SlideShare ist ein Scribd-Unternehmen logo
1 von 34
Deep Hedging of Derivatives Using
Reinforcement Learning
Hull et al. working paper
๋ฐœํ‘œ์ž : ์œค์ง€์ƒ
Graduate School of Information. Yonsei Univ.
Machine Learning & Computational Finance Lab.
1. Introduction
2. Hedging
3. Setting Hedging model
4. Experiments
5. Conclusion
INDEX
1 Introduction
1. Introduction
When someone conduct risk management, hedging is very common and
important thing to do
But theoretical hedging cannot be fitted to real-world problem exactly because
of market friction
1. Introduction
Hedging is sequential optimal control task
&
RL is sequential optimal control task
Then can we implement RL to hedging task to
reduce total hedging cost?
2 Hedging
2. Hedging
Hedging
Short 1 call option
๐ถ๐‘‡ = max(๐‘†๐‘‡ โˆ’ ๐พ, 0)
Long 1 call option
Underlying asset
-0.4 +0.4
+1
-3
+5
+3
+6
+0.9 - 0.9
+1.2
-1.2
+0.7
+2
-0.7
-2
P&L P&L
Stock
movement
me
2. Hedging
Hedging
Short 1 call option
๐ถ๐‘‡ = max(๐‘†๐‘‡ โˆ’ ๐พ, 0)
Long 1 call option
Underlying asset
-0.4 +0.4
+1
-3
+5
+3
+6
+0.9 - 0.9
+1.2
-1.2
+0.7
+2
-0.7
-2
Margin call
P&L P&L
Stock
movement
cashflow
-1.4
Margin call
-2
Total cashflow of naked(not hedging) position = -3.4
me
2. Hedging
Hedging
Short 1 call option
๐ถ๐‘‡ = max(๐‘†๐‘‡ โˆ’ ๐พ, 0)
Long 1 call option
Underlying asset
-0.4 +0.4
+1
-3
+5
+3
+6
+0.9 - 0.9
+1.2
-1.2
+0.7
+2
-0.7
-2
P&L P&L
Stock
movement
cashflow P&L from hedge
+0.3
-0.8
+1.4
+0.8
+1.6
Total cashflow of hedged position = 0
me
2. Hedging
Hedging
2. Hedging
Delta-hedging
โˆ†= ๐‘ ๐‘‘1 =
๐œ•๐ถ
๐œ•๐‘†
So when we take position amount of โˆ†, portfolio profit is almost zero
If volatility of underlying asset is very high, or hedging period is too wide, hedge will
not be effective
2. Hedging
Delta-hedging
Theoretically, CONTINUOUS Delta-hedging with NO transaction cost can make
perfect hedged portfolio.
2. Hedging
Delta-hedging
Theoretically, CONTINUOUS Delta-hedging with NO transaction cost can make
perfect hedged portfolio.
Hedging more
frequently
Decrease
Transaction Cost
3 Setting Hedging model
3. Setting Hedging model
State
1. The holding of the asset
during the previous time period((๐‘– โˆ’ 1)ฮ”๐‘ก~๐‘–ฮ”๐‘ก) : ๐ป๐‘–โˆ’1
2. The asset price at time(๐‘–ฮ”๐‘ก) : ๐‘†๐‘–
3. The time to maturity : (๐‘› โˆ’ ๐‘–)ฮ”๐‘ก
Action
The amount of the asset to be held from time ๐‘–ฮ”๐‘ก to time (๐‘– + 1)ฮ”๐‘ก : ๐ป๐‘–
State & Action
โ€ข Time-step : ฮ”๐‘ก
โ€ข The life of the option : ๐‘›ฮ”๐‘ก
3. Setting Hedging model
Accounting P&L formulation
๐‘…๐‘–+1 = ๐‘‰๐‘–+1 โˆ’ ๐‘‰๐‘– + ๐ป๐‘– ๐‘†๐‘–+1 โˆ’ ๐‘†๐‘– โˆ’ ๐œ…|๐‘†๐‘–+1 ๐ป๐‘–+1 โˆ’ ๐ป๐‘– |
When we derive reward function as accounting P&L formulation,
reward function to minimize can be:
where
โ€ข ๐‘‰๐‘– : Derivatives value at time-step ๐‘–ฮ”๐‘ก
โ€ข ๐‘†๐‘– : Underlying asset value at time-step ๐‘–ฮ”๐‘ก
โ€ข ๐ป๐‘–: Position of underlying asset relative to position of derivatives
โ€ข ๐œ… : Trading cost parameter
In addition, there are an initial reward โˆ’๐œ…|๐‘†0๐ป0| and final reward โˆ’๐œ…|๐‘†๐‘›๐ป๐‘›|
to set up(liquidate) the hedge position at first(last) time-step
if long, positive value
if short, negative value
3. Setting Hedging model
Cash Flow Formulation
๐‘…๐‘–+1 = ๐‘†๐‘–+1 ๐ป๐‘– โˆ’ ๐ป๐‘–+1 โˆ’ ๐œ…|๐‘†๐‘–+1 ๐ป๐‘–+1 โˆ’ ๐ป๐‘– |
When we derive reward function as cash flow formulation,
reward function to minimize can be:
where
โ€ข ๐‘†๐‘– : Underlying asset value at time-step ๐‘–ฮ”๐‘ก
โ€ข ๐ป๐‘–: Position of underlying asset relative to position of derivatives
โ€ข ๐œ… : Trading cost parameter
In addition, there are other rewards
โ€ข Initial rewards : โˆ’๐‘†0๐ป0 โˆ’ ๐œ…|๐‘†0๐ป0| at first time-step
โ€ข final rewards : ๐‘†๐‘›๐ป๐‘› โˆ’ ๐œ… ๐‘†๐‘›๐ป๐‘› + ๐‘๐‘Ž๐‘ฆ๐‘œ๐‘“๐‘“ ๐‘œ๐‘“ ๐‘‘๐‘’๐‘Ÿ๐‘–๐‘ฃ๐‘Ž๐‘ก๐‘–๐‘ฃ๐‘’๐‘  at last time-step
if long, positive value
if short, negative value
3. Setting Hedging model
Approach Comparison
๐‘†๐‘–+1 ๐ป๐‘– โˆ’ ๐ป๐‘–+1 โˆ’ ๐œ…|๐‘†๐‘–+1 ๐ป๐‘–+1 โˆ’ ๐ป๐‘– |
๐‘‰๐‘–+1 โˆ’ ๐‘‰๐‘– + ๐ป๐‘– ๐‘†๐‘–+1 โˆ’ ๐‘†๐‘– โˆ’ ๐œ…|๐‘†๐‘–+1 ๐ป๐‘–+1 โˆ’ ๐ป๐‘– |
Accounting P&L approach reward
โˆ’๐œ…|๐‘†0๐ป0|
โˆ’๐œ…|๐‘†๐‘›๐ป๐‘›|
At time-step 1
At time-step 2~(๐‘› โˆ’ 1)
At time-step ๐‘›
Cash Flow approach reward
At time-step 1
At time-step 2~(๐‘› โˆ’ 1)
At time-step ๐‘›
โˆ’๐‘†0๐ป0 โˆ’ ๐œ…|๐‘†0๐ป0|
๐‘†๐‘›๐ป๐‘› โˆ’ ๐œ… ๐‘†๐‘›๐ป๐‘› + ๐‘‰
๐‘›
3. Setting Hedging model
Approach Comparison
๐‘†๐‘–+1 ๐ป๐‘– โˆ’ ๐ป๐‘–+1 โˆ’ ๐œ…|๐‘†๐‘–+1 ๐ป๐‘–+1 โˆ’ ๐ป๐‘– |
๐‘‰๐‘–+1 โˆ’ ๐‘‰๐‘– + ๐ป๐‘– ๐‘†๐‘–+1 โˆ’ ๐‘†๐‘– โˆ’ ๐œ…|๐‘†๐‘–+1 ๐ป๐‘–+1 โˆ’ ๐ป๐‘– |
Accounting P&L approach reward
โˆ’๐œ…|๐‘†0๐ป0|
โˆ’๐œ…|๐‘†๐‘›๐ป๐‘›|
At time-step 1
At time-step 2~(๐‘› โˆ’ 1)
At time-step ๐‘›
Cash Flow approach reward
At time-step 1
At time-step 2~(๐‘› โˆ’ 1)
At time-step ๐‘›
โˆ’๐‘†0๐ป0 โˆ’ ๐œ…|๐‘†0๐ป0|
