiency levels. On one hand, they allow for a material increase in competition – both within the industry and from newcomers – likely eroding margins and leading to consolidation. On the other, the increased transparency which comes with them could trigger a positive volume effect and a radical cost reduction. Getting there first usually represents a material advantage…
  • 11. RegObs Special Report 07 December 2022 ◆ 11 MB survey suggests A.I. and cost efficiency in focus… To better understand the current state of play when it comes to adoption of new technologies in the banking sphere, we enquired European banks on two fronts. The first – our “benchmark survey” – is more simplistic and standardised, helping us to map out an industry standard. We thank the c.20 participating banks which, for the sake of anonymity, we won’t name. The second - our case study almanac – is described in the next chapter. We summarise our findings in a synthetic aggregate benchmark embodied by our “Banca Benchmark”, a large, fictional European bank: 1) Banca Benchmark spends a somewhat modest €0.5bn per year on developing applications that use DLT/MLAI/QC technologies, although the overwhelming majority (>80%) is going into AI (mostly chatbots). 2) Banca Benchmark mostly aims at better cost efficiency, with revenue enhancement and customer experience also important, but to a lesser extent. 3) Banca Benchmark conceives Fintechs as partners in the digital journey. 4) Banca Benchmark believes that adoption of these technologies will not just cut costs but change the entire role that banks play within the industry, emblematic perhaps of a transformation of the whole banking business model, with Risk Management & Control, Digital Channels and Back Office & IT most exposed to this transformation. 5) Banca Benchmark is more focused on innovating back office functions and cutting costs rather than transforming products and the way they are conceived and offered. MB-designed benchmark and a use-case surveys to map EU banks tech activity Looking primarily at DLT/MLAI/QC technologies we see as defining the new banking landscape, we wanted to understand better what banks had already been doing (or were planning on doing). As such, we queried banks along two directions. The first was through a survey designed to build a benchmark across our respondent banks, we refer to this as our “Benchmark Survey”. Secondly, we asked banks to showcase their successes which we aggregated in our almanac in the next chapter. The Mediobanca Innovative Technology Benchmark Survey Our benchmark survey consisted of six questions focused primarily on gauging the size and allocation of the investment budget into the three technology groups, as well as what part of the bank was likely to be most impacted by the adoption of the technologies. Whilst not mentioning any bank by name, we share the results on an aggregated basis across the c.20 respondent banks. Banca Benchmark We aggregate the responses received into a fictional Banca Benchmark, representing EU banks. This is strong of c.€250bn market capitalisation, investing a modest €0.5bn a year on A.I./machine learning, quantum computing and DLT, with an overwhelming dominance of A.I. and mainly for risk management, digital channels and back office purposes. The main goal is usually cost saving, with projects more and more in partnerships with fintechs.
  • 12. RegObs Special Report 07 December 2022 ◆ 12 Picture 1: Benchmark – Typical response from the benchmark survey (embodied by “Banca Benchmark”) Source: Mediobanca Securities €0.5bn annual budget appears small Based on size of the amount of budgeted investment into the three technologies and investment horizon, we detected an aggregate €0.5bn annual budget. On one hand, this hopefully reflects more carefully investments in the three technologies (DLT/MLAI/QC) as opposed to more generic IT budgets. On the other, this suggests that on average banks are not yet investing massive amounts in this direction. Cost efficiency and machine learning the focus We asked respondents how they were splitting their investments across the different technologies and the ultimate purpose of the investment. On average, nearly one-third of the investment was directed at cost efficiency. Whilst, around 20% was aimed at boosting revenues, enhancing customer experience and improving operational risk. How this was split across the different technologies was much more skewed, with c.80% going into MLAI. The remainder was split 15% into DLT and just 3% into Quantum. Figure 1: Investment split (purpose, %) Source: Mediobanca Securities Figure 2: Investment split (technology, %) Source: Mediobanca Securities Market Cap: €252bn Investment (€mn): €0.5bn/year Investment (split): AI/ML 82% DLT 15% Quantum 3% Main part of bank: Risk management/control Digital Channels Back office/IT Main goal: Cost sav ings View on FinTechs: Partners Banca Benchmark 30% 22% 21% 17% 10% Cost efficiency Revenue Boost Customer Experience Fraud, Operational Risk Other Risk Management 56% 26% 15% 3% Machine Learning Automation DLT Quantum
  • 13. RegObs Special Report 07 December 2022 ◆ 13 Yet, there are still considerable differences between individual banks Although there is little doubt most banks see cost efficiency savings and AI (automation and machine learning) as the foremost purpose/technology where they are investing this budget, there is still a fair bit of heterogeneity across the different banks. For example, Bank 2, 3, 12 and 16 are pretty much all bang on average in terms of purpose of their investments. Contrast this with Bank 13, 9, 4 or 14, for example, where customer experience, cost efficiency, fraud/operational risk or risk management are much more important features than for the rest of our sample. Picture 2: Investment split (purpose, %) - distribution from the mean (=0) for each bank in the sample Source: Mediobanca Securities In terms of the split across the different technologies Bank 2 and Bank 3 are again pretty much in-line with the consensus. Bank 4 meanwhile has put a much higher weighting on Quantum investments than for the rest of the sample and bank 14 is investing primarily in automation. Picture 3: Investment split (technology, %) – distribution from the mean (=0) for each bank in the sample Source: Mediobanca Securities FinTechs seen as partners. New role for banks post-adoption of these technologies Our survey suggests banks are not scared about FinTechs sending them to extinction. As we transition to this new banking landscape, banks were overwhelmingly positive about the role of FinTechs in that journey. Nearly 90% view them as partners, as opposed to either rivals or just temporary players. Once widespread adoption of the technologies is realised, banks unsurprisingly see better cost efficiency as a major outcome (in-line with what they had said previously about the main goal being to cut costs). Perhaps more interesting, is that even more of our sample of banks recognise that these technologies are likely to fundamentally change the business model and the overarching role banks play within the -2 -1 0 1 2 3 Bank 1 Bank 2 Bank 3 Bank 4 Bank 5 Bank 6 Bank 7 Bank 8 Bank 9 Bank 10 Bank 11 Bank 12 Bank 13 Bank 14 Bank 15 Bank 16 Distance from mean (Z-Score) Cost efficiency Revenue Boost Customer Experience Fraud, Operational Risk Other Risk Management -2 -1 0 1 2 3 4 Bank 1 Bank 2 Bank 3 Bank 4 Bank 5 Bank 6 Bank 7 Bank 8 Bank 9 Bank 10 Bank 11 Bank 12 Bank 13 Bank 14 Bank 15 Bank 16 Distance from mean (Z-Score) Automation Quantum DLT Machine Learning
  • 14. RegObs Special Report 07 December 2022 ◆ 14 industry. It goes beyond just cutting costs, or better risk management. Fears over the possibility of revenue erosion are further down the pecking order. Figure 3: View on FinTechs Source: Mediobanca Securities Figure 4: Largest impact of these technologies for banks Source: Mediobanca Securities Risk management, digital channels & back office most susceptible to tech disruption Finally, we asked our recipients what part of the bank they would see as most impacted by the adoption of our three technologies. There were three areas that were clearly identified as the most susceptible: risk management & control; digital channels; and back office and IT. With each being included in the “Top 3” by 11 of the sample of 18 banks that responded to this question. Picture 4: Part of the bank most impacted by technological disruption Source: Mediobanca Securities; *18 banks (rather than 16) responded to this question. 89% 5% 6% Partner Temporary Player Contender 50% 39% 11% 0% 0% 0% 10% 20% 30% 40% 50% 60% Role of banks in the financial industry Cut costs Improve fraud/operational risk management Erode revenues Improve other aspects of risk management 11x 11x 11x 7x 6x 3x 2x 2x 0x 0x 0x 2x 4x 6x 8x 10x 12x Risk management and control Digital channels Back office and IT Payments and transactions Securities and capital markets Credit granting Physical network Financial advisory Deposit gathering Treasury No. of respondent banks that included within their "Top 3"
