Value at Risk
The objective of this presentation is to introduce
an alternative way of looking at risk.
Value at Risk (VaR)
• Definition:
– Value at Risk is an estimate of the worst possible loss an investment
could realize over a given time horizon, under normal market conditions
(defined by a given level of confidence).
• VaR answers the question: 'What is the maximum loss we can expect with X
% confidence over a given time period (T)?’
Example: We are 99% confident that it will not lose more than $1 million in
the next 10 days.
• So, the VaR informs us how much (in dollars, in dirhams or in any other
currency) will be lost at maximum if a very negative and unexpected event
takes place with a given probability.
• VaR can measure the risk of many types of financial securities (i.e., stocks,
bonds, commodities, foreign exchange, off-balance-sheet derivatives such as
futures, forwards, swaps, and options, and etc.)
Value at Risk (VaR)
• Key Components of VaR:
1) Confidence Level: Typically 90%, 95% or 99%, which indicates
how sure we are about the estimation.
For example, at 95%, we are 95% confident that the loss will not
exceed a certain amount.
2) Time Horizon: The period over which the risk is measured, such as
one day, one week, or one month.
3) Loss Amount: The dollar value or percentage of the investment
expected to be lost.
VaR and Regulatory Capital
• How is VaR used?
Regulatory agencies (e.g., Basel Accords) use VaR to
determine how much capital banks should hold.
Example: If a bank's 10-day 99% VaR is $50 million, it
must hold at least 3 times that amount ($150 million)
as market risk capital.
For credit risk, Basel II requires banks to hold capital
based on a 1-year 99.9% VaR.
Advantages of VaR
• Advantages of VaR:
Provides a clear, single number for senior management summarizing
the total risk in a portfolio of assets.
Value at Risk is a dollar value risk measure, as opposed to the other
measurements of risk in the financial industry such as: beta and
standard deviation.
Useful for regulators, risk managers, and traders
Answers the question: 'How bad can things get?'
• Limitations:
Does not account for extreme losses (tail risk)
it relies on historical data and the assumption that past
performance is indicative of future risks, which may not always hold
true.
Calculation of VaR
• VaR Calculation Methods:
1. Parametric (Variance-Covariance): Based on
normal distribution.
2. Historical Simulation: Uses historical market
data.
3. Monte Carlo Simulation: Models random
scenarios.
First: VAR using Variance-Covariance approach
Example 1: what is the 99% VAR for a Portfolio with a mean of $0 and standard
deviation of $10 million.
VAR = 0 – 2.33* 10 = 99% VaR is -2.33 × 10 = -$23.3 million.
Example 2: What is the 99% VAR of a portfolio that has a mean gain of $2 million
and a standard deviation of $10 million over 6 months.
VaR = $2 - 2.33 × $10 = -$21.3 million
• Conclusion: This means there is a 1% chance the portfolio will lose more than
$21.3 million in the next 6 months.
Practice Question 1
• A portfolio has an expected return of 5% per year
and a standard deviation of 8% per year. The
portfolio’s current value is $1,000,000.
• Calculate the 1-year Value at Risk (VaR) for this
portfolio at the following confidence levels:
– 90% confidence
– 95% confidence
– 99% confidence
• Interpret the results and explain what the VaR at
each confidence level indicates.
Solution
1) VaR at 90% Confidence Level:
VaR=(0.05−1.645*0.08)*1,000,000
VaR = VaR=− 81,600
We are 90% confident that the portfolio’s potential loss will not exceed $81,600
over the next year.
OR: There’s a 10% chance that the portfolio could lose more than $81,600 over
the next year.
2) VaR at 95% Confidence Level:
VaR=(0.05−1.96*0.08)*1,000,000
VaR=−106,800
3) VaR at 99% Confidence Level:
VaR=(0.05−2.33*0.08)*1,000,000
VaR=−136,400
Practice Question 2
A portfolio has an expected daily return of 0.1%
and a daily standard deviation of 1.5%. The
portfolio’s current value is $2,000,000.
1) Calculate the 1-day Value at Risk (VaR) for this
portfolio at the following confidence levels:
– 90% confidence
– 95% confidence
– 99% confidence
2) Interpret the results and explain what the VaR at
each confidence level indicates.
3) Calculate the 5 days var using 99% level.
Solution
• VaR at 90% Confidence Level:
VaR=(0.001−1.645*0.015)*2,000,000
VaR = VaR=−47,350
We are 90% confident that the portfolio’s potential loss will not exceed $47,350
for one day.
OR, There’s a 10% chance that the portfolio could lose more than $47,350 over
the next day.
• VaR at 95% Confidence Level:
VaR=(0.001−1.96*0.015)*2,000,000
VaR=−56,800
• VaR at 99% Confidence Level:
VaR=(0.001−2.33*0.015)*2,000,000
VaR=−67,900
Solution
3) To calculate the 5-day VaR, we need to adjust
the standard deviation by multiplying it by the
square root of the time period (in this case, 5
days), while keeping the expected return the
same.
• VaR at 99% Confidence Level:
VaR=(0.001−2.33*0.015*(5)^0.5)*2,000,000
VaR=−154,301
Practice Question 3
• The monthly returns for a portfolio are normally
distributed with a mean of 15% and standard deviation of
10%. Assume that 40 000 000 dirhams is invested.
Calculate the following using a 99% confidence level.
a) VaR for one month.
b) VaR for 7 months.
a) VAR = (0.15−2.33*0.10)* 40,000,000 = −3,320,000
b) VAR = (0.15−2.33*0.10*(7) ^0.5)* 40,000,000 =
−18,658,402
Practice Question 4
• The returns for a portfolio are normally
distributed with a mean of 12% and standard
deviation of 8%. Assume that 15 000 000
dirhams is invested. Calculate the following
using a 95% confidence level. The frequency of
the data is on a weekly basis.
•
• VaR for one week.
• VaR for 5 weeks.
Solution
a) VAR = (0.00230−1.96*0.011178)* 15,000,000 = −294,210
b) VAR (0.00230−1.96*0.011178*(5)^0.5)* 15,000,000 = −700,455
Expected Shortfall (ES)
• Expected Shortfall (ES), also known as Conditional Value
at Risk (CVaR), is a risk measure that quantifies the
expected loss in the tail of the distribution beyond the
Value at Risk (VaR) level.
• Unlike VaR, which only considers the magnitude of losses
at a and below a certain confidence level, ES takes into
account the magnitude of losses that exceed the VaR level
and provides a more comprehensive measure of risk.
• Mathematically, ES can be expressed as the expected
value of all losses that exceed the VaR at a certain
confidence level.
Normal Distribution Assumption
• Normal Distribution Assumption in VaR:
VaR is often calculated assuming returns follow
a normal distribution.
This is true in some cases but not always
accurate for financial markets.
• Limitations:
• Fat tails or skewed distributions can cause
underestimation of extreme risks.