๐‘†๐‘›๐ป๐‘› โˆ’ ๐œ… ๐‘†๐‘›๐ป๐‘› + ๐‘‰
๐‘›
When we use Accounting P&L approach
reward, we should know derivatives
pricing model
3. Setting Hedging model
Approach Comparison
โ€ฆ
โ€ฆ
Time-step
Time-step
reward
reward
๏‚ง Accounting P&L approach rewards are almost zero-near value.
โ†’ to minimize cost (reward), model just train to make rewards at every time step equal zero
๏‚ง However, Cash Flow approach rewards are not similar each other.
โ†’ to minimize cost, model should learn pricing model and is hard to converge because of
credit assignment problem
3. Setting Hedging model
๐‘Œ ๐‘ก = ๐”ผ ๐ถ๐‘ก + ๐‘ ๐”ผ ๐ถ๐‘ก
2
โˆ’ ๐”ผ ๐ถ๐‘ก
2
Model in this work
๐น ๐‘†๐‘ก, ๐‘Ž = ๐‘„1(๐‘†๐‘ก, ๐‘Ž) + ๐‘ ๐‘„2(๐‘†๐‘ก, ๐‘Ž) โˆ’ ๐‘„1 ๐‘†๐‘ก, ๐‘Ž 2
Two Q-values are introduced,
๐‘„1 estimates the expected cost for state-action combinations
๐‘„1 โ‰ˆ ๐”ผ ๐ถ๐‘ก
๐‘„1 estimates the expected value of the square of the cost for state-action combinations
๐‘„2 โ‰ˆ ๐”ผ ๐ถ๐‘ก
2
Expectation of
hedging cost
volatility of
hedging cost
Set cost equation ๐‘Œ ๐‘ก to minimize
where ๐”ผ ๐ถ๐‘ก is expectation of hedging cost for time ๐‘ก ~ maturity
3. Setting Hedging model
Model in this work
Critic ๐‘„1& ๐‘„2 update with loss function:
๐‘…๐‘ก+1 + ๐›พ๐‘„1 ๐‘†๐‘ก+1, ๐œ‹ ๐‘†๐‘ก+1 โˆ’ ๐‘„1 ๐‘†๐‘ก, ๐ด๐‘ก; ๐‘ค1
2
๐‘…๐‘ก+1
2
+ ๐›พ2
๐‘„2 ๐‘†๐‘ก+1, ๐œ‹ ๐‘†๐‘ก+1 + 2๐›พ๐‘…๐‘ก+1๐‘„1 ๐‘†๐‘ก+1, ๐œ‹ ๐‘†๐‘ก+1 โˆ’ ๐‘„2 ๐‘†๐‘ก, ๐ด๐‘ก; ๐‘ค2
2
Actor ๐œ‹ update as:
๐œƒ โ† ๐œƒ โˆ’ ๐›ผโˆ‡๐œƒ๐น(๐‘†๐‘ก, ๐œ‹ ๐‘†๐‘ก; ๐œƒ )
โˆ‡๐œƒ๐น ๐‘†๐‘ก, ๐œ‹ ๐‘†๐‘ก; ๐œƒ = โˆ‡๐œƒ๐‘„1(๐‘†๐‘ก, ๐‘Ž) + ๐‘ฉ(โˆ‡๐œƒ๐‘„2 ๐‘†๐‘ก, ๐‘Ž โˆ’ 2๐‘„1 ๐‘†๐‘ก, ๐‘Ž โˆ‡๐œƒ๐‘„1 ๐‘†๐‘ก, ๐‘Ž
where ๐‘ฉ =
๐‘
2
๐‘„2 ๐‘†๐‘ก, ๐‘Ž โˆ’ ๐‘„1 ๐‘†๐‘ก, ๐‘Ž 2 โˆ’
1
2
Since expected value of ๐‘„2 ๐‘†๐‘ก, ๐ด๐‘ก =expected value of ๐‘…๐‘ก+1 + ๐›พ๐‘„1 ๐‘†๐‘ก+1, ๐‘Ž 2
,
4Experiments
4. Experiments
Simulation Test
I. Geometric Brownian Motion Test
II. Stochastic Volatility Test
4. Experiments
Setting
โ€ข We are in short position on 1 call option of different time-to-maturity
1. 1-month
2. 3-months
โ€ข Strike price of call option ๐พ = ๐‘†0 (ATM at time-step 0)
โ€ข We can only use underlying stock to hedge.
โ€ข Using DDPG algorithm.
โ€ข Implement the prioritized experience replay method.
โ€ข Using Accounting P&L approach.
4. Experiments
I. Geometric Brownian Motion Test
where
๐‘†: Stock price
C: call option price
q: dividend yield
๐‘…๐‘–+1 = ๐‘‰๐‘–+1 โˆ’ ๐‘‰๐‘– + ๐ป๐‘– ๐‘†๐‘–+1 โˆ’ ๐‘†๐‘– โˆ’ ๐œ…|๐‘†๐‘–+1 ๐ป๐‘–+1 โˆ’ ๐ป๐‘– |
๐น ๐‘†๐‘ก, ๐‘Ž = ๐‘„1(๐‘†๐‘ก, ๐‘Ž) + ๐‘ ๐‘„2(๐‘†๐‘ก, ๐‘Ž) โˆ’ ๐‘„1 ๐‘†๐‘ก, ๐‘Ž 2
๐‘‘๐‘† = ๐œ‡๐‘†๐‘‘๐‘ก + ๐œŽ๐‘†๐‘‘๐‘ง
๐ถ = ๐‘†0๐‘’โˆ’๐‘ž๐‘‡
๐‘ ๐‘‘1 โˆ’ ๐พ๐‘’โˆ’๐‘Ÿ๐‘‡
๐‘ ๐‘‘2
๐‘‘1 =
ln
๐‘†0
๐พ
+ ๐‘Ÿโˆ’๐‘ž+
๐œŽ2
2
๐‘‡
๐œŽ ๐‘‡
๐‘‘2 = ๐‘‘1 โˆ’ ๐œŽ ๐‘‡
๐œ‡ = 5%, ๐‘Ÿ = 0, ๐‘ž = 0, ๐œŽ = 20%, ๐œ… = 1%, ๐‘ = 1.5
4. Experiments
I. Geometric Brownian Motion Test
<1-month call option>
<3-months call option>
4. Experiments
II. Stochastic Volatility Test
When an option is ATM, implied volatility is approximately ๐œŽ0๐ต
taking ๐œŽ0๐ต into Black-Scholes model as input ๐œŽ, we can value a call option
SABER model (๐›ฝ = 1)
๐‘‘๐‘† = ๐œ‡๐‘†๐‘‘๐‘ก + ๐œŽ๐‘†๐‘‘๐‘ง1
๐‘‘๐œŽ = ๐‘ฃ๐œŽ๐‘‘๐‘ง2
๐”ผ ๐‘‘๐‘ง1๐‘‘๐‘ง2 = ๐œŒ๐‘‘๐‘ก
where ๐‘ฃ: volatility of volatility
๐œŒ = โˆ’0.4, ๐œŽ0 = 20%, ๐‘ฃ = 60%, others = equal
๐น0 = ๐‘†0๐‘’ ๐‘Ÿโˆ’๐‘ž ๐‘‡
๐ต = 1 +
๐œŒ๐‘ฃ๐œŽ0
4
+
2โˆ’3๐œŒ2 ๐‘ฃ
24
๐‘‡
๐œ™ =
๐‘ฃ
๐œŽ0 ln
๐น0
๐พ
๐œ’ = ln
1โˆ’2๐œŒ+๐œ™2+๐œ™โˆ’๐œŒ
1โˆ’๐œŒ
4. Experiments
II. Stochastic Volatility Test
Our model is compared with 2 delta-hedging strategy
1. Bartlett Delta : Delta calculated by SABER
2. Practitioner Delta : Delta calculated by market implied volatility
4. Experiments
II. Stochastic Volatility Test
<1-month call option>
<3-months call option>
4. Experiments
a. our hedge instrument position is close to theoretical hedge position: Delta hedging
b. our hedge instrument position is much less than theoretical hedge position: being under-hedging
c. our hedge instrument position is much more than theoretical hedge position: being over-hedging
Since transaction cost is significant,
model donโ€™t take hedge position as much as model required
4. Experiments
Since transaction cost is significant,
model donโ€™t take hedge position as much as model required
When 0.6 delta is required and we take 0.5 delta hedge position, model take 0.1 delta more
When 0.9 delta is required and we take 0.5 delta hedge position, model take only 0.25 delta more
When 0.2 delta is required and we take 0.5 delta hedge position, model take only -0.2 delta more
5 Conclusion
1. Use not only simulated data but real-world data
2. More well-structured architecture is needed
3. Practical hedging method like hedging vol as well as delta-hedging should be
controlled by RL
4. Adaptive transaction cost can be introduced
5. Conclusion

