(VCV) Value at Risk Example: Calculating Variance Covariance

Calculating Variance Covariance
(VCV) Value at Risk Example

Data – Price time series
Historical price data for Gold and WTI have been obtained for the period 1-Jun-2011 to
29-Jun-2012 from onlygold.com and eia.gov respectively. It is the time over which the
risk is to be evaluated. Figure 1 shows an extract of the daily time series data:
The Return Series
The first step for any of the VaR approaches is the determination of the return
series. This is achieved by taking the natural logarithm of the ratio of successive
prices as shown in Figure 2. AN alternative approach is to use the formula: (P1-
P0/P0)*100.
Variance Covariance – Simple Moving

Average (SMA)
Next SMA daily volatility is calculated. The formula is as follows:

Rt is the rate of return at time t. E(R) is the mean of the return distribution which
can be obtained in EXCEL by taking the average of the return series, i.e. AVERAGE
(array of return series). Sum the squared differences of Rt over E(R) across all data
points and divide the result by the number of returns in the series less one to obtain
the variance.
The square root of the result is the standard deviation or SMA volatility of the return
series. Alternately, the volatility can be calculated directly in EXCEL by using STDEV
function, applied to the return series, as shown in Figure 3:

The daily SMA volatility for Gold in Cell F18 is calculated as STDEV (array of
Gold return series). The daily SMA volatility for Gold is 1.4377 % and for WTI is
1.9856%.
VCV (SMA) daily VaR
What is the daily SMA VaR for Gold and WTI at a confidence level of 99%?

Daily VaR = Volatility or standard deviation of return
series × z- value of the inverse of the standard normal
cumulative distribution function (CDF) corresponding
with a specified confidence level.

Daily VaR for Gold calculated in Cell F16 is the product of the daily SMA volatility
(Cell F18) and the z-value of the inverse of the standard normal CDF for 99%. In
EXCEL the inverse z-score at the 99% confidence level is calculated as
NORMSINV (99%) = 2.326. Hence, daily VaR for Gold and WTI at 99%
confidence level works out to 3.3446% and 4.6192% respectively.

J-day holding SMA VaR – Scenario 1

The definition of VaR mentioned above considers three things, maximum loss,
probability and holding period.
In Basel II and Basel III a ten-day holding period is a standard assumption.
How do you incorporate the holding period into your calculations?
What is the holding SMA VaR for WTI & Gold for a holding period 10 days at a
confidence level of 99%?
Holding period VaR = Daily VaR × SQRT (holding period in days)
Where SQRT(.) is EXCEL’s square root function.
This is demonstrated for the WTI and Gold in Figure 5 below:

The 10-day holding VaR for Gold at 99% confidence level (Cell F15) is
calculated by multiplying Daily VaR (Cell F17) with the square root of the holding
period (Cell F16). This works out to be 10.5767% for Gold and 14.6073% for WTI.
J-day holding SMA VaR – Scenario 2

Let’s consider the following question:
What is the holding SMA VaR for Gold & WTI for a holding period 252
days at a confidence level of 75%?
Note that 252 days are taken to represent trading days in a year.
The methodology is the same as used before for calculating the 10-day holding
SMA VaR at a 99% confidence level, except that the confidence level and holding
period are changed.
Hence, we first determine the daily VaR at the 75% confidence level.
Recall that the daily VaR is the product of the daily SMA volatility of underlying
returns and the inverse z-score (here calculated for 75%, i.e. NORMSINV(75%)=
0.6745). The resulting daily VaR is then multiplied with the square root of 252
days to arrive at the holding VaR.
This is illustrated in Figure 6 below:

252-day holding VaR at 75% for Gold (Cell F15) is the product of the daily
VaR calculated at 75% confidence level (Cell F17) and the square root of the
holding period (Cell F16). It is 15.3940% for Gold and 21.2603% for WTI. The
daily VaR in turn is the product of the daily SMA volatility (Cell F19) and the
inverse z-score associated with the confidence level (Cell F18).

VaR Historical Simulation Approach
Ordered Returns
Unlike the VCV approach to VaR there is no assumption made about the
underlying return distribution in the historical simulation approach. VaR is based
on the actual return distribution which in turn is based on the data set used in the
calculations. The starting point for the calculation of VaR for us then is the return
series derived earlier.
We will take this return series and use it to calculate a histogram.
You can find the Histogram tool under the Data Analysis tab in Excel. If for some
reason you don’t see a data analysis tab in your version of Excel, go to Excel
Options, chose Addins and then simply add the Data Analysis Addin to enable
this tab.
To generate the Histogram select the daily return series by calculating the
percentage change in prices from one day to the next and use that series as your
input range.
Opt for New worksheet ply, cumulative percentage and chart output (see below)
to see a graphical representation of the Histogram as well as a supporting table.
When you press ok, Excel will create a new tab for you and show the histogram
that you see below. (the example refers to other time series, only for
exemplificative purposes).
Our first order of business is to reorder the series in ascending order, from
smallest return to largest return:
As a risk professional our focus is on the downside which in the histogram below
is marked at -2.77%. The upside is about 1.99+%
Daily Historical Simulation VaR
There are 270 returns in the series. At the 99% confidence level, the daily VaR
under this method equals the return corresponding to the index number
calculated as follows:
(1-confidence level)*Number of returns where the result

is rounded down to the nearest integer.
This integer represents the index number for a given return as shown below:

The return corresponding to that index number is the daily historical simulation
VaR.

The VLOOKUP function lookups the return to the corresponding index value from
the order return data set. Note that the formula takes the absolute value of the
result. For example at the 99% confidence level the integer number works out to
2. For Gold this corresponds with the return of -5.5384% or 5.5384% in absolute
terms, i.e. there is a 1% chance that the price of Gold will fall by more than
5.5384% over a holding period of 1 day.
10-day holding Historical Simulation VaR

As for the VCV approach the holding VaR is equal to the daily VaR times the
square root of the holding period. For Gold this works out to 5.5384%*SQRT(10)
= 17.5139%.
Amount of worst case loss

So what is the amount of worst case loss for Gold over a 10-day holding period
that will only be exceeded 1 day in 100 days (i.e. 99% confidence level)
calculated using the Historical Simulation approach?
Worst Case Loss for Gold @ 99% confidence level over a 10-day holding period
= Market Value of Gold * 10-day VaR% = (1598.50*100)* 17.5139%= USD
27,996.
There is a 1% chance that the value of Gold in the portfolio will lose an amount
greater than USD 27,996 over a holding period of 10-days.

(VCV) Value at Risk Example: Calculating Variance Covariance

Uploaded by

Copyright:

Available Formats

(VCV) Value at Risk Example: Calculating Variance Covariance

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

(VCV) Value at Risk Example: Calculating Variance Covariance

Uploaded by

Copyright:

Available Formats

Calculating Variance Covariance

(VCV) Value at Risk Example

Variance Covariance – Simple Moving

J-day holding SMA VaR – Scenario 1

In Basel II and Basel III a ten-day holding period is a standard assumption.

How do you incorporate the holding period into your calculations?

Holding period VaR = Daily VaR × SQRT (holding period in days)

Where SQRT(.) is EXCEL’s square root function.

This is demonstrated for the WTI and Gold in Figure 5 below:

J-day holding SMA VaR – Scenario 2

This is illustrated in Figure 6 below:

We will take this return series and use it to calculate a histogram.

(1-confidence level)*Number of returns where the result

10-day holding Historical Simulation VaR

Amount of worst case loss

You might also like