Introduction to
Value at Risk
1
Toshifumi Kuga CEO of TOSHI STATS SDN. BHD
Ver1.0
Needs for Value at Risk
• How much market risk is the company taking?
• Is the list of company s positions by products good enough to
capture risk profile?
• Maximum potential loss with likelihood is needed
• Market risk can be defined based on probability
• Definition of Value at Risk (VaR) : The maximum likely loss over
some target period
2
Advantage of Value at Risk
• Common consistent measure of risk
• VaR enable us aggregate the risks of sub-positions into firm wide
portfolio
• Consideration on Interdependence and correlations among risk
factors
• VaR is probabilistic
• VaR is expressed as lost money
3
Three key questions for VaR
1. Holding period
• How can we set time horizon?
• Depend on the liquidity of markets
• Depend on the investment horizon
• Depend on regulations
4
Three key questions for VaR
2. Confidence interval
• How can we say 99% ?
• What assumption is used?
3. Observation period for data
• Why is it important to understand VaR?
• Are market events included in observation period?
5
Two key components of VaR
1. Exposures
• Exposures come from each product/inventory in the
company
• Exposures are captured based on risk factors such as
interest rates, stock price and foreign currencies
• Exposures are calculated by front office system (with
pricing models) and fed to VaR system
6
Two key components of VaR
2. Variance/Covariance on returns
• Created based on market movement/trends in the past
• Historical time series data of markets during observation
period is used
• If observation periods are different, Variance/Covariance on
returns are also different even though kind of exposures are
same in the portfolio
7
Exposures in detail
• Where does market risk come from?
• Example: if Japanese investors have US equities, what risk
factors do they have?
• Answer : two risk factors 1. US stocks 2. JPY/USD rate
• Exposure: Sensitivities to risk factors which the products in
the position/inventory have
8
Exposures in detail
• Examples of Exposures
• Exposure of Equity
• Exposure of Fixed Income
• Exposure of Foreign currencies
• Delta,Gamma and Vega are often used for derivative products
9
Variance-Covariance approach
• Normal distribution is usually assumed for risk factors movement
• Volatilities
• Definition : the annualized standard deviation of return
• Historical Volatilities : Assumed that volatility is constant
• Covariance/Correlation
• It defines inter-dependence among risk factors movement
• Noted : Volatility and Covariance/Correlation tends to change over time in reality
10
Calculation method for VaR
• Historical method
• No statistical assumption used about distribution of risk factors
movement
• Variance-Covariance method
• Normal distribution is usually assumed
• Monte Carlo Simulation
• Normal distribution is usually assumed
11
Advantage of Historical method
• No need for assumptions about probability distributions of
risk factors movement
• Good to capture of tail risk, which rarely happens
• Easier to implement VaR model than other methods because
it is simple
12
Limitaions of VaR
• Is VaR a perfect KRI to monitor risk of portfolios?
• What will happen beyond the confidence level?
• VaR tells us the most we can lose if tail events do not occur (Ex. the most
we can lose over 99% of time)
• VaR itself gives us no indication of how much the loss might be if tail
events occur
• Stress tests are needed to capture the risk which can not be identified by
VaR
13
Back Testing
• Comparison between the estimated VaR and the actual
results of P/L of a portfolio
• A relative large number of extreme observations indicates
that our risk measures are probably too low
• Back Testing are required by regulators in many countries.
14
Thanks for your attentions
!
• TOSHI STATS SDN. BHD, Digital-learning center for statistical computing in Asia
• CEO : Toshifumi Kuga, Certified financial services auditor
• Company web site : www.toshistats.net
• Company FB page : www.facebook.com/toshistatsco
• Please do not hesitate to send your opinion and massages about our courses to
us !
15
Disclaimer
• TOSHI STATS SDN. BHD. and I do not accept any responsibility or
liability for loss or damage occasioned to any person or property
through using materials, instructions, methods or ideas contained
herein, or acting or refraining from acting as a result of such use.
TOSHI STATS SDN. BHD. and I expressly disclaim all implied
warranties, including merchantability or fitness for any particular
purpose. There will be no duty on TOSHI STATS SDN. BHD. and me
to correct any errors or defects in the codes and the software.
© 2014 TOSHI STATS SDN. BHD. All rights reserved
16

Introduction to VaR

  • 1.