Weitere รคhnliche Inhalte

Was ist angesagt?

microbiologia, enfermedades transmitidas por los alimentos
microbiologia, enfermedades transmitidas por los alimentosmicrobiologia, enfermedades transmitidas por los alimentos
microbiologia, enfermedades transmitidas por los alimentosYazz Macias G
ย 
Introducciรณn Enfermedades Infecciosas
Introducciรณn Enfermedades InfecciosasIntroducciรณn Enfermedades Infecciosas
Introducciรณn Enfermedades InfecciosasBenรญcio Araรบjo
ย 
Coliformes fecales
Coliformes fecales Coliformes fecales
Coliformes fecales KarlyToapanta
ย 
Incremento de la Resiliencia Climรกtica basada en Ecosistemas de Comunidades R...
Incremento de la Resiliencia Climรกtica basada en Ecosistemas de Comunidades R...Incremento de la Resiliencia Climรกtica basada en Ecosistemas de Comunidades R...
Incremento de la Resiliencia Climรกtica basada en Ecosistemas de Comunidades R...AIDA_Americas
ย 
Contaminacion del agua
Contaminacion del aguaContaminacion del agua
Contaminacion del aguaFranklin Flores
ย 
Saneamiento Ambiental - Trabajo de Investigaciรณn
Saneamiento Ambiental - Trabajo de InvestigaciรณnSaneamiento Ambiental - Trabajo de Investigaciรณn
Saneamiento Ambiental - Trabajo de InvestigaciรณnRobin Gomez Peรฑa
ย 
Infecciones Meningocรณcicas
Infecciones MeningocรณcicasInfecciones Meningocรณcicas
Infecciones MeningocรณcicasGeramel De la Cruz
ย 

Was ist angesagt? (10)

microbiologia, enfermedades transmitidas por los alimentos
microbiologia, enfermedades transmitidas por los alimentosmicrobiologia, enfermedades transmitidas por los alimentos
microbiologia, enfermedades transmitidas por los alimentos
ย 
Introducciรณn Enfermedades Infecciosas
Introducciรณn Enfermedades InfecciosasIntroducciรณn Enfermedades Infecciosas
Introducciรณn Enfermedades Infecciosas
ย 
Colera ExpoDila
Colera ExpoDilaColera ExpoDila
Colera ExpoDila
ย 
Coliformes fecales
Coliformes fecales Coliformes fecales
Coliformes fecales
ย 
Cianobacterias y cianotoxinas
Cianobacterias y cianotoxinasCianobacterias y cianotoxinas
Cianobacterias y cianotoxinas
ย 
Incremento de la Resiliencia Climรกtica basada en Ecosistemas de Comunidades R...
Incremento de la Resiliencia Climรกtica basada en Ecosistemas de Comunidades R...Incremento de la Resiliencia Climรกtica basada en Ecosistemas de Comunidades R...
Incremento de la Resiliencia Climรกtica basada en Ecosistemas de Comunidades R...
ย 
Contaminacion del agua
Contaminacion del aguaContaminacion del agua
Contaminacion del agua
ย 
Colera ok
Colera okColera ok
Colera ok
ย 
Saneamiento Ambiental - Trabajo de Investigaciรณn
Saneamiento Ambiental - Trabajo de InvestigaciรณnSaneamiento Ambiental - Trabajo de Investigaciรณn
Saneamiento Ambiental - Trabajo de Investigaciรณn
ย 
Infecciones Meningocรณcicas
Infecciones MeningocรณcicasInfecciones Meningocรณcicas
Infecciones Meningocรณcicas
ย 

ร„hnlich wie PPT - Deep Hedging OF Derivatives Using Reinforcement Learning

PPT - Adaptive Quantitative Trading : An Imitative Deep Reinforcement Learnin...
PPT - Adaptive Quantitative Trading : An Imitative Deep Reinforcement Learnin...PPT - Adaptive Quantitative Trading : An Imitative Deep Reinforcement Learnin...
PPT - Adaptive Quantitative Trading : An Imitative Deep Reinforcement Learnin...Jisang Yoon
ย 
Financial Trading as a Game: A Deep Reinforcement Learning Approach
Financial Trading as a Game: A Deep Reinforcement Learning ApproachFinancial Trading as a Game: A Deep Reinforcement Learning Approach
Financial Trading as a Game: A Deep Reinforcement Learning Approach่ฌ™็›Š ้ปƒ
ย 
"Quantitative Trading as a Mathematical Science" by Dr. Haksun Li, Founder an...
"Quantitative Trading as a Mathematical Science" by Dr. Haksun Li, Founder an..."Quantitative Trading as a Mathematical Science" by Dr. Haksun Li, Founder an...
"Quantitative Trading as a Mathematical Science" by Dr. Haksun Li, Founder an...Quantopian
ย 
Intro to Quantitative Investment (Lecture 1 of 6)
Intro to Quantitative Investment (Lecture 1 of 6)Intro to Quantitative Investment (Lecture 1 of 6)
Intro to Quantitative Investment (Lecture 1 of 6)Adrian Aley
ย 
Reinforcement learning Research experiments OpenAI
Reinforcement learning Research experiments OpenAIReinforcement learning Research experiments OpenAI
Reinforcement learning Research experiments OpenAIRaouf KESKES
ย 
Intro to Quant Trading Strategies (Lecture 8 of 10)
Intro to Quant Trading Strategies (Lecture 8 of 10)Intro to Quant Trading Strategies (Lecture 8 of 10)
Intro to Quant Trading Strategies (Lecture 8 of 10)Adrian Aley
ย 
Daa unit 1
Daa unit 1Daa unit 1
Daa unit 1jinalgoti
ย 
IFTA2020 Kei Nakagawa
IFTA2020 Kei NakagawaIFTA2020 Kei Nakagawa
IFTA2020 Kei NakagawaKei Nakagawa
ย 
Differential Machine Learning Masterclass
Differential Machine Learning MasterclassDifferential Machine Learning Masterclass
Differential Machine Learning MasterclassAntoine Savine
ย 
Demystifying deep reinforement learning
Demystifying deep reinforement learningDemystifying deep reinforement learning
Demystifying deep reinforement learning์žฌ์—ฐ ์œค
ย 
Reinfrocement Learning
Reinfrocement LearningReinfrocement Learning
Reinfrocement LearningNatan Katz
ย 
Intro to Quantitative Investment (Lecture 6 of 6)
Intro to Quantitative Investment (Lecture 6 of 6)Intro to Quantitative Investment (Lecture 6 of 6)
Intro to Quantitative Investment (Lecture 6 of 6)Adrian Aley
ย 
Deep Reinforcement Learning
Deep Reinforcement LearningDeep Reinforcement Learning
Deep Reinforcement LearningMeetupDataScienceRoma
ย 
adversarial robustness lecture
adversarial robustness lectureadversarial robustness lecture
adversarial robustness lectureMuhammadAhmedShah2
ย 
daa-unit-3-greedy method
daa-unit-3-greedy methoddaa-unit-3-greedy method
daa-unit-3-greedy methodhodcsencet
ย 
Quantitativetechniqueformanagerialdecisionlinearprogramming 090725035417-phpa...
Quantitativetechniqueformanagerialdecisionlinearprogramming 090725035417-phpa...Quantitativetechniqueformanagerialdecisionlinearprogramming 090725035417-phpa...
Quantitativetechniqueformanagerialdecisionlinearprogramming 090725035417-phpa...kongara
ย 
Market Risk Modelling
Market Risk ModellingMarket Risk Modelling
Market Risk Modellingav vedpuriswar
ย 
Optimum Investment Selection process-Nov 9-2013
Optimum Investment Selection process-Nov 9-2013Optimum Investment Selection process-Nov 9-2013
Optimum Investment Selection process-Nov 9-2013Gary Crosbie
ย 