  • 15. RegObs Special Report 07 December 2022 ◆ 15 …seconded by our “use case” almanac The second survey we sent out was to map the “use-cases” banks have been working on as anecdotal evidence of their interest and successes. We give full details on the individual cases here (where we have permission to disclose). We aggregate the use cases in a fictional “Banca Case”, composed of 26 cases across 14 respondent banks. The majority (c.70%) of Banca Case relates to MLAI (mostly “Chatbot”/Virtual Assistants), ¼ in DLT and just a couple in QC. In general, they were built for internal use rather than to be used alongside other peers. Most cases were developed in Partnership with Fintechs/technology providers, although internal resources were also (at least partially) employed in ½ of cases: usually around 4 FTEs. The journey from inception to production was short (279 days), and on average each use case had now been in active operation for a similar time. Consistent with Banca Benchmark, Banca Case’s goal tended to be cost efficiency where, for a budget of around €1.3m, a return of €1.9m was expected. Digital Channels and Back Office & IT were the departments most impacted, followed by Securities & Capital Markets and Risk Management & Control, with far fewer use-cases related to Payments than indicated in the Benchmark survey. Use Case Survey Our use case survey was filled out by 14 respondent banks, which gave us a total of 26 existing use cases for the three investigated technologies. It was designed to be less restrictive than the benchmark survey, with more freedom for respondents to go into greater detail on one (or more) use cases. Yet, it is still possible to make some general observations about the types of use cases that our banks have implemented. Banca Case: our aggregate benchmark We aggregate the responses from the use cases survey into a fictional Banca Case, representing EU banks. This is strong of €205bn market capitalisation and showcasing 26 use cases. 2/3 of these apply MLAI, ¼ DLT and c.1/20 QC, with an overwhelming cost cutting scope mainly for digital channels and Backoffice. ¾ of use cases have involved a partner, with 50% internal development and 50% outsourcing. c.2/3 of cases involve standalone adoption and 1/3 adopt a cooperative approach. The 26 cases have had an average time-to-production of 280 days, €1.3m investment and €1.9m return. Picture 5: Use cases - snapshot of the findings from the use cases sent back by banks Source: Mediobanca Securities Use-cases dominated by MLAI applications, but many interesting DLT applications also in pipeline… Given the overwhelming majority under our benchmark survey stated they were investing in AI technologies, it is of little surprise that most use cases were also based on this technology (70% of total use cases). Although, it was also clear that there was a bias towards classifying the use-case as Technology split: AI/ML (66%); DLT (28%); Quantum (6%) Partnership: Partner (77%); No partner (23%) Development: Internal (50%); Outsourced (50%) Main user(s): Stand-alone adoption (65%); Co-operativ e (35%) Time to production: c.280 days Budget: €1.3mn Return: €1.9mn Expected ROI: 40-50% Main purpose: Cost efficiency Main part of bank: Digital Channels Back Office/IT Use Cases
  • 16. RegObs Special Report 07 December 2022 ◆ 16 machine learning, even when it sounded much more aligned with the technical definition of automation. Within AI, the largest proportion of focus areas were for some form of ChatBot application (CallBot, Virtual Assistant, VoiceBot) with a number of banks providing use cases in this area. DLT-based applications made up around a quarter of the overall use-cases. Here we saw quite varied applications, with very little cross-over between banks. As we illustrate below, this includes post- trade services, Bitcoin trading/custody and digital surety amongst others. There weren’t many use cases for Quantum technology, with this making up just 6% of the overall use- case examples we received. As we stated previously, the technology is still very much in its infancy with the focus more on partnering with FinTechs and BigTech organisations in advance of the eventual Quantum revolution. What has been done so far is targeted at more efficient, quicker and more comprehensive dynamic portfolio optimisation. Picture 6: Area of focus for use-cases, split by type of technology employed Source: Mediobanca Securities …whilst typically built with a partner (sometimes supported by internal staff) and designed for standalone use by developing the firm rather than in co-operation with peers In 3/4 of our sample (77%), these use cases were built either entirely or partially with the support of a partner, with around half of the use cases completely outsourced to external organisations. Where the use case was built internally, it was implemented on average by the equivalent of 4 FTEs. Finally, we highlight that whilst the majority were built to be used solely by the respondent bank, about 1/3 were designed to be used co-operatively with other organisations. Picture 7: Use cases mostly built with a partner and designed for stand-alone use Source: Mediobanca Securities 280-day time to production On average, it took each use case around 280 days to move to the production phase. This works out to a little over a year given c.250 trading days in a year. As we demonstrate below, it consists of around 3-4 months (based on 20 trading days in a month) for each of the initial exploratory analysis and proof- of-concept phases. The development phase then takes a little longer at around 4-5 months (although likely slightly exaggerated by a couple banks consolidating the analysis and POC phases into the development phase). The pre-production phase then takes a further 3 months, on average. Automation Machine Learning DLT Quantum Back-office processing Foreclosure Post-trade serv ices Dynamic portfolio optimisation Employee Q&A MIFID compliance Bitcoin trading/custody Asset allocation Loan decision making Chatbot Trav el insurance Financial health analysis Callbot Digital surety Structured product issuance DCM book building Cash-for-collateral lending Legal email automation Salary-backed loans Checking reciprocal accounts Record of carbon credits
  • 17. RegObs Special Report 07 December 2022 ◆ 17 At the time of receiving survey responses, use cases had been operational for a little over a year on average. As such, we would point out that whilst most of the use cases we have been provided are fairly newly implemented, banks are very likely also working on other projects using these technologies that are in pre-production phase and not disclosed, yet. Picture 8: Breakdown of days for each stage of production Source: Mediobanca Securities; *POC = Proof of Concept Cost efficiency the dominant purpose of use-cases, customer experience also important In our use case survey, we also asked respondents to tick which of the below five options were a targeted benefit of the use case. They could tick as many as deemed pertinent. Once again, we saw cost efficiency come out on top, where around two-thirds of use cases flagged this as a purpose. Customer experience was also considered an important feature in the use cases, with half mentioning this as a purpose. Revenue boost and other risk management were both ticked in nearly a third of the total use cases. Fraud and operational risk management was relevant to c.20% of all use cases. Picture 9: Percentage of use cases that ticked this purpose as one of their options Source: Mediobanca Securities Estimated return of around 40-50%, but some doubts on consistency of data inputs One aspect of the survey where we had initial doubts about the quality of responses was with regards to asking for absolute numbers on the budget and projected return of the projects. The outcome of our survey was that on average, around €1.3mn was budgeted for each of the use cases. Meanwhile, the anticipated return was on the order of €1.9mn. The way these numbers were reported makes identifying whether they were on an annualised or overall basis ambiguous. Assuming simply that both 65% 50% 31% 31% 19% 0% 10% 20% 30% 40% 50% 60% 70% Cost efficiency Customer experience Revenue boost Other risk mgmt. Fraud/Op. Risk % of total use cases that ticked this as an option
