I need some help with an economics paper dealing with VALUE AT RISK.?
Does anyone know what the implications for an institution using VAR?
- JoeyVLv 710 years agoFavorite Answer
1) VaR (often) assumes that losses are distributed according to a Normal distn. In many situations that is clearly not the case as the recent financial meltdown made pretty clear. If losses are skewed or long-tailed then this is probably not captured by VaR.
2) VaR uses backward looking data to estimate forward volatility. For instance suppose that you own a stock of a biotech company that releases clinical trials data tomorrow but are estimating VaR using past vol on this stock. Your VaR estimate will show little risk, but tomorrow's announcement will make or break the company.
3) VaR depends on a methodology used to derive it and many different methods might produce different VaR numbers. How much would you trust a risk estimate if another plausible derivation returned a much different number?
4) VaR relies on stable correlations between assets, but correlations often depend on magnitude of asset price changes.
5) VaR says nothing about what happens after you reach the VaR limit. Thus, two companies may have a 95% VaR of 2% but given that they exceed the VaR limit of 2% the expected loss for company A might be 3% but for company B it might be 25%. VaR would suggest they are equally risky.
6) VaR almost never accounts for market friction, liquidity problems, operational risk, etc. that can be highly important in determining the risk of a pile of assets.
BTW - As a hedge fund risk manager for 10 years, I calculated VaR on every portfolio every day. It's an excellent measure of something like exposure under stable market conditions. More often than not I would like the VaR to be higher (because it means we are productively using capital) rather than lower. Risk problems tend to come from things like partial VaR where the contribution to VaR from an asset or asset class is too high. If VaR is your only measure of risk, then you would be an idiot and not much of a risk manager but it's still useful.
- John WLv 710 years ago
The ill fated Long Term Capital Management fund and the Eifuku fund used Value At Risk extensively. The problem with VaR is that it's a reporting mechanism nothing more, it doesn't provide you with any guidance as to what an acceptable risk is. The use of VaR relies on your intuition to avoid excessive risk. Many economists argue that only human intuition can properly evaluate the risk but if you optimize for the log utility of wealth thereby using geometric/logarithmic evaluations of risk (The Kelly Criterion) then at least you have a reasonable limit for acceptable risk. The book "Fortune's Formula" by William Poundstone has accounts of institutions that used VaR.
- Anonymous10 years ago
It was not until the crash of 2008 that it was widely recognised that the efficient market hypothesis was no more than a statement of a general tendency, and that major risks could occasionally arise from statistically unpredictable patterns of investor conduct, and it was not until then that it was generally realised that policies that were beneficial to individual investors could be harmful if they were adopted by all of them. Professor Shin of Princeton University accepts that the resulting shortcomings of economic theory had played a big rôle in that disaster, and reports that the "race is on" to fill the gaps in economic theory and to add a new perspective to macroeconomics by the incorporation into it of a new theory of financial economics. It consensus had already formed in the course of 2009 that internationally agreed "financial regulation" had become necessary, two G20 summit meetings of the leaders the major developed and developing countries had been held for that purpose , and a variety of concrete proposals had been formulated.