ร„hnlich wie PPT - Deep Hedging OF Derivatives Using Reinforcement Learning (20)

PPT - Adaptive Quantitative Trading : An Imitative Deep Reinforcement Learnin...
PPT - Adaptive Quantitative Trading : An Imitative Deep Reinforcement Learnin...PPT - Adaptive Quantitative Trading : An Imitative Deep Reinforcement Learnin...
PPT - Adaptive Quantitative Trading : An Imitative Deep Reinforcement Learnin...
ย 
Financial Trading as a Game: A Deep Reinforcement Learning Approach
Financial Trading as a Game: A Deep Reinforcement Learning ApproachFinancial Trading as a Game: A Deep Reinforcement Learning Approach
Financial Trading as a Game: A Deep Reinforcement Learning Approach
ย 
"Quantitative Trading as a Mathematical Science" by Dr. Haksun Li, Founder an...
"Quantitative Trading as a Mathematical Science" by Dr. Haksun Li, Founder an..."Quantitative Trading as a Mathematical Science" by Dr. Haksun Li, Founder an...
"Quantitative Trading as a Mathematical Science" by Dr. Haksun Li, Founder an...
ย 
Intro to Quantitative Investment (Lecture 1 of 6)
Intro to Quantitative Investment (Lecture 1 of 6)Intro to Quantitative Investment (Lecture 1 of 6)
Intro to Quantitative Investment (Lecture 1 of 6)
ย 
Reinforcement learning Research experiments OpenAI
Reinforcement learning Research experiments OpenAIReinforcement learning Research experiments OpenAI
Reinforcement learning Research experiments OpenAI
ย 
Intro to Quant Trading Strategies (Lecture 8 of 10)
Intro to Quant Trading Strategies (Lecture 8 of 10)Intro to Quant Trading Strategies (Lecture 8 of 10)
Intro to Quant Trading Strategies (Lecture 8 of 10)
ย 
Daa unit 1
Daa unit 1Daa unit 1
Daa unit 1
ย 
IFTA2020 Kei Nakagawa
IFTA2020 Kei NakagawaIFTA2020 Kei Nakagawa
IFTA2020 Kei Nakagawa
ย 
static_hedge
static_hedgestatic_hedge
static_hedge
ย 
Differential Machine Learning Masterclass
Differential Machine Learning MasterclassDifferential Machine Learning Masterclass
Differential Machine Learning Masterclass
ย 
Demystifying deep reinforement learning
Demystifying deep reinforement learningDemystifying deep reinforement learning
Demystifying deep reinforement learning
ย 
Reinfrocement Learning
Reinfrocement LearningReinfrocement Learning
Reinfrocement Learning
ย 
Schema anf
Schema anfSchema anf
Schema anf
ย 
Intro to Quantitative Investment (Lecture 6 of 6)
Intro to Quantitative Investment (Lecture 6 of 6)Intro to Quantitative Investment (Lecture 6 of 6)
Intro to Quantitative Investment (Lecture 6 of 6)
ย 
Deep Reinforcement Learning
Deep Reinforcement LearningDeep Reinforcement Learning
Deep Reinforcement Learning
ย 
adversarial robustness lecture
adversarial robustness lectureadversarial robustness lecture
adversarial robustness lecture
ย 
daa-unit-3-greedy method
daa-unit-3-greedy methoddaa-unit-3-greedy method
daa-unit-3-greedy method
ย 
Quantitativetechniqueformanagerialdecisionlinearprogramming 090725035417-phpa...
Quantitativetechniqueformanagerialdecisionlinearprogramming 090725035417-phpa...Quantitativetechniqueformanagerialdecisionlinearprogramming 090725035417-phpa...
Quantitativetechniqueformanagerialdecisionlinearprogramming 090725035417-phpa...
ย 
Market Risk Modelling
Market Risk ModellingMarket Risk Modelling
Market Risk Modelling
ย 
Optimum Investment Selection process-Nov 9-2013
Optimum Investment Selection process-Nov 9-2013Optimum Investment Selection process-Nov 9-2013
Optimum Investment Selection process-Nov 9-2013
ย 

