Help me out on this one… Regarding the practical application of value at risk (VAR) for portfolio managers, which of the following statements is FALSE? VAR can: A) be used to set risk limits on an absolute level. B) be used to identify the macroeconomic factors that have the greatest impact on overall portfolio performance. C) not be used to set risk limits relative to a benchmark. …
Is it B?
Yeah, feel pretty good about B two changes later.
Nope, it’s C
Let’s make this a two-for: QUESTION 2 Which of the following factors is the common weakness in historical and Monte Carlo Simulation approach to VAR estimation? A) Both assume that historical variance-covariance matrix is stable. B) For some assets you may face model risk. C) A lot of data is needed for time period of interest.
Nope, going A for #2.
Post the answers tonight if you get a chance while I’m thinking about it.
I remembered that historical doesn’t use a model for #2, so that fixed my choice at A rather than B. Also, VaR assumptions do include stable covariance matrices which is why that is a key stress test in the process. For #1, with C I was initially thinking of absolute financial institution and portfolio VaR, but then I remembered that benchmark VaR could also be used to provide relative risk management data with detailed breakouts. Two different VaR methods. C was false. Pylon, aren’t you east coast and employed? What are you doing up?
- A 2. C
Black Swan Wrote: ------------------------------------------------------- > > Pylon, aren’t you east coast and employed? What > are you doing up? Yes. I am employed and on the East coast (and I am grateful for both as well, actually). I have developed a really unhealthy study habit of falling asleep with my son room after reading him a book (around 8 PM). I’m just usually so exhausted after work, I can’t help it. Then I wake up in the middle of the night, refreshed and ready to study. So tonight I slept from 8:30 PM to about 12:30 AM, and have now been up for 2 1/2 hours hitting the books. But soon to bed, as my alarm clock will go off at 5:30AM. Anyway… here are the answers (which I think suck, but more on that tomorrow): QUESTION 1 Regarding the practical application of value at risk (VAR) for portfolio managers, which of the following statements is FALSE? VAR can: A) be used to set risk limits on an absolute level. B) be used to identify the macroeconomic factors that have the greatest impact on overall portfolio performance. C) not be used to set risk limits relative to a benchmark. The correct answer was C) not be used to set risk limits relative to a benchmark. VAR can be used to set risk limits for a portfolio – either on an absolute level or on a relative basis versus a benchmark. QUESTION 2 Which of the following factors is the common weakness in historical and Monte Carlo Simulation approach to VAR estimation? A) Both assume that historical variance-covariance matrix is stable. B) For some assets you may face model risk. C) A lot of data is needed for time period of interest. Your answer: A was incorrect. The correct answer was B) For some assets you may face model risk. The historical method uses actual returns for the position in question. An advantage of the historical method is not having to assume any particular distribution. A disadvantage is that it assumes past performance is representative of what can occur in the future, which may not be the case. The Monte Carlo simulation method for calculating VAR usually involves generating random numbers with a computer. The generated numbers represent possible returns of the asset or portfolio. An advantage is that Monte Carlo simulation does not require the normality assumption and can accommodate the required assumptions for complex relationships. A disadvantage is the requirement for many managerial assumptions and a great deal of computer time and calculations. The historical method and Monte Carlo Simulation both suffer from modeling risk. ------- OK… here’s why I disagree with Schweser… for question 1 I don’t really know how you would use VAR to set risk limits relative to a benchmark. Would you calculate the VAR of the benchmark and then set a limit as a percetage of that? If so, I don’t remember that being covered in the curriculum. So I went with B. I certainly don’t see how VAR identifies macro-economic risk factors… I thought that was for factor push analysis (or whatever they call it now). For question 2, I think A is a good answer (actually, I think it’s the best answer). Even though the historical approach does not assume normality, the results it gives are still dependent upon a variance / covariance matrix – just one that isn’t calculated and may not even be calculable. It doesn’t matter – the returns are still dependent upon that matirx and it’s stability over time (I’m assuming we all agree that the Monte Carlo results also depend on matrix stability). Anyway, I’ll be curious to read your thoughts when I get to work in four hours. Good night for now.
- B 2. A (definitely got to be correct on this one)
I am used to VaR being treated as a top level absolute measure as well. But then I remembered seeing relative benchmark VaR explained as a method of breaking out variance contributions such as the information ratio. I believe I read this in L2 pm, but it could have been elsewhere. For #2, I agree with your point 110%.
1C. I like it because you can set the SD relative to the benchmark and calculate VAR as relative to the benchmark and of course the normal way as an absolute measure. In fact, I know some institutions who do it. 2A. I don’t get B. ‘For some assets’ is not true, you will face model risk for all assets. With A, you do play around with these in a monte-carlo (if you want) so it’s not true. I stick to A for 2.
I missed the word “limits” in question 1 for A and since VAR can’t be used to create absolute limits on risk, so I thought A is false (thus the correct answer). I thought B to be correct because you can use VAR to test the sensitivity of various underlying factors, you can run VAR analysis for each of the factors and see which one has the largest VAR at a given probability. I believe given this “C” should be the answer as you can conduct VAR for benchmark and porfolio and compare or you might also do VAR on the returns generated above the benchmark and then set risk limits. For second one, I believe A is an advantage they have over the variance-covariance method, because they don’t have any such underlying assumptions, MC is just random number generation and historical is just a crude measure on the historical data, I didn’t choose B because it was very general and all the models (even variance covariance) suffer that risk; I chose C as historical needs a lot of data and MC also uses a lot of data for analysis. I hope we don’t make such mistakes on 6th of June.
For #2, is Schweser refering to the inappropriate application of model that results in model risk (i.e. using BSM (instead of Binomial) to value an american call option etc)? Another question on model risk - so is the Analytical VAR method (aka variance-covariance method) also affected by model risk?
Black Swan Wrote: ------------------------------------------------------- > I am used to VaR being treated as a top level > absolute measure as well. But then I remembered > seeing relative benchmark VaR explained as a > method of breaking out variance contributions such > as the information ratio. I believe I read this > in L2 pm, but it could have been elsewhere. > > For #2, I agree with your point 110%. I don’t believe VAR was in the L2 curriculum. I don’t remember anything like this or even how it would work.
mcpass Wrote: ------------------------------------------------------- > 1C. I like it because you can set the SD relative > to the benchmark and calculate VAR as relative to > the benchmark and of course the normal way as an > absolute measure. In fact, I know some > institutions who do it. > I don’t remember this from the curriculum at all – do you? Can you explain this method a little better? What do you mean by “set the SD relative to the b’mark”??? Do you mean that if the assets under question have an E® = benchmark and a SD < b’mark, then there will be negative VAR? Again… not in the curriculum was it? So you’ve made a weak case that “C” might be the right answer, but done nothing to tell me why the answer “B” would be wrong. I’m not persuaded. > 2A. I don’t get B. ‘For some assets’ is not true, > you will face model risk for all assets. With A, > you do play around with these in a monte-carlo (if > you want) so it’s not true. > > I stick to A for 2. Agreed…
plyon Wrote: >but done nothing to tell me why the >answer “B” would be wrong. I’m not persuaded. I agree with you. What they describe in B is closer to scenario analysis (or stress testing). By the way both questions are too ambiguous. I wouldn’t lose sleep over them.
> I wouldn’t lose sleep over them. Ironic, given that I was studying at 3 AM.
I dont think historical variance-covariance matrix is relevant in the historical VAR method, correct?