Historical VaR for Futures and Options Portfolios
Summary
The discussion outlines a historical Value at Risk approach for a portfolio of futures and options. It proposes applying past instrument returns to current positions to estimate historical dollar profit and loss, then selecting a lower-tail percentile as the VaR estimate. Because derivatives provide leveraged exposure, the answer says to use each contract’s notional value rather than the cash deposited or initial margin to scale market risk.
The response suggests calculating risk by position and summing the resulting dollar VaRs. That shortcut has an important limitation: adding individual VaRs generally does not produce portfolio VaR, because joint outcomes and correlations affect portfolio losses. Historical portfolio VaR is more commonly estimated by applying each historical market move to all positions on the same date, then ranking the combined portfolio outcomes. The exchange provides no worked calculation or validation, and it does not explain how to model options’ nonlinear price sensitivity or choose holding periods and data windows.
Key ideas
- Historical VaR estimates a loss threshold from ranked past profit and loss outcomes.
- Derivative exposure should be scaled using contract notionals rather than the cash margin deposit.
- Portfolio VaR depends on how positions move together, so adding separate position VaRs can misstate total risk.
- Options may require repricing under historical market moves because their value does not change linearly with the underlying.
Tags
Full text
# Value at Risk for Portfolio of Futures # Value at Risk for Portfolio of Futures I'm working in a very small commodity trading company. They are not used to excel at all, so i built their trading sheet to follow open positions & past positions. Now they asked me to calculate the VaR of the total portfolio. I have never done that, so I read a lot of ressources on the internet, especially topics on here, but none of them really explain on how to do it precisely. I'm calculating historical VaR The portfolio is made of futures & options They dont roll the futures. They keep the positions for a relatively short time (2-3 weeks) We have deposited 20k$ at the bank So I tried doing it: I choose the number of windows (1 year) so 250 cases I collected historical return of futures on t-1 I applied that return on t-1 to today futures position and record dollar P&L This is my first question. What does it mean to today's futures position ? I saw someone on youtube multiply it by the value of my portfolio (20k$) ? I repeat but change t-1 to t-2 and so on until I have 250 cases I then choose my significance level (5%) I rank those 250 P&L and get 5%th lowest percentile (top 5% most negative P&L) But the result is only for one futures position, how do i add different VaR to have the total VaR of the porfolio ? Because i have futures & options, and I don't know how to calculate the weight of each positions. Some people say to use the initial margin that i keep at the broker, but i wanted to be sure. Thank you in advance for your help ## Answer by SuavestArt (score 2, accepted) https://quant.stackexchange.com/a/77101 Futures and options are financial derivatives, and one key feature is their inherent leverage, which means you don't need to have the full cash equivalent to finance the market risk they represent. You'll need to know the contract size of each instrument so as to get their notional value. In the context of historical VaR calculations, estimating correlations is not a concern. You can calculate VaR for each position by using the time series of returns and the notional value of each position. Then, add up these results to determine the portfolio VaR in dollar value. As mentioned earlier, expressing VaR in percentage terms doesn't provide meaningful insights for a portfolio of derivatives.
Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.