Kรผrzlich hochgeladen

Call Girls Indiranagar Just Call ๐Ÿ‘— 9155563397 ๐Ÿ‘— Top Class Call Girl Service B...
Call Girls Indiranagar Just Call ๐Ÿ‘— 9155563397 ๐Ÿ‘— Top Class Call Girl Service B...Call Girls Indiranagar Just Call ๐Ÿ‘— 9155563397 ๐Ÿ‘— Top Class Call Girl Service B...
Call Girls Indiranagar Just Call ๐Ÿ‘— 9155563397 ๐Ÿ‘— Top Class Call Girl Service B...only4webmaster01
ย 
Call Girls In Bellandur โ˜Ž 7737669865 ๐Ÿฅต Book Your One night Stand
Call Girls In Bellandur โ˜Ž 7737669865 ๐Ÿฅต Book Your One night StandCall Girls In Bellandur โ˜Ž 7737669865 ๐Ÿฅต Book Your One night Stand
Call Girls In Bellandur โ˜Ž 7737669865 ๐Ÿฅต Book Your One night Standamitlee9823
ย 
DATA SUMMIT 24 Building Real-Time Pipelines With FLaNK
DATA SUMMIT 24  Building Real-Time Pipelines With FLaNKDATA SUMMIT 24  Building Real-Time Pipelines With FLaNK
DATA SUMMIT 24 Building Real-Time Pipelines With FLaNKTimothy Spann
ย 
๐Ÿ‘‰ Amritsar Call Girl ๐Ÿ‘‰๐Ÿ“ž 6367187148 ๐Ÿ‘‰๐Ÿ“ž Just๐Ÿ“ฒ Call Ruhi Call Girl Phone No Amri...
๐Ÿ‘‰ Amritsar Call Girl ๐Ÿ‘‰๐Ÿ“ž 6367187148 ๐Ÿ‘‰๐Ÿ“ž Just๐Ÿ“ฒ Call Ruhi Call Girl Phone No Amri...๐Ÿ‘‰ Amritsar Call Girl ๐Ÿ‘‰๐Ÿ“ž 6367187148 ๐Ÿ‘‰๐Ÿ“ž Just๐Ÿ“ฒ Call Ruhi Call Girl Phone No Amri...
๐Ÿ‘‰ Amritsar Call Girl ๐Ÿ‘‰๐Ÿ“ž 6367187148 ๐Ÿ‘‰๐Ÿ“ž Just๐Ÿ“ฒ Call Ruhi Call Girl Phone No Amri...karishmasinghjnh
ย 
April 2024 - Crypto Market Report's Analysis
April 2024 - Crypto Market Report's AnalysisApril 2024 - Crypto Market Report's Analysis
April 2024 - Crypto Market Report's Analysismanisha194592
ย 
VIP Model Call Girls Hinjewadi ( Pune ) Call ON 8005736733 Starting From 5K t...
VIP Model Call Girls Hinjewadi ( Pune ) Call ON 8005736733 Starting From 5K t...VIP Model Call Girls Hinjewadi ( Pune ) Call ON 8005736733 Starting From 5K t...
VIP Model Call Girls Hinjewadi ( Pune ) Call ON 8005736733 Starting From 5K t...SUHANI PANDEY
ย 
Call Girls Bommasandra Just Call ๐Ÿ‘— 7737669865 ๐Ÿ‘— Top Class Call Girl Service B...
Call Girls Bommasandra Just Call ๐Ÿ‘— 7737669865 ๐Ÿ‘— Top Class Call Girl Service B...Call Girls Bommasandra Just Call ๐Ÿ‘— 7737669865 ๐Ÿ‘— Top Class Call Girl Service B...
Call Girls Bommasandra Just Call ๐Ÿ‘— 7737669865 ๐Ÿ‘— Top Class Call Girl Service B...amitlee9823
ย 
Detecting Credit Card Fraud: A Machine Learning Approach
Detecting Credit Card Fraud: A Machine Learning ApproachDetecting Credit Card Fraud: A Machine Learning Approach
Detecting Credit Card Fraud: A Machine Learning ApproachBoston Institute of Analytics
ย 
Thane Call Girls 7091864438 Call Girls in Thane Escort service book now -
Thane Call Girls 7091864438 Call Girls in Thane Escort service book now -Thane Call Girls 7091864438 Call Girls in Thane Escort service book now -
Thane Call Girls 7091864438 Call Girls in Thane Escort service book now -Pooja Nehwal
ย 
โžฅ๐Ÿ” 7737669865 ๐Ÿ”โ–ป Bangalore Call-girls in Women Seeking Men ๐Ÿ”Bangalore๐Ÿ” Esc...
โžฅ๐Ÿ” 7737669865 ๐Ÿ”โ–ป Bangalore Call-girls in Women Seeking Men  ๐Ÿ”Bangalore๐Ÿ”   Esc...โžฅ๐Ÿ” 7737669865 ๐Ÿ”โ–ป Bangalore Call-girls in Women Seeking Men  ๐Ÿ”Bangalore๐Ÿ”   Esc...
โžฅ๐Ÿ” 7737669865 ๐Ÿ”โ–ป Bangalore Call-girls in Women Seeking Men ๐Ÿ”Bangalore๐Ÿ” Esc...amitlee9823
ย 
Capstone Project on IBM Data Analytics Program
Capstone Project on IBM Data Analytics ProgramCapstone Project on IBM Data Analytics Program
Capstone Project on IBM Data Analytics ProgramMoniSankarHazra
ย 
Call Girls In Attibele โ˜Ž 7737669865 ๐Ÿฅต Book Your One night Stand
Call Girls In Attibele โ˜Ž 7737669865 ๐Ÿฅต Book Your One night StandCall Girls In Attibele โ˜Ž 7737669865 ๐Ÿฅต Book Your One night Stand
Call Girls In Attibele โ˜Ž 7737669865 ๐Ÿฅต Book Your One night Standamitlee9823
ย 
Mg Road Call Girls Service: ๐Ÿ“ 7737669865 ๐Ÿ“ High Profile Model Escorts | Banga...
Mg Road Call Girls Service: ๐Ÿ“ 7737669865 ๐Ÿ“ High Profile Model Escorts | Banga...Mg Road Call Girls Service: ๐Ÿ“ 7737669865 ๐Ÿ“ High Profile Model Escorts | Banga...
Mg Road Call Girls Service: ๐Ÿ“ 7737669865 ๐Ÿ“ High Profile Model Escorts | Banga...amitlee9823
ย 
Discover Why Less is More in B2B Research
Discover Why Less is More in B2B ResearchDiscover Why Less is More in B2B Research
Discover Why Less is More in B2B Researchmichael115558
ย 
Predicting Loan Approval: A Data Science Project
Predicting Loan Approval: A Data Science ProjectPredicting Loan Approval: A Data Science Project
Predicting Loan Approval: A Data Science ProjectBoston Institute of Analytics
ย 
โžฅ๐Ÿ” 7737669865 ๐Ÿ”โ–ป mahisagar Call-girls in Women Seeking Men ๐Ÿ”mahisagar๐Ÿ” Esc...
โžฅ๐Ÿ” 7737669865 ๐Ÿ”โ–ป mahisagar Call-girls in Women Seeking Men  ๐Ÿ”mahisagar๐Ÿ”   Esc...โžฅ๐Ÿ” 7737669865 ๐Ÿ”โ–ป mahisagar Call-girls in Women Seeking Men  ๐Ÿ”mahisagar๐Ÿ”   Esc...
โžฅ๐Ÿ” 7737669865 ๐Ÿ”โ–ป mahisagar Call-girls in Women Seeking Men ๐Ÿ”mahisagar๐Ÿ” Esc...amitlee9823
ย 
Call me @ 9892124323 Cheap Rate Call Girls in Vashi with Real Photo 100% Secure
Call me @ 9892124323  Cheap Rate Call Girls in Vashi with Real Photo 100% SecureCall me @ 9892124323  Cheap Rate Call Girls in Vashi with Real Photo 100% Secure
Call me @ 9892124323 Cheap Rate Call Girls in Vashi with Real Photo 100% SecurePooja Nehwal
ย 

Kรผrzlich hochgeladen (20)