  • 18. RegObs Special Report 07 December 2022 ◆ 18 were provided on an overall basis, we get a basic implied return of 40-50%. However, we would suggest taking this with a big “pinch of salt”. Digital channels & back office again most important, but shake-up in rest of pecking order… Lastly, we asked banks what part of the bank they thought would be most revolutionised by their use case. Given that we had previously asked the same question to banks in relation to the three technologies during our benchmark survey, we found it interesting to compare the two. Digital Channels and Back Office & IT were consonant with what we had found with our benchmark, where both were considered the most transformed (10 use-cases out of 26 mentioning them). Interestingly, Risk Management & Control was significantly less relevant to our use cases vis-à-vis with the benchmark. Where Securities & Capital Markets, in particular, took its place for the use-cases. As a final point here, we stress the more marginal proportion of banks referring to Payments & Transactions in their Top 3 for the use-case survey. We’d suggest that a logical possibility here is that Risk Management & Control or Payments innovations are further away from the production stage for most banks yet are both areas where they are investing considerable portions of the overall transformation budget. Picture 10: Part of the bank most revolutionised by each of the use cases (versus benchmark survey) Source: Mediobanca Securities; *Two axis are not comparable: lbs is out of possible 26x and rhs is out of possible 18x. 0x 2x 4x 6x 8x 10x 12x 0x 2x 4x 6x 8x 10x 12x Digital channels Back office and IT Securities and capital markets Risk management and control Financial advisory Credit granting Physical network Payments and transactions Treasury Deposit gathering No. banks that included within their "Top 3" No. use cases that included within their "Top 3" No. of use cases that included within their "Top 3" (lhs) No. of banks in benchmark that included within their "Top 3" (rhs)
  • 19. RegObs Special Report 07 December 2022 ◆ 19 Good effort, but far from transformative innovation yet Our two surveys show coherent results, broad enough to depict the current location of banks along their innovation journey. While banks are trying hard to innovate, most of their efforts are going in a linear evolution of the current operating model, with heavy investment in chatbots as the “new automation”, not quite a technological revolution yet, in our view. Short inception-to- production timing, the high usage of external staff and the focus on efficiency-oriented projects can both reflect good project management and cost discipline and room to increase ambition along the innovation vector. There is something to be said here about the current career incentives for innovation leaders: showing short term results plays against truly transforming multi-year projects. Governance likely also plays a role in the low penetration of DLT, in our view, reflecting the low propensity to share and coordinate with peers/competitors and the need to control the project/investment. The absence of payments reflects the exit of this lucrative scale business by banks, potentially a strategic mistake in the long term and a drawback for the overall innovation know-how, in our view. Finally, it is just too early for QC to take room, but it will come. The overarching conclusion is that banks know they must change, but they are doing so too slowly and linearly, in our view. Coherence between Mediobanca’ s surveys Our two surveys provided coherent results, comforting us on the overall picture these describe and hence allowing us to draw some preliminary conclusions on where European banks are on their innovation journey now. Trying hard… We see banks are trying hard to innovate, with a multitude of different initiatives, dedicated structures, more and more experiments and a growing attention to the innovation world. …good inception to production timing… The surveys confirm banks have short inception-to-production timing, showing project management efficiency, but also, possibly, suggesting they could increase their ambition and aim for more complex targets. …but not daring enough, yet 70% focus on chatbot development suggests to us that the current focus is on moving further on efficiency and the evolution of the automation process of the past decades. More ambitious projects, potentially rethinking the current business model and production processes, only represent the fringe of innovation investments, at best, for now. Low DLT penetration: cooperation is the real hurdle We find a particularly low penetration of DLT initiatives in the innovation pipeline. In our view, the main reason for this is intrinsic to DLT, i.e. the necessity to open and cooperate with competitors. We find a focus on projects which can be fully controlled and owned by the individual banks. This approach is a hurdle to the development of effective DLT-based initiatives, which instead require the creation of shared networks with win-win characteristics for networks participants. More internalisation to own and develop Banks seem to rely heavily on external providers/consultants for their innovation projects. The more the internalisation of the staff involved, the more the ownership and the ability to give birth to concatenated developments which can transform products/processes over time. A cultural shift based on internal staff is needed for change to take root and flourish.
  • 20. RegObs Special Report 07 December 2022 ◆ 20 Efficiency and digital channel/IT focus rather than rethinking products/processes Our surveys highlighted the banks’ focus on efficiency and on digital channels/IT. These projects tend to be easier to justify and provide a faster payback time, to the advantage of the sponsors who need to show results. Yet, transformations take time and are hardly plausible through small, cost-cutting- focused projects. The issue with this is that multi-year programs tend to work against career progress. More innovative proposals require a change in the incentive structure of innovation leaders, we would argue. Payments are out of scope Payments seem to fall out of the scope of banks. More and more banks have abandoned this very large and lucrative space, to the advantage of specialised payment companies. Yet, payments remain a core product provided by banks and one which is at the forefront of the digital revolution. It is just too early for Quantum Computing… but it will come If we had to rank the technologies by stage of implementation/maturity, we would place MLAI up top, DLT runner-up and QC in third place, at a distance. The supreme computational power this technology could provide would come in handy for the transition of banks into data companies, but we are just not there yet, and we understand why banks place it last in the priority list. It will come, but it is just too early now. Pass mark for the effort, but much more is required to drive a mutation in the species Our surveys show banks are experimenting, they are active, they are getting their hands dirty to learn. Yet, the horizon is still too narrow and the ambition too low for us to see the intention to radically change course, mutate, change skin. We are still in a slow, risk averse and mistake averse attitude. The perception of threat is likely still very low, and the survival instinct is yet to kick in for banks to understand faster, more radical change (a mutation) is due for the perpetuation of the species. A vision of how to get more ambitions in the coming chapters We will argue what the mutation could be and how what the path to survival is over the next chapters of the note.
