375 reputation
211
bio website
location
age
visits member for 1 year, 5 months
seen 22 hours ago

"The same past data can confirm a theory and its exact opposite! If you survive until tomorrow, it could mean that either a) you are more likely to be immortal or b) that you are closer to death."


May
14
comment Calculate alpha (CAPM) in “cross country-portfolio”
Well, I was using simple returns, not exponential.
Apr
29
comment Stochastic Calculus in Quantitative analysis
You can also look in bjorks book Arbitrage Theory in continous time which is relatively comprehensive and not too technical. On the downside he is not very consistent with his notations inbetween chapters (it is as they are written at different times throughout his career and then mashed together).
Apr
29
comment Why are options called what they are called?
Just a remark: I dont think he would ask why they were called put and call if HE thought it was obvious.
Apr
27
comment Backtesting Period
I'm not entirely sure how your backtest procedure will look like, but I think this article may be of interest for you: papers.ssrn.com/sol3/papers.cfm?abstract_id=2308659 They talk about minimal backtest length in a sense to avoid overfitting, which is not entirely what you're asking for. It still might give you some ideas on what period to choose from that point of view.
Apr
17
comment Engle Granger test returns a 0 in matlab, while correlation factor is .80+. Am I doing something wrong?
To answer your question. Correlation and cointegration is basically not the same thing. I wish I could answer your other question as well on Johansens (I think that was the test you used), unfortunately I don't know that test well enough to ensure my credibility.
Apr
16
comment Engle Granger test returns a 0 in matlab, while correlation factor is .80+. Am I doing something wrong?
I advice you to read a book on time series analysis. It is strongly recommended to know what you're doing in statistics.
Apr
16
comment Significance of Data
Is the question to recommend a book or to help you interpret the results? In case of the latter, then you should consider giving more details of the problem. It looks like you might be testing for stationarity of the cointegrated process (pair). It's very unclear though...
Apr
13
comment Pricing of a simple contingent claim
Hmm, it might be necessary in some cases. It worked nicely now since you took the power of an exponential, so you could just multiply the $\beta$ into the exponent and then the expectation became very easy. If this is not the case you should try to find the dynamics "of the whole claim" instead.
Apr
12
comment CVaR/VaR Ratio as alpha goes to 1
Well, is that the expression you really want to compute the limit of? If the denominator is the CVaR then you have the wrong expression I believe. It should be something like $\phi(\Phi^{-1}(p))/p$.
Apr
10
comment Divergent or Convergent Strategies? Which is the way to go?
Of course, having multiple strategies is certainly better. That requires a lot of recourses thought, in particular human capital.
Apr
9
comment Divergent or Convergent Strategies? Which is the way to go?
Well Ok, obviously I can't really argue against it since I haven't seen the figures. What you mention last is more to my point though. So the weaker funds do niche strategies that for some period of time makes a lot of money. But then, as a result of some rare event, they lose a lot of money. In other words they have a very fat left tail. This is typical characteristic of the convergent strategy. My point is now that even though it is a niche strategy, it must not be convergent. If they instead deployed a divergent strategy they could survive these rare events.
Apr
9
comment Divergent or Convergent Strategies? Which is the way to go?
I know, after googling the terms I can see that the notions are not widely applied in finance. Nonetheless, I think it is a useful way of thinking. Of course you can classify in terms equity, FX, commodities and what not. It doesn't tell you much more than what you're actually investing in. With regards to your last paragraph, that is not the CLT. Frankly, I think it's both naive and ludicrous to think that the aggregated returns of a hedge fund would be normally distributed just because they have multiple strategies (or in any case).
Mar
29
comment Options with a stochastic strike
Well, I wouldn't say that the strike is stochastic, because that is not entirely true. I mean the strike is always K, but the whole claim is either 0 or some value X depending on what happens up until maturity. On the other hand, one could see the strike as either K or $\pm \infty$ (depending on if it is a call or put).
Mar
28
comment Options with a stochastic strike
I think one reason for having Asian Options, and other similar instruments, is that players with large positions in vanilla options have great incentive to "manipulate" share prices in order to avoid great losses or being able to exercise one's options, even if it means incurring losses from selling off/buying lots of shares in the underlying. With Asian options, however, this becomes much more difficult.
Feb
21
comment Finding the dynamics of a dividend paying asset under arbitrary numeraire
Ok, I just realized that the dividends process in domestic currency that I use is not correct. $D \cdot E$ is just the accumulated dividends process in foreign currency, evaluated in domestic currency. This is not really what I wanted. The correct dynamics are $dD_d = \delta S_d dt = \delta E \cdot S dt$, i.e. the dividends process received in domestic currency at each infinitesimal time point. Then the equations will solve nicely. I guess it is not possible to find an EMM under those conditions I had firstly.
Feb
13
comment Lower bound of ITM Calls when computing Implied Volatility
Yes, that appeared to be the case. For deep ITM calls the spread was wide so even the average bid-ask was below the lower bound. Thank you
Feb
13
comment Lower bound of ITM Calls when computing Implied Volatility
Oh of course. The lower bound was kind of obvious. I'm surprised I overlooked that. Anyway, you made some other good points to take into consideration. Thank you!
Dec
25
comment What are the general limitations of Gaussian copulas with regards to the range of joint pdf's it can approximate?
I'm not familiar with the term nataf transformation. But as previous commentator said, the gaussian copula has neither lower nor upper tail dependence. To clarify, copulas are used to model dependence between r.v's, in particular different copulas has different tail characteristics. For instance, t-copulas has both lower and upper tail dependence, and clayton copulas only has lower tail dependence. I removed my earlier comment on this because I wasn't sure if this is what you were asking for.
Dec
6
comment Choosing a weak learner
I believe this question is better suited for Stack Overflow.
Nov
22
comment Optimizing stochastic functions numerically
Thank you for your comment. It looks interesting. I wanted to try it out, but I could get it to install properly, gotta give it another try later.