Call Girls Indiranagar Just Call ๐Ÿ‘— 9155563397 ๐Ÿ‘— Top Class Call Girl Service B...
Call Girls Indiranagar Just Call ๐Ÿ‘— 9155563397 ๐Ÿ‘— Top Class Call Girl Service B...Call Girls Indiranagar Just Call ๐Ÿ‘— 9155563397 ๐Ÿ‘— Top Class Call Girl Service B...
Call Girls Indiranagar Just Call ๐Ÿ‘— 9155563397 ๐Ÿ‘— Top Class Call Girl Service B...
ย 
Call Girls In Bellandur โ˜Ž 7737669865 ๐Ÿฅต Book Your One night Stand
Call Girls In Bellandur โ˜Ž 7737669865 ๐Ÿฅต Book Your One night StandCall Girls In Bellandur โ˜Ž 7737669865 ๐Ÿฅต Book Your One night Stand
Call Girls In Bellandur โ˜Ž 7737669865 ๐Ÿฅต Book Your One night Stand
ย 
DATA SUMMIT 24 Building Real-Time Pipelines With FLaNK
DATA SUMMIT 24  Building Real-Time Pipelines With FLaNKDATA SUMMIT 24  Building Real-Time Pipelines With FLaNK
DATA SUMMIT 24 Building Real-Time Pipelines With FLaNK
ย 
Call Girls In Shalimar Bagh ( Delhi) 9953330565 Escorts Service
Call Girls In Shalimar Bagh ( Delhi) 9953330565 Escorts ServiceCall Girls In Shalimar Bagh ( Delhi) 9953330565 Escorts Service
Call Girls In Shalimar Bagh ( Delhi) 9953330565 Escorts Service
ย 
๐Ÿ‘‰ Amritsar Call Girl ๐Ÿ‘‰๐Ÿ“ž 6367187148 ๐Ÿ‘‰๐Ÿ“ž Just๐Ÿ“ฒ Call Ruhi Call Girl Phone No Amri...
๐Ÿ‘‰ Amritsar Call Girl ๐Ÿ‘‰๐Ÿ“ž 6367187148 ๐Ÿ‘‰๐Ÿ“ž Just๐Ÿ“ฒ Call Ruhi Call Girl Phone No Amri...๐Ÿ‘‰ Amritsar Call Girl ๐Ÿ‘‰๐Ÿ“ž 6367187148 ๐Ÿ‘‰๐Ÿ“ž Just๐Ÿ“ฒ Call Ruhi Call Girl Phone No Amri...
๐Ÿ‘‰ Amritsar Call Girl ๐Ÿ‘‰๐Ÿ“ž 6367187148 ๐Ÿ‘‰๐Ÿ“ž Just๐Ÿ“ฒ Call Ruhi Call Girl Phone No Amri...
ย 
Anomaly detection and data imputation within time series
Anomaly detection and data imputation within time seriesAnomaly detection and data imputation within time series
Anomaly detection and data imputation within time series
ย 
April 2024 - Crypto Market Report's Analysis
April 2024 - Crypto Market Report's AnalysisApril 2024 - Crypto Market Report's Analysis
April 2024 - Crypto Market Report's Analysis
ย 
VIP Model Call Girls Hinjewadi ( Pune ) Call ON 8005736733 Starting From 5K t...
VIP Model Call Girls Hinjewadi ( Pune ) Call ON 8005736733 Starting From 5K t...VIP Model Call Girls Hinjewadi ( Pune ) Call ON 8005736733 Starting From 5K t...
VIP Model Call Girls Hinjewadi ( Pune ) Call ON 8005736733 Starting From 5K t...
ย 
Call Girls Bommasandra Just Call ๐Ÿ‘— 7737669865 ๐Ÿ‘— Top Class Call Girl Service B...
Call Girls Bommasandra Just Call ๐Ÿ‘— 7737669865 ๐Ÿ‘— Top Class Call Girl Service B...Call Girls Bommasandra Just Call ๐Ÿ‘— 7737669865 ๐Ÿ‘— Top Class Call Girl Service B...
Call Girls Bommasandra Just Call ๐Ÿ‘— 7737669865 ๐Ÿ‘— Top Class Call Girl Service B...
ย 
Detecting Credit Card Fraud: A Machine Learning Approach
Detecting Credit Card Fraud: A Machine Learning ApproachDetecting Credit Card Fraud: A Machine Learning Approach
Detecting Credit Card Fraud: A Machine Learning Approach
ย 
Thane Call Girls 7091864438 Call Girls in Thane Escort service book now -
Thane Call Girls 7091864438 Call Girls in Thane Escort service book now -Thane Call Girls 7091864438 Call Girls in Thane Escort service book now -
Thane Call Girls 7091864438 Call Girls in Thane Escort service book now -
ย 
Abortion pills in Doha Qatar (+966572737505 ! Get Cytotec
Abortion pills in Doha Qatar (+966572737505 ! Get CytotecAbortion pills in Doha Qatar (+966572737505 ! Get Cytotec
Abortion pills in Doha Qatar (+966572737505 ! Get Cytotec
ย 
โžฅ๐Ÿ” 7737669865 ๐Ÿ”โ–ป Bangalore Call-girls in Women Seeking Men ๐Ÿ”Bangalore๐Ÿ” Esc...
โžฅ๐Ÿ” 7737669865 ๐Ÿ”โ–ป Bangalore Call-girls in Women Seeking Men  ๐Ÿ”Bangalore๐Ÿ”   Esc...โžฅ๐Ÿ” 7737669865 ๐Ÿ”โ–ป Bangalore Call-girls in Women Seeking Men  ๐Ÿ”Bangalore๐Ÿ”   Esc...
โžฅ๐Ÿ” 7737669865 ๐Ÿ”โ–ป Bangalore Call-girls in Women Seeking Men ๐Ÿ”Bangalore๐Ÿ” Esc...
ย 
Capstone Project on IBM Data Analytics Program
Capstone Project on IBM Data Analytics ProgramCapstone Project on IBM Data Analytics Program
Capstone Project on IBM Data Analytics Program
ย 
Call Girls In Attibele โ˜Ž 7737669865 ๐Ÿฅต Book Your One night Stand
Call Girls In Attibele โ˜Ž 7737669865 ๐Ÿฅต Book Your One night StandCall Girls In Attibele โ˜Ž 7737669865 ๐Ÿฅต Book Your One night Stand
Call Girls In Attibele โ˜Ž 7737669865 ๐Ÿฅต Book Your One night Stand
ย 
Mg Road Call Girls Service: ๐Ÿ“ 7737669865 ๐Ÿ“ High Profile Model Escorts | Banga...
Mg Road Call Girls Service: ๐Ÿ“ 7737669865 ๐Ÿ“ High Profile Model Escorts | Banga...Mg Road Call Girls Service: ๐Ÿ“ 7737669865 ๐Ÿ“ High Profile Model Escorts | Banga...
Mg Road Call Girls Service: ๐Ÿ“ 7737669865 ๐Ÿ“ High Profile Model Escorts | Banga...
ย 
Discover Why Less is More in B2B Research
Discover Why Less is More in B2B ResearchDiscover Why Less is More in B2B Research
Discover Why Less is More in B2B Research
ย 
Predicting Loan Approval: A Data Science Project
Predicting Loan Approval: A Data Science ProjectPredicting Loan Approval: A Data Science Project
Predicting Loan Approval: A Data Science Project
ย 
โžฅ๐Ÿ” 7737669865 ๐Ÿ”โ–ป mahisagar Call-girls in Women Seeking Men ๐Ÿ”mahisagar๐Ÿ” Esc...
โžฅ๐Ÿ” 7737669865 ๐Ÿ”โ–ป mahisagar Call-girls in Women Seeking Men  ๐Ÿ”mahisagar๐Ÿ”   Esc...โžฅ๐Ÿ” 7737669865 ๐Ÿ”โ–ป mahisagar Call-girls in Women Seeking Men  ๐Ÿ”mahisagar๐Ÿ”   Esc...
โžฅ๐Ÿ” 7737669865 ๐Ÿ”โ–ป mahisagar Call-girls in Women Seeking Men ๐Ÿ”mahisagar๐Ÿ” Esc...
ย 
Call me @ 9892124323 Cheap Rate Call Girls in Vashi with Real Photo 100% Secure
Call me @ 9892124323  Cheap Rate Call Girls in Vashi with Real Photo 100% SecureCall me @ 9892124323  Cheap Rate Call Girls in Vashi with Real Photo 100% Secure
Call me @ 9892124323 Cheap Rate Call Girls in Vashi with Real Photo 100% Secure
ย 