  • 21. RegObs Special Report 07 December 2022 ◆ 21 “breaking the bank”: isolating the species at risk New technologies are likely to disrupt businesses/products which are highly standardised. This chapter executes a surgical procedure - never attempted before – to separate the standardised businesses from the customised businesses of the twelve most representative European banks. This predicates on the assumption that standardised businesses are more porous to re-engineering on new technology than customised businesses. We identify 2/3 of operating income, 60% of costs and 50% of loans belong to standardised banking. New technologies (such as DLT, AI and quantum computing) are usually seen as a threat to the status- quo, as a rapid adoption in various sectors of the economy could have the potential to change current rules of the game. However, this would not necessarily mean that they are threats for current players if these are quick enough in adapting and improving their own business models. In this context, new technologies can be seen as (inevitable) enablers of future growth. Banks are not exempt from this, and emerging technologies could change how things are currently made. Clear examples could be:  Industrialise/streamline credit granting - A common DLT platform could facilitate, accelerate and optimize the process to grant credit to customers. Real estate transactions involve a large number of legal, financial and real-estate intermediaries acting for the buyer, seller and lender, each adding fees and time to the transaction. A shared/permissioned blockchain for mortgage loans could be a powerful application for DLT technology.  Revolutionise the liability structure - The adoption of a specific type of CBDC could change banks’ funding structure (see our report on Digital Euro);  Democratise the exchange of securities - The tokenization of financial assets could change the way they are distributed to clients and how they are managed by banks (i.e. banks could need less FTE to manage the same products/clients through the implementation of smart contracts). They could theoretically even disintermediate banks in full via direct P2P transactions or perhaps leaving banks as custodians.  Take efficiency to another planet - Could further increase the efficiency of banks by reducing costs and increasing volumes. Standardisation vs customisation Yet, the transformation would require banks to invest massively in new technologies to adjust their business model for all these potential changes. Moreover, we believe that the more a financial product is standardised, the more it is exposed to technological change, while customised products, i.e. tailor- made products with a large component of human advisory/interaction, are less exposed to technological disruption.
  • 22. RegObs Special Report 07 December 2022 ◆ 22 Picture 11: banks’ product matrix based on high/low impact from new technologies Source: Mediobanca Securities In the following section, we try to identify which businesses are more exposed to this, and which could be the impacts to volumes, margins, and costs. Standardised products account for 2/3 of operating income Creating the European megabank to dissect its business model We dissected the business models of the 12 most representative banks in Europe, which together represent >€480bn in market cap or c€15trn in total assets, across eight European markets: France, Spain, Italy, UK, Germany, Switzerland, Nordics & Benelux. In other words, our sample is composed of the major European investment, commercial and diversified banks. From here, we generally refer to this as the “aggregate”. Net interest income & fees were split into “standardised” and “customised” buckets… Our main goal is to understand which business of our banks’ sample can be challenged, improved, or even substituted by new emerging technologies and players. In the last decades, banks have continued to grow and diversify their business models by adding new sources of revenues and shrinking their cost base to increase profitability. They have expanded in asset management, insurance and strengthened their services to corporates by offering more sophisticated and tailor-made products. We believe new emerging technologies could represent a threat for some core businesses of banks. To identify which business is more exposed to this trend, we split the banks’ main sources of revenues (NII and fees) in two categories: standardised and customised businesses as a proxy of exposure to technological disruption: i) standardised, i.e. businesses based on a standardised model across the banking sector, in which there is little customisation of the services provided to clients (clear example are mortgages, consumer loans, payment and FX services, mass market of asset management and insurance products); ii) customised, i.e. all businesses where services are generated on an ad hoc basis, tailored on client needs (clear example are corporate and SME loans, private banking and wealth management services, corporate finance and advisory services). For the NII decomposition we used the data from the H121 EBA transparency exercise. By taking the geographical exposures of each bank for each category (i.e. mortgages, consumer, corporate and SME loans), and applying the average interest rates of each country for each exposure, we estimate a Mortgages Consumer loans Payments FX transactions Brokerage Underwriting fees Distribution of insurance products Distribution of saving products Corporate and SME loans IB services Private banking fees Wealth management fees Insurance products Standardised Customised Advisory services Potential impact from new technologies Low High
  • 23. RegObs Special Report 07 December 2022 ◆ 23 theoretical interest income generated by standardised and customised businesses. Then we applied these proportions to the 4Y average interest income from loans of each bank in our sample. All other sources of income in the NII have been equally split among the four categories. For the funding side, we assume that mortgage exposures are financed by covered bonds (taking the average yield of each country’s covered bond curve), consumer loans by securitizations or senior non- preferred securities, while the cost of funding for corporate and SME loans is based on the difference between total interest expenses and the cost of funding to finance mortgages and consumer loans. While the loan book is evenly split, NII is skewed more towards standardised products We classified mortgages and other personal loans as “standardised” products, whereas corporate and SME lending is considered “customised”, needing a more personalised approach making it more difficult to industrialise. Based on this classification, the loan book of our aggregate bank was pretty much split 50/50. Yet, NII was much more skewed towards the standardised products (c.2/3) versus 1/3 from customised businesses. Picture 12: Composition of aggregated loan book (standardized vs customized, %) Source: Mediobanca Securities Picture 13: Composition of net interest income (standardized vs customized, %) Source: Mediobanca Securities Fee income also tilted more towards standardisation (c.60%) As for fees, we have assumed that asset management fees, insurance distribution, custody, brokerage, payments, underwriting, and FX fees are all standardised revenues. Meanwhile, any fees related to wealth management, corporate finance, advisory and in-house insurance factories are customised revenue streams. On this basis, we found that roughly 60% of banking fees relate to standardised products/processes, with the other 40% being customised. Table 1: Revenues decomposition: standardised vs customised Revenue decomposition (€bn) % NII [standardised] 94 66% NII [customised] 48 34% NII [total] 142 100% Fees [standardised] 57 59% Fees [customised] 40 41% Fees [total] 97 100% Revenues [standardised] 151 63% Revenues [customised] 88 37% Revenues [total] 240 100% Source: Mediobanca Securities Mortgages 38% Consumer credit 10% Corporate 39% SME 13% Mortgages 36% Consumer credit 30% Corporate 22% SME 12%