    Introduction to Value atRisk 1 Toshifumi Kuga CEO of TOSHI STATS SDN. BHD Ver1.0
  • 2.
    Needs for Valueat Risk • How much market risk is the company taking? • Is the list of company s positions by products good enough to capture risk profile? • Maximum potential loss with likelihood is needed • Market risk can be defined based on probability • Definition of Value at Risk (VaR) : The maximum likely loss over some target period 2
  • 3.
    Advantage of Valueat Risk • Common consistent measure of risk • VaR enable us aggregate the risks of sub-positions into firm wide portfolio • Consideration on Interdependence and correlations among risk factors • VaR is probabilistic • VaR is expressed as lost money 3
  • 4.
    Three key questionsfor VaR 1. Holding period • How can we set time horizon? • Depend on the liquidity of markets • Depend on the investment horizon • Depend on regulations 4
  • 5.
    Three key questionsfor VaR 2. Confidence interval • How can we say 99% ? • What assumption is used? 3. Observation period for data • Why is it important to understand VaR? • Are market events included in observation period? 5
  • 6.
    Two key componentsof VaR 1. Exposures • Exposures come from each product/inventory in the company • Exposures are captured based on risk factors such as interest rates, stock price and foreign currencies • Exposures are calculated by front office system (with pricing models) and fed to VaR system 6
  • 7.
    Two key componentsof VaR 2. Variance/Covariance on returns • Created based on market movement/trends in the past • Historical time series data of markets during observation period is used • If observation periods are different, Variance/Covariance on returns are also different even though kind of exposures are same in the portfolio 7
  • 8.
    Exposures in detail •Where does market risk come from? • Example: if Japanese investors have US equities, what risk factors do they have? • Answer : two risk factors 1. US stocks 2. JPY/USD rate • Exposure: Sensitivities to risk factors which the products in the position/inventory have 8
  • 9.
    Exposures in detail •Examples of Exposures • Exposure of Equity • Exposure of Fixed Income • Exposure of Foreign currencies • Delta,Gamma and Vega are often used for derivative products 9
  • 10.
    Variance-Covariance approach • Normaldistribution is usually assumed for risk factors movement • Volatilities • Definition : the annualized standard deviation of return • Historical Volatilities : Assumed that volatility is constant • Covariance/Correlation • It defines inter-dependence among risk factors movement • Noted : Volatility and Covariance/Correlation tends to change over time in reality 10
  • 11.
    Calculation method forVaR • Historical method • No statistical assumption used about distribution of risk factors movement • Variance-Covariance method • Normal distribution is usually assumed • Monte Carlo Simulation • Normal distribution is usually assumed 11
  • 12.
    Advantage of Historicalmethod • No need for assumptions about probability distributions of risk factors movement • Good to capture of tail risk, which rarely happens • Easier to implement VaR model than other methods because it is simple 12
  • 13.
    Limitaions of VaR •Is VaR a perfect KRI to monitor risk of portfolios? • What will happen beyond the confidence level? • VaR tells us the most we can lose if tail events do not occur (Ex. the most we can lose over 99% of time) • VaR itself gives us no indication of how much the loss might be if tail events occur • Stress tests are needed to capture the risk which can not be identified by VaR 13
  • 14.
    Back Testing • Comparisonbetween the estimated VaR and the actual results of P/L of a portfolio • A relative large number of extreme observations indicates that our risk measures are probably too low • Back Testing are required by regulators in many countries. 14
  • 15.
    Thanks for yourattentions ! • TOSHI STATS SDN. BHD, Digital-learning center for statistical computing in Asia • CEO : Toshifumi Kuga, Certified financial services auditor • Company web site : www.toshistats.net • Company FB page : www.facebook.com/toshistatsco • Please do not hesitate to send your opinion and massages about our courses to us ! 15
  • 16.
    Disclaimer • TOSHI STATSSDN. BHD. and I do not accept any responsibility or liability for loss or damage occasioned to any person or property through using materials, instructions, methods or ideas contained herein, or acting or refraining from acting as a result of such use. TOSHI STATS SDN. BHD. and I expressly disclaim all implied warranties, including merchantability or fitness for any particular purpose. There will be no duty on TOSHI STATS SDN. BHD. and me to correct any errors or defects in the codes and the software. © 2014 TOSHI STATS SDN. BHD. All rights reserved 16