PPT - Deep Hedging OF Derivatives Using Reinforcement Learning

  • 1. Deep Hedging of Derivatives Using Reinforcement Learning Hull et al. working paper ๋ฐœํ‘œ์ž : ์œค์ง€์ƒ Graduate School of Information. Yonsei Univ. Machine Learning & Computational Finance Lab.
  • 2. 1. Introduction 2. Hedging 3. Setting Hedging model 4. Experiments 5. Conclusion INDEX
  • 4. 1. Introduction When someone conduct risk management, hedging is very common and important thing to do But theoretical hedging cannot be fitted to real-world problem exactly because of market friction
  • 5. 1. Introduction Hedging is sequential optimal control task & RL is sequential optimal control task Then can we implement RL to hedging task to reduce total hedging cost?
  • 7. 2. Hedging Hedging Short 1 call option ๐ถ๐‘‡ = max(๐‘†๐‘‡ โˆ’ ๐พ, 0) Long 1 call option Underlying asset -0.4 +0.4 +1 -3 +5 +3 +6 +0.9 - 0.9 +1.2 -1.2 +0.7 +2 -0.7 -2 P&L P&L Stock movement me
  • 8. 2. Hedging Hedging Short 1 call option ๐ถ๐‘‡ = max(๐‘†๐‘‡ โˆ’ ๐พ, 0) Long 1 call option Underlying asset -0.4 +0.4 +1 -3 +5 +3 +6 +0.9 - 0.9 +1.2 -1.2 +0.7 +2 -0.7 -2 Margin call P&L P&L Stock movement cashflow -1.4 Margin call -2 Total cashflow of naked(not hedging) position = -3.4 me
  • 9. 2. Hedging Hedging Short 1 call option ๐ถ๐‘‡ = max(๐‘†๐‘‡ โˆ’ ๐พ, 0) Long 1 call option Underlying asset -0.4 +0.4 +1 -3 +5 +3 +6 +0.9 - 0.9 +1.2 -1.2 +0.7 +2 -0.7 -2 P&L P&L Stock movement cashflow P&L from hedge +0.3 -0.8 +1.4 +0.8 +1.6 Total cashflow of hedged position = 0 me
  • 11. 2. Hedging Delta-hedging โˆ†= ๐‘ ๐‘‘1 = ๐œ•๐ถ ๐œ•๐‘† So when we take position amount of โˆ†, portfolio profit is almost zero If volatility of underlying asset is very high, or hedging period is too wide, hedge will not be effective
  • 12. 2. Hedging Delta-hedging Theoretically, CONTINUOUS Delta-hedging with NO transaction cost can make perfect hedged portfolio.
  • 13. 2. Hedging Delta-hedging Theoretically, CONTINUOUS Delta-hedging with NO transaction cost can make perfect hedged portfolio. Hedging more frequently Decrease Transaction Cost
  • 15. 3. Setting Hedging model State 1. The holding of the asset during the previous time period((๐‘– โˆ’ 1)ฮ”๐‘ก~๐‘–ฮ”๐‘ก) : ๐ป๐‘–โˆ’1 2. The asset price at time(๐‘–ฮ”๐‘ก) : ๐‘†๐‘– 3. The time to maturity : (๐‘› โˆ’ ๐‘–)ฮ”๐‘ก Action The amount of the asset to be held from time ๐‘–ฮ”๐‘ก to time (๐‘– + 1)ฮ”๐‘ก : ๐ป๐‘– State & Action โ€ข Time-step : ฮ”๐‘ก โ€ข The life of the option : ๐‘›ฮ”๐‘ก
  • 16. 3. Setting Hedging model Accounting P&L formulation ๐‘…๐‘–+1 = ๐‘‰๐‘–+1 โˆ’ ๐‘‰๐‘– + ๐ป๐‘– ๐‘†๐‘–+1 โˆ’ ๐‘†๐‘– โˆ’ ๐œ…|๐‘†๐‘–+1 ๐ป๐‘–+1 โˆ’ ๐ป๐‘– | When we derive reward function as accounting P&L formulation, reward function to minimize can be: where โ€ข ๐‘‰๐‘– : Derivatives value at time-step ๐‘–ฮ”๐‘ก โ€ข ๐‘†๐‘– : Underlying asset value at time-step ๐‘–ฮ”๐‘ก โ€ข ๐ป๐‘–: Position of underlying asset relative to position of derivatives โ€ข ๐œ… : Trading cost parameter In addition, there are an initial reward โˆ’๐œ…|๐‘†0๐ป0| and final reward โˆ’๐œ…|๐‘†๐‘›๐ป๐‘›| to set up(liquidate) the hedge position at first(last) time-step if long, positive value if short, negative value
  • 17. 3. Setting Hedging model Cash Flow Formulation ๐‘…๐‘–+1 = ๐‘†๐‘–+1 ๐ป๐‘– โˆ’ ๐ป๐‘–+1 โˆ’ ๐œ…|๐‘†๐‘–+1 ๐ป๐‘–+1 โˆ’ ๐ป๐‘– | When we derive reward function as cash flow formulation, reward function to minimize can be: where โ€ข ๐‘†๐‘– : Underlying asset value at time-step ๐‘–ฮ”๐‘ก โ€ข ๐ป๐‘–: Position of underlying asset relative to position of derivatives โ€ข ๐œ… : Trading cost parameter In addition, there are other rewards โ€ข Initial rewards : โˆ’๐‘†0๐ป0 โˆ’ ๐œ…|๐‘†0๐ป0| at first time-step โ€ข final rewards : ๐‘†๐‘›๐ป๐‘› โˆ’ ๐œ… ๐‘†๐‘›๐ป๐‘› + ๐‘๐‘Ž๐‘ฆ๐‘œ๐‘“๐‘“ ๐‘œ๐‘“ ๐‘‘๐‘’๐‘Ÿ๐‘–๐‘ฃ๐‘Ž๐‘ก๐‘–๐‘ฃ๐‘’๐‘  at last time-step if long, positive value if short, negative value
  • 18. 3. Setting Hedging model Approach Comparison ๐‘†๐‘–+1 ๐ป๐‘– โˆ’ ๐ป๐‘–+1 โˆ’ ๐œ…|๐‘†๐‘–+1 ๐ป๐‘–+1 โˆ’ ๐ป๐‘– | ๐‘‰๐‘–+1 โˆ’ ๐‘‰๐‘– + ๐ป๐‘– ๐‘†๐‘–+1 โˆ’ ๐‘†๐‘– โˆ’ ๐œ…|๐‘†๐‘–+1 ๐ป๐‘–+1 โˆ’ ๐ป๐‘– | Accounting P&L approach reward โˆ’๐œ…|๐‘†0๐ป0| โˆ’๐œ…|๐‘†๐‘›๐ป๐‘›| At time-step 1 At time-step 2~(๐‘› โˆ’ 1) At time-step ๐‘› Cash Flow approach reward At time-step 1 At time-step 2~(๐‘› โˆ’ 1) At time-step ๐‘› โˆ’๐‘†0๐ป0 โˆ’ ๐œ…|๐‘†0๐ป0| ๐‘†๐‘›๐ป๐‘› โˆ’ ๐œ… ๐‘†๐‘›๐ป๐‘› + ๐‘‰ ๐‘›
  • 19. 3. Setting Hedging model Approach Comparison ๐‘†๐‘–+1 ๐ป๐‘– โˆ’ ๐ป๐‘–+1 โˆ’ ๐œ…|๐‘†๐‘–+1 ๐ป๐‘–+1 โˆ’ ๐ป๐‘– | ๐‘‰๐‘–+1 โˆ’ ๐‘‰๐‘– + ๐ป๐‘– ๐‘†๐‘–+1 โˆ’ ๐‘†๐‘– โˆ’ ๐œ…|๐‘†๐‘–+1 ๐ป๐‘–+1 โˆ’ ๐ป๐‘– | Accounting P&L approach reward โˆ’๐œ…|๐‘†0๐ป0| โˆ’๐œ…|๐‘†๐‘›๐ป๐‘›| At time-step 1 At time-step 2~(๐‘› โˆ’ 1) At time-step ๐‘› Cash Flow approach reward At time-step 1 At time-step 2~(๐‘› โˆ’ 1) At time-step ๐‘› โˆ’๐‘†0๐ป0 โˆ’ ๐œ…|๐‘†0๐ป0| ๐‘†๐‘›๐ป๐‘› โˆ’ ๐œ… ๐‘†๐‘›๐ป๐‘› + ๐‘‰ ๐‘› When we use Accounting P&L approach reward, we should know derivatives pricing model