  • 24. RegObs Special Report 07 December 2022 ◆ 24 …with associated costs backed out from the C/I ratio of specialised firms/subsidiaries Now that revenues have been split between standardised and customised pools, we needed a way to identify the associated costs. To do this, we constructed a representative sample of either highly specialised firms/pure players or business units that were focused on each of the various revenue sources. For instance, for Asset Management fees we took Amundi, DWS, UBS Asset Management, Schroders and HSBC Asset Management. Where we then took the average cost-income ratio to estimate the costs associated with each of the revenue streams, giving us a way to obtain the standardised and customised costs (and ultimately how the pre-provision profit was split across standardised and customised). Suggestive of a roughly 60%/40% split on costs between standardised/customised Our assumed cost-income ratio ended up at around 53% for both standardised and customised products. Given most revenues came from standardised sources, naturally we also saw this was the main driver for costs at around 60% of the total. Picture 14: Assumed cost-income ratios per activity (derived from samples of pure plays) Source: Mediobanca Securities, Benchmark companies: SAN UK – Retail, LLOY – Retail, ABN – Retail, NWG – Retail, SAN US, BARC CCP, ACA Consumer Credit, BKT CB, ISP CB, DBK CB, ABN CB, LLOY CB, AMUN, DWS, UBS AM, Schroders AM, HSBC AM, ACA Insurance, BNP Insurance, ISP Insurance, CABK Insurance, KBC Insurance, HSBC Insurance, BNP Security Services, ACA Asset Servicing, BNY Mellon Securities Services, NEXI, ACA Financing Activities, BNP GM, GLE GMIS, GS GM, ACA Wealth, ABN PB, BNP Global Banking, Rothschild, ACA Leasing & Factoring. c.2/3 of operating income is coming from standardised streams Putting our P&L back together, we compute that 65% of operating income relates to standardised products and services, whilst only around 35% are customised. Across our sample we also see a fair bit of deviation, ranging from as much as 85% of operating profits coming from standardised, down to 43%. 81% 69% 61% 49% 48% 44% 43% 29% 80% 61% 58% 57% 55% 54% 51% 51% 44% 23% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% Custody/Trust Markets Asset Management Consumer banks Mortgage banks Underwriting Payments/FX Insurance Wealth Management Other Asset Management Commercial banks Corp. Fin/ Advisory Factoring Other Guarantees/ Commitments Loans Corporate banks Insurance/Other brokerage
  • 25. RegObs Special Report 07 December 2022 ◆ 25 Picture 15: Operating profit split into standardized and customized buckets for our sample Source: Mediobanca Securities Table 2: Aggregate bank PBT decomposition - standardised vs customised (€bn) Simulation % NII [standardised] 94 66% NII [customised] 48 34% NII [total] 142 100% Fees [standardised] 57 59% Fees [customised] 40 41% Fees [total] 97 100% Revenues [standardised] 151 63% Revenues [customised] 88 37% Revenues [total] 240 100% Expenses [standardised] -74 61% Expenses [customised] -47 39% Expenses [total] -121 100% Pre-provision profit [standardised] 77 65% Pre-provision profit [customised] 42 35% Pre-provision profit [total] 119 100% Source: Mediobanca Securities, company data, EBA 85% 80% 77% 71% 65% 64% 63% 58% 49% 49% 46% 43% 65% 15% 20% 23% 29% 35% 36% 37% 42% 51% 51% 54% 57% 35% 0% 20% 40% 60% 80% 100% Bank 6 Bank 1 Bank 12 Bank 11 Bank 5 Bank 4 Bank 7 Bank 10 Bank 3 Bank 9 Bank 2 Bank 8 Aggregate Standardised Customised
  • 26. RegObs Special Report 07 December 2022 ◆ 26 Digital disruption carries 30-50% margin compression… We look at prior sectors disrupted by technological transformation such as the film rental, photography, local taxi and retail bookstores sectors. Anecdotal evidence points towards 30-50% margin compression. You can read our in-depth case studies here. Netflix, Amazon, Uber: 30-50% margin compression from digital disruption To understand how margins on standardised products might be squeezed, we look to other sectors that have undergone sweeping technological change for inspiration. Across the four case studies we looked at, we estimate roughly 30-50% margin compression. At the top of the range is the c.50% compression in the movie rental sector where digital streaming services (think Netflix/Amazon Prime) crushed the high margins enjoyed by brick & mortar incumbents (Blockbuster). Whilst the shift from physical bookstores (Barnes & Noble) to the digitalised Amazon model where books were instead purchased online and delivered via a sprawling distribution network saw a compression of c.30%. We also looked at the advent of Uber for the local public transport sector and the launch of digital cameras for Kodak. Our detailed work on case studies can be found here. Picture 16: Margin transformation of photography sector Source: Mediobanca Securities, Factset Picture 17: Margin transformation of movie rental sector Source: Mediobanca Securities, Factset Picture 18: Margin transformation of taxi industry Source: Mediobanca Securities, Factset, Gov.UK (link) – Data on taxis/PHVs given on bi-annual basis pre-2017 so have interpolated interceding years. Picture 19: Margin transformation of book retail sector Source: Mediobanca Securities, Factset 0% 10% 20% 30% 40% 50% 60% 70% 80% 1980 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Kodak Fuji Film First commercially marketed digital camera in 1990. 20% 25% 30% 35% 40% 45% 50% 55% 60% 65% 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 Blockbuster Netflix 40 60 80 100 120 140 160 20% 25% 30% 35% 40% 45% 50% 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 Uber Founded Taxis/PHVs [London, rhs] Addison Lee 15% 20% 25% 30% 35% 40% 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Barnes & Noble Amazon
  • 27. RegObs Special Report 07 December 2022 ◆ 27 …or standardised revenues -30%; costs -60% to offset We simulate three possible scenarios for the tech-driven disruption of the standardised banking sector. These are informed by the degree of margin compression we have witnessed during the digital transformation of several case studies. Our base case uses the mid-point of 40% margin compression to NII and fees, resulting in roughly -30% standardised revenue decline (not all fees – insurance, AM, custody, placement and underwriting – suffer the same margin erosion with e.g. payments/FX taking the full haircut) that would require a c.60% decrease in standardised costs to be fully offset. This would imply that standardised banking would have to collapse the cost-income ratio from c50% today to <30%, something even the most efficient of the famously lean Nordic banks do not reach. Our optimistic and pessimistic scenarios instead suggest -22% and -36% of revenue compression, necessitating c.-45% and c.-75% cost contraction, respectively, to maintain profitability unchanged. 