  • 20. 3. Setting Hedging model Approach Comparison โ€ฆ โ€ฆ Time-step Time-step reward reward ๏‚ง Accounting P&L approach rewards are almost zero-near value. โ†’ to minimize cost (reward), model just train to make rewards at every time step equal zero ๏‚ง However, Cash Flow approach rewards are not similar each other. โ†’ to minimize cost, model should learn pricing model and is hard to converge because of credit assignment problem
  • 21. 3. Setting Hedging model ๐‘Œ ๐‘ก = ๐”ผ ๐ถ๐‘ก + ๐‘ ๐”ผ ๐ถ๐‘ก 2 โˆ’ ๐”ผ ๐ถ๐‘ก 2 Model in this work ๐น ๐‘†๐‘ก, ๐‘Ž = ๐‘„1(๐‘†๐‘ก, ๐‘Ž) + ๐‘ ๐‘„2(๐‘†๐‘ก, ๐‘Ž) โˆ’ ๐‘„1 ๐‘†๐‘ก, ๐‘Ž 2 Two Q-values are introduced, ๐‘„1 estimates the expected cost for state-action combinations ๐‘„1 โ‰ˆ ๐”ผ ๐ถ๐‘ก ๐‘„1 estimates the expected value of the square of the cost for state-action combinations ๐‘„2 โ‰ˆ ๐”ผ ๐ถ๐‘ก 2 Expectation of hedging cost volatility of hedging cost Set cost equation ๐‘Œ ๐‘ก to minimize where ๐”ผ ๐ถ๐‘ก is expectation of hedging cost for time ๐‘ก ~ maturity
  • 22. 3. Setting Hedging model Model in this work Critic ๐‘„1& ๐‘„2 update with loss function: ๐‘…๐‘ก+1 + ๐›พ๐‘„1 ๐‘†๐‘ก+1, ๐œ‹ ๐‘†๐‘ก+1 โˆ’ ๐‘„1 ๐‘†๐‘ก, ๐ด๐‘ก; ๐‘ค1 2 ๐‘…๐‘ก+1 2 + ๐›พ2 ๐‘„2 ๐‘†๐‘ก+1, ๐œ‹ ๐‘†๐‘ก+1 + 2๐›พ๐‘…๐‘ก+1๐‘„1 ๐‘†๐‘ก+1, ๐œ‹ ๐‘†๐‘ก+1 โˆ’ ๐‘„2 ๐‘†๐‘ก, ๐ด๐‘ก; ๐‘ค2 2 Actor ๐œ‹ update as: ๐œƒ โ† ๐œƒ โˆ’ ๐›ผโˆ‡๐œƒ๐น(๐‘†๐‘ก, ๐œ‹ ๐‘†๐‘ก; ๐œƒ ) โˆ‡๐œƒ๐น ๐‘†๐‘ก, ๐œ‹ ๐‘†๐‘ก; ๐œƒ = โˆ‡๐œƒ๐‘„1(๐‘†๐‘ก, ๐‘Ž) + ๐‘ฉ(โˆ‡๐œƒ๐‘„2 ๐‘†๐‘ก, ๐‘Ž โˆ’ 2๐‘„1 ๐‘†๐‘ก, ๐‘Ž โˆ‡๐œƒ๐‘„1 ๐‘†๐‘ก, ๐‘Ž where ๐‘ฉ = ๐‘ 2 ๐‘„2 ๐‘†๐‘ก, ๐‘Ž โˆ’ ๐‘„1 ๐‘†๐‘ก, ๐‘Ž 2 โˆ’ 1 2 Since expected value of ๐‘„2 ๐‘†๐‘ก, ๐ด๐‘ก =expected value of ๐‘…๐‘ก+1 + ๐›พ๐‘„1 ๐‘†๐‘ก+1, ๐‘Ž 2 ,
  • 24. 4. Experiments Simulation Test I. Geometric Brownian Motion Test II. Stochastic Volatility Test
  • 25. 4. Experiments Setting โ€ข We are in short position on 1 call option of different time-to-maturity 1. 1-month 2. 3-months โ€ข Strike price of call option ๐พ = ๐‘†0 (ATM at time-step 0) โ€ข We can only use underlying stock to hedge. โ€ข Using DDPG algorithm. โ€ข Implement the prioritized experience replay method. โ€ข Using Accounting P&L approach.
  • 26. 4. Experiments I. Geometric Brownian Motion Test where ๐‘†: Stock price C: call option price q: dividend yield ๐‘…๐‘–+1 = ๐‘‰๐‘–+1 โˆ’ ๐‘‰๐‘– + ๐ป๐‘– ๐‘†๐‘–+1 โˆ’ ๐‘†๐‘– โˆ’ ๐œ…|๐‘†๐‘–+1 ๐ป๐‘–+1 โˆ’ ๐ป๐‘– | ๐น ๐‘†๐‘ก, ๐‘Ž = ๐‘„1(๐‘†๐‘ก, ๐‘Ž) + ๐‘ ๐‘„2(๐‘†๐‘ก, ๐‘Ž) โˆ’ ๐‘„1 ๐‘†๐‘ก, ๐‘Ž 2 ๐‘‘๐‘† = ๐œ‡๐‘†๐‘‘๐‘ก + ๐œŽ๐‘†๐‘‘๐‘ง ๐ถ = ๐‘†0๐‘’โˆ’๐‘ž๐‘‡ ๐‘ ๐‘‘1 โˆ’ ๐พ๐‘’โˆ’๐‘Ÿ๐‘‡ ๐‘ ๐‘‘2 ๐‘‘1 = ln ๐‘†0 ๐พ + ๐‘Ÿโˆ’๐‘ž+ ๐œŽ2 2 ๐‘‡ ๐œŽ ๐‘‡ ๐‘‘2 = ๐‘‘1 โˆ’ ๐œŽ ๐‘‡ ๐œ‡ = 5%, ๐‘Ÿ = 0, ๐‘ž = 0, ๐œŽ = 20%, ๐œ… = 1%, ๐‘ = 1.5
  • 27. 4. Experiments I. Geometric Brownian Motion Test <1-month call option> <3-months call option>
  • 28. 4. Experiments II. Stochastic Volatility Test When an option is ATM, implied volatility is approximately ๐œŽ0๐ต taking ๐œŽ0๐ต into Black-Scholes model as input ๐œŽ, we can value a call option SABER model (๐›ฝ = 1) ๐‘‘๐‘† = ๐œ‡๐‘†๐‘‘๐‘ก + ๐œŽ๐‘†๐‘‘๐‘ง1 ๐‘‘๐œŽ = ๐‘ฃ๐œŽ๐‘‘๐‘ง2 ๐”ผ ๐‘‘๐‘ง1๐‘‘๐‘ง2 = ๐œŒ๐‘‘๐‘ก where ๐‘ฃ: volatility of volatility ๐œŒ = โˆ’0.4, ๐œŽ0 = 20%, ๐‘ฃ = 60%, others = equal ๐น0 = ๐‘†0๐‘’ ๐‘Ÿโˆ’๐‘ž ๐‘‡ ๐ต = 1 + ๐œŒ๐‘ฃ๐œŽ0 4 + 2โˆ’3๐œŒ2 ๐‘ฃ 24 ๐‘‡ ๐œ™ = ๐‘ฃ ๐œŽ0 ln ๐น0 ๐พ ๐œ’ = ln 1โˆ’2๐œŒ+๐œ™2+๐œ™โˆ’๐œŒ 1โˆ’๐œŒ
  • 29. 4. Experiments II. Stochastic Volatility Test Our model is compared with 2 delta-hedging strategy 1. Bartlett Delta : Delta calculated by SABER 2. Practitioner Delta : Delta calculated by market implied volatility
  • 30. 4. Experiments II. Stochastic Volatility Test <1-month call option> <3-months call option>
  • 31. 4. Experiments a. our hedge instrument position is close to theoretical hedge position: Delta hedging b. our hedge instrument position is much less than theoretical hedge position: being under-hedging c. our hedge instrument position is much more than theoretical hedge position: being over-hedging Since transaction cost is significant, model donโ€™t take hedge position as much as model required
  • 32. 4. Experiments Since transaction cost is significant, model donโ€™t take hedge position as much as model required When 0.6 delta is required and we take 0.5 delta hedge position, model take 0.1 delta more When 0.9 delta is required and we take 0.5 delta hedge position, model take only 0.25 delta more When 0.2 delta is required and we take 0.5 delta hedge position, model take only -0.2 delta more
  • 34. 1. Use not only simulated data but real-world data 2. More well-structured architecture is needed 3. Practical hedging method like hedging vol as well as delta-hedging should be controlled by RL 4. Adaptive transaction cost can be introduced 5. Conclusion