20-40% revenue attrition from digital disruption… Picture 20 shows the revenue attrition of the standardised business of the twelve banks in our sample under the base, optimistic and pessimistic scenarios from digital disruption. On aggregate, this indicates c.20-40% revenue compression of standardised revenues. Picture 20: Summary of standardised revenue compression under pessimistic, base & optimistic scenarios (% revenue) Source: Mediobanca Securities …calling for c.45-75% cost cuts to keep profitability of standardised banking unchanged Picture 21 shows the equivalent cost cuts of the standardised business of the twelve banks in our sample required to keep profitability unchanged under the base, optimistic and pessimistic scenarios of revenue attrition from digital disruption. On aggregate, this indicates c.45-75% cost cuts, a magnitude so great to require a rethink of the cost structure rather than linear cuts to the current business model, in our view. Picture 21: cost cuts to offset standardised revenue compression under pessimistic, base & optimistic scenarios, % Source: Mediobanca Securities -43% -41% -40% -40% -37% -35% -34% -32% -30% -29% -28% -25% -36% -35% -34% -33% -32% -29% -28% -28% -26% -24% -24% -22% -21% -29% -27% -27% -25% -24% -22% -22% -21% -20% -19% -18% -17% -16% -22% -50% -45% -40% -35% -30% -25% -20% -15% -10% -5% 0% Bank 6 Bank 8 Bank 12 Bank 1 Bank 7 Bank 9 Bank 5 Bank 3 Bank 4 Bank 11 Bank 10 Bank 2 Aggregate Pessimistic Base Optimistic -85% -73% -73% -69% -69% -62% -52% -48% -45% -41% -39% -25% -73% -60% -53% -52% -51% -49% -46% -39% -36% -34% -31% -29% -19% -59% -41% -37% -36% -35% -34% -32% -28% -26% -24% -22% -20% -14% -45% -90% -80% -70% -60% -50% -40% -30% -20% -10% 0% Bank 6 Bank 12 Bank 7 Bank 8 Bank 1 Bank 9 Bank 5 Bank 11 Bank 4 Bank 10 Bank 3 Bank 2 Aggregate Pessimistic Base Optimistic
  • 28. RegObs Special Report 07 December 2022 ◆ 28 Scenario analysis: From the Amazon to Netflix experience In this chapter we illustrate in more detail our methodology and the results of our scenario analysis (base, optimistic and pessimistic) and its impacts on banks’ standardised business. We trim NII by haircutting the credit spread on standardised loan products For NII we isolated how much of the net interest margin is represented by the risk-free rate and how much is coming from credit spread. To compute the overall risk-free rate for each of the companies in our sample we first took the average duration for mortgages, consumer, corporate and SMEs for all the relevant countries. Allowing us to get the relevant risk-free rate for the different types of loans in each country. Again, using the EBA database, we were able to calculate the implied Group level risk-free rates based on their how exposures were split between the different types of loans and geographies. The credit spread was then simply assumed to be the delta between this risk-free rate and the net interest margin. For our 12-bank sample the overall credit spread was found to be c.200bps with consumer credit offering the widest margin (c.700bps) and mortgages the lowest (c.135bps). Under each of the different scenarios we then hit the credit spread for the standardised products (i.e. mortgages/consumer) based on what we had learnt from our historic case study examples: • Optimistic scenario (-30%) = the Amazon experience • Base scenario (-40%) = the Fuji experience • Pessimistic scenario (-50%) = the Netflix experience Not all fees are equal: standardisation score given to determine margin pressure Our approach for fees was a little different given the breadth of different fee-earning activities conducted by our banks. As such, we tried to adjust the magnitude of the margin cuts in each scenario to a sliding scale. In this instance, we chose a scale from 0 to 5, where 0 represented a customised product such as Wealth Management. We assigned this to each of our “customised” fee sources. At the other end of the spectrum we had completely standardised products or services such as FX fees. In terms of margin compression for each, we assumed that the most standardised products would see compression akin to the top of the range from our case study analysis (i.e. -50%), whilst customised products would be untouched. We then allowed for a 10% step between each of the sub-levels and assumed that the optimistic scenario would be 5% less compression and the pessimistic 5% more compression than this base case. Picture 22: MB assumed sliding scale from customized to standardized sources of fee income Source: Mediobanca Securities
  • 29. RegObs Special Report 07 December 2022 ◆ 29 We summarise both the NII and fee compression in the below table… Picture 23: Margin compression under our different scenarios Source: Mediobanca Securities #1: Base Case – The “Fuji Experience” Standardised revenues decline roughly -29%, 2/3 explained by NII compression We begin by looking at our base case. For this we use the mid-point of our 30-50% margin compression range under the case study examples we gave earlier. A scenario which most closely follows what happened to Fuji film in the photography industry during the shift away from film. Hence, we coin this the Fuji scenario. Here, we find that on aggregate there is a -29% revenue when applying our margin compression assumptions. This is split into roughly -19% coming from standardised NII and a further - 10% from standardised fee income. However, there is also a wide disparity around this mean across the individual banks in our sample (standard deviation = 4%). At the extremes we see Bank 6 facing nearly -35% revenue decline, with almost all of that coming from the NII side. The opposite is true for Bank 2, where the revenue impact is more inconsequential at -21% and most of that is from lower fees. Picture 24: Base Case – Impact on NII & fee income (% total revenues) Source: Mediobanca Securities Mortgages and consumer both equal drivers of the NII compression Going into a little more detail on what is driving the changes in standardised revenues, we split out the two standardised elements (mortgages and consumer). The picture here is very mixed at the Current Optimistic Base Pessimistic 100% 70% 60% 50% 0 100% 100% 100% 100% 1 100% 95% 90% 85% 2 100% 85% 80% 75% 3 100% 75% 70% 65% 4 100% 65% 60% 55% 5 100% 55% 50% 45% Margin Cut (% of initial revenue) Fee rank Net Interest Income Standardised Customised -35% -34% -33% -32% -29% -28% -28% -26% -24% -24% -22% -21% -29% -40% -35% -30% -25% -20% -15% -10% -5% 0% Bank 6 Bank 8 Bank 12 Bank 1 Bank 7 Bank 9 Bank 5 Bank 3 Bank 4 Bank 11 Bank 10 Bank 2 Aggregate NII Fees
  • 30. RegObs Special Report 07 December 2022 ◆ 30 individual bank level, which sort of balances out at the aggregate level to be roughly equal contributions from mortgages and consumer loans. This is because while mortgages tend to make up more of the overall loan book, credit spreads on consumer loans are much wider than for mortgages. As such, the two effects more or less offset each other on aggregate. Picture 25: Base Case – standardised NII impact split by product (% total revenues) Source: Mediobanca Securities Highly standardised ‘payments’ element dominates the story for fee compression On the fee side, we also illustrate the contribution from the various fee buckets towards the total. We do caution here that the way that different banks disclose their fee breakdown is not entirely consonant with one another. Clearly, the main driver here is payments which makes up a hefty chunk of the total (half of the -10%), although this is likely to also be picking up some of the FX fees as well for example. This reflects the fact that payments are a major contributor to the fee income line and at least in our view, are a highly standardised product. Picture 26: Base Case – Fee impact split by product/service (% total revenues) Source: Mediobanca Securities -28% -28% -22% -20% -19% -17% -16% -16% -14% -13% -13% -4% -19% -30% -25% -20% -15% -10% -5% 0% Bank 6 Bank 1 Bank 7 Bank 12 Bank 8 Bank 11 Bank 5 Bank 10 Bank 3 Bank 4 Bank 9 Bank 2 Aggregate Mortgages Consumer -16% -15% -15% -13% -12% -11% -11% -7% -7% -7% -6% -4% -10% -18% -16% -14% -12% -10% -8% -6% -4% -2% 0% Bank 2 Bank 9 Bank 8 Bank 12 Bank 3 Bank 5 Bank 4 Bank 7 Bank 10 Bank 11 Bank 6 Bank 1 Aggregate Custody/Trust Underwriting/ Placement Insurance FX Fees Asset Management Securities brokerage Payments
  • 31. RegObs Special Report 07 December 2022 ◆ 31 2/3 reduction in costs required to hold profitability steady… New technologies should lead to some cost synergies as banks would be more efficient in offering the same product. Yet, it is hard at this stage to estimate the amount of cost synergies new technologies could free up. In our base case scenario, we estimate that core standardised revenues could decrease by c30%, meaning that if costs remain the same, the operating income could decrease by c60% on aggregate level. Hence, we estimate that cost should decrease by c60% to offset the margin pressure and leaving unchanged the aggregated profitability of our aggregate standardised bank. Picture 27: Base Case – potential impact to the standardized operating income Source: Mediobanca Securities …implying <30% C/I ratio post cuts in the standardised bank… We show what the cost cuts required to neutralise revenue erosion would do to the cost-income ratio. On aggregate, the C/I for the standardized business would need to drop from c.50% to <30%. This does vary quite a bit across our sample of banks though. At the top of the range, Bank 6 would be required to slash costs by c.74% to get to a C/I ratio of around 12%. With this being one fourth of the cost- income of the most efficient banks in Europe today, getting here will be far from easy (and realistically require a lot of work also on the revenue side). Bank 2 on the other hand would only require costs to come down around 36%, i.e. some are genetically fitter for survival than others. Picture 28: Base Case – cost cutting required to offset margin compression Source: Mediobanca Securities -19% -10% -29% -57% -59% 0 20,000 40,000 60,000 80,000 100,000 120,000 140,000 160,000 180,000 200,000 Revenues NII margin pressure Fee margin pressure Revenues post margin pressure Costs New Operating income Cost cutting Initial operating income % change in revenues % change in operating income % fall in costs required to offset revenue margin compression -60% -53% -52% -51% -49% -46% -39% -36% -34% -31% -29% -19% -42% -80% -60% -40% -20% 0% 20% 40% 60% Bank 6 Bank 12 Bank 7 Bank 8 Bank 1 Bank 9 Bank 5 Bank 11 Bank 4 Bank 10 Bank 3 Bank 2 Aggregate C/I (pre) C/I (post) Change in C/I (%) C/I ratio (%) C/I ratio (%) Change in C/I (%)
  • 32. RegObs Special Report 07 December 2022 ◆ 32 …implying <40% C/I ratio of standardised+customised banking, at Nordic level Finally, considering that banks are a mix of standardised and customised businesses, usually sharing costs among businesses, we illustrate the impact of our analysis to the aggregate banks’ sample taking into consideration both standardised and customised businesses. Total combined revenues would decline by 18% (assuming no revenue erosion in the customised bank), while costs should drop by 36% to perpetuate the initial profitability. This would imply aggregate C/I ratio dropping to <40% from c.50%, for reference, in line with best in class Nordic banks. Picture 29: Base Case – potential impact to the aggregate operating income (standardized + custumised business) Source: Mediobanca Securities #2: Bookshop scenario - Optimistic case Standardised revenues decline by 22% If the banking sector faced margin compression more akin in magnitude to that of book retailers post- Amazon (i.e. 30% margin compression), revenue compression would be more benign at -22% (vs -29% in our base case). In this more optimistic scenario, the range of revenue compression corridor for individual banks would narrow to -27% to -16%. Picture 30: Optimistic Case – Impact on standardized revenues Source: Mediobanca Securities ​ -37% -18% -36% 0 50,000 100,000 150,000 200,000 250,000 Revenues Margin pressure Revenues post margin pressure Costs New Operating income Cost cutting Initial operating income % change in revenues % change in operating % fall in costs required to offset revenue margin compression -27% -27% -25% -24% -22% -22% -21% -20% -19% -18% -17% -16% -22% -30% -25% -20% -15% -10% -5% 0% Bank 8 Bank 6 Bank 12 Bank 1 Bank 9 Bank 7 Bank 5 Bank 3 Bank 4 Bank 11 Bank 10 Bank 2 Aggregate NII Fees
  • 33. RegObs Special Report 07 December 2022 ◆ 33 c.1/2 reduction in costs required to hold profitability steady In our optimistic case scenario, we estimate that core standardised revenues could decrease by c22%, meaning that if costs remain the same, the operating income of the standardised bank could decrease by c44%. Hence, we estimate that cost should decrease by c45% to offset the margin pressure and leaving the aggregated profitability of our banks’ sample unchanged. In this case the C/I ratio for the standardised bank would need to fall to c.35% from c.50%. Picture 31: Optimistic Case – potential impact to the standardized operating income Source: Mediobanca Securities #3: Movie rental scenario - Pessimistic case Standardised revenues decline by 36% The final scenario we show is for a more “pessimistic case”. This more closely traces the very high margin compressions we have seen in industries like movie & TV rental/streaming of around -50%. Here we might observe standardised revenues coming down around -36% on aggregate. In this instance, we could even see some banks have around 40% of their total standardised revenues eroded, with a minimum of 25% for the least impacted names. Picture 32: Pessimistic Case – Impact on standardized revenues Source: Mediobanca Securities -14% -8% -22% -44% -45% 0 20,000 40,000 60,000 80,000 100,000 120,000 140,000 160,000 180,000 200,000 Revenues NII margin pressure Fee margin pressure Revenues post margin pressure Costs New Operating income Cost cutting Initial operating income % change in revenues % change in operating income % fall in costs required to offset revenue margin compression -43% -41% -40% -40% -37% -35% -34% -32% -30% -29% -28% -25% -36% -50% -45% -40% -35% -30% -25% -20% -15% -10% -5% 0% Bank 6 Bank 8 Bank 12 Bank 1 Bank 7 Bank 9 Bank 5 Bank 3 Bank 4 Bank 11 Bank 10 Bank 2 Aggregate NII Fees
  • 34. RegObs Special Report 07 December 2022 ◆ 34 Cutting cost base by c73% to retain the same profitability… In our pessimistic case scenario, we estimate that core revenues could decrease by c36% (vs 29% and 22% in our base and optimistic scenarios, respectively), leading to 70% cut of the operating income, without taking into consideration any cost optimization. Hence, we estimate that cost should decrease by c73% to offset the margin pressure, to perpetuate the profitability pre-disruption. Picture 33: Pessimistic Case – potential impact to the aggregated operating income Source: Mediobanca Securities …implying C/I c.20% in the standardised bank; c.35% at standardised+customised level With 73% cost cuts required to offset the margin compression in the pessimistic scenario, our sample of standardised banks would be required to bring their C/I ratio down significantly to c.20%. Leaving the customised perimeter unchanged, this would imply c.35% C/I at standardised+customised level, well below the current best-in-class Nordic banks. The transformation of standardised banking The current aggregated core revenues (NII+fees) for the 12 banks in our sample works out to be around €240bn (split c.€150bn standardised and c.€90bn customised). Whereas the relevant expense base to support these revenues is c.€120bn (split c.€75bn standardised and c.€45bn customised). This gives us a pre-provision profit of roughly €120bn, again split 60-65% standardised and 35-40% customised. The table shows the deep transformation of standardised banking versus the stability in customised banking. -24% -12% -36% -70% -73% 0 20,000 40,000 60,000 80,000 100,000 120,000 140,000 160,000 180,000 200,000 Revenues NII margin pressure Fee margin pressure Revenues post margin pressure Costs New Operating income Cost cutting Initial operating income % change in revenues % change in operating income % fall in costs required to offset revenue margin compression
  • 35. RegObs Special Report 07 December 2022 ◆ 35 Table 3: Summary of aggregate P&L under adverse, base and optimistic scenarios Source: Mediobanca Securities, EBA Current Pessimistic Base Optimistic NII [standardised] 94 57 65 72 NII [customised] 48 48 48 48 NII [total] 142 106 113 120 Fees [standardised] 57 40 43 46 Fees [customised] 40 40 40 40 Fees [total] 97 80 83 85 Rev enues [standardised] 151 97 107 118 Rev enues [customised] 88 88 88 88 Revenues [total] 240 185 196 206 Revenue deterioration [% current] -23% -18% -14% Expenses [standardised] (74) (20) (30) (41) Expenses [customised] (47) (47) (47) (47) Expenses [total] (121) (67) (77) (87) Cost cutting [% current] 0% -45% -36% -28% C/I [standardised] 49% 21% 28% 35% C/I [customised] 53% 53% 53% 53% C/I [total] 51% 36% 39% 42%
  • 36. RegObs Special Report 07 December 2022 ◆ 36 The great (cost) reset: scaling up DLT to do the trick Despite being in vogue, DLT technology is still disproportionately used merely in small, experimental projects. This makes it hard to extrapolate what it could/would do, if applied to a scale business, either assuming the trilemma will be solved at some point or that permissioned- DLT can do so already. Our intense dialogue with industry experts helps us to move from theory to practice in formulating the cost function of a (permissioned) DLT network, featuring fixed costs which progressively handover to variable transaction costs as volumes pick up. The Spunta DLT case validates our function which lands c.2-3x above its costs (a spread of €2-4m), i.e. embedding a margin of conservatism. Our conclusion is that DLT technology could provide banks with astronomic cost efficiencies in some of their businesses. Lack of evidence on blockchain’s running costs An often quoted paper (link) estimates DLT technology “could reduce banks’ infrastructure costs attributed to cross-border payments, securities trading and regulatory compliance by between $15- 20bn”. While there are oceans of narrative about the immense cost-saving opportunity of digitisation, there is very little practical evidence about the quantum involved. For example, there is no easy evidence of what the running costs of a DLT-based platform look like. Sunk development costs, linear opex until scale economies kick in to flatten the curve Logically, the cost function of DLT should involve sunk development costs. These will vary with the complexity of the applications that the DLT network is intended to operate and to the number of nodes required. Counterintuitively, all things are considered, a permissionless network is likely to cost more than a permissioned one. In fact, while the former may require lower sunk investment to setup the network, it will also require higher operating costs to encompass transparency with privacy, entailing the archive and accessibility of ever larger amounts of data for an overall significant overhead. Instead, permissioned DLT networks will be leaner on operating costs once the higher sunk setup investment on machinery is satisfied, i.e. resulting in lower individual transaction costs when implementing the tool on large volumes. We have spoken to industry experts to try to gauge the shape of the cost function of a permissioned DLT network, which we summarise here: 1. Hardware per node – blockchain ledger and part dedicated to support the application - €30- 35k per annum 2. Service per node – staff, hardware maintenance, helpdesk – 0.1x FTE in case of fragmentation in network managers, 0.025-0.05x when optimising 3. Private network connection per node – running the DLT network on a private network incurs into costs depending on the type of network, the throughput etc - €12-15k per annum 4. Hardware - One single hardware component such as transaction validation / notary functions - €150k per annum regardless of the number of nodes in the DLT network 5. Finally, the cost of DLT software licencing. This can vary dramatically from the menu price (€0.3/transaction even reaching levels 100x smaller) In essence, what we have gathered is a cost function looking like: 𝐴𝑛𝑛𝑢𝑎𝑙 𝐶𝑜𝑠𝑡 = €150,000 + (€35,000 + 0.05 ∗ 𝐹𝑇𝐸 𝑐𝑜𝑠𝑡 + €15,000) ∗ 𝑛 + 𝑡 (𝑘) ∗ 𝑘 Where n = number of DLT nodes, t = the transaction cost and k = the number of transactions, with t progressively decreasing as k expands.
  • 37. RegObs Special Report 07 December 2022 ◆ 37 Spunta, the largest DLT network in operation We have dug far and wide to find a living example of DLT network to gauge the costs related to operating it. Spunta, is a DLT-based network comprised of 100 Italian banks, dedicated to interbank account reconciliation governed by ABI Lab, part of the Italian Banking Association. Spunta’s application “verifies the matching of correspondent accounts that involve two different banks. The interbank reconciliation procedure in Italy is linked to processes traditionally carried out by the back office and aimed at reconciling the transaction flows that generate accounting entries in the mutual accounts in Italy and at managing pending transactions. Up to now [then for today’s reader], reconciliation was based on bilateral registers with a low level of standardisation and operating processes that were not very advanced. The implementation of a blockchain-based process using Distributed Ledger Technology (DLT) for interbank reconciliations in Italy makes it possible to automatically detect non-matching transactions using a shared algorithm that standardises both the process and the single communication channel, and provides a comprehensive view of the transactions among the interested parties. As a consequence, the principles of the new Spunta envisage full visibility of the transactions and those of the counterparty; rapid management of the flows with daily, rather than monthly, reconciliations; shared rules for the symmetrical reconciliation of transactions between counterparty banks; and the integrated management of communications and processes in the event of an imbalance”. See https://www.abilab.it/en/aree-ricerca/blockchain- dlt/spunta-banca-dlt. Spunta only operates one hour a day, carrying out c.350m transactions a year. “If this application will be applied to more complex cases, working at its full capacity, it was estimated that the infrastructure would be able to manage a total volume of 8.4 billion transactions. As term of comparison, Bitcoin’s blockchain in the entire 2020 managed 113 million transactions”. Spunta’s costs: €2-3m a year So how much does the largest DLT network cost? Going through the annual reports of ABI Lab, the entity operating Spunta for ABI, we were able to detect annual revenues budgeted from Blockchain & DLT of €2-3m per annum, growing slightly in 2021 (so for 2022). As we would expect ABI Lab to operate Spunta at a breakeven regime for the ABI affiliate banks, we consider this a good proxy for the comprehensive costs of the Spunta DLT application. Picture 34: blockchain cost evolution, € Source: Mediobanca Securities, ABI LAB annual report Fixed costs handover to variable costs beyond 10bn transactions Picture 35 plots the cost evolution from our cost function calibrated for DLT networks of 100, 320 and 640 nodes, ramping up the volume of transactions from 10m to 10bn. This illustrative example shows the high incidence on total cost of the network size and of other fixed cost when the network is relatively new and transaction volumes are low. Instead, when transactions start to grow in the - 500,000 1,000,000 1,500,000 2,000,000 2,500,000 3,000,000 2019 2020 2021