207 reputation
15
bio website
location
age
visits member for 2 years
seen Feb 4 at 15:24

Apr
19
comment Is there any thing out there as a substitute for KDB?
The minimal 2-core setup of kdb+ is actually pretty reasonable: $25K per year including maintenance.
Aug
5
comment Analyzing an incomplete set of trades
That's an interesting approach. I hadn't considered a statistical solution to fill in the missing trades. But I don't see how I could estimate any probability distributions when I don't even know how much of the flow I'm missing; I could be seeing 100% of their trades or 5%. And the market is FX, so short selling is quite natural and common.
Aug
5
comment Implications of the Riemann hypothesis in finance?
Are you sure he was implying that it had any applications in finance? Maybe he was just suggesting that they should devote their time to studying pure math.
Aug
4
comment Analyzing an incomplete set of trades
It's actually an OTC market, so I know everything about the individual entities and can link them to their trades.
Jul
30
comment Market making in thinly traded assets
The lack of correlation is intrinsic to this asset. There simply aren't any other traded assets that have any measurable relation to its value (observed or otherwise).
Apr
20
comment Do I need a copula to accurately estimate the VaR of a portfolio of risky assets?
Thanks, one last question. You say that a Gaussian copula suffers the same tail risk problems as a multivariate normal, but what about a Gaussian copula with non-Gaussian marginals? Couldn't that fit fat tails?
Apr
20
comment Do I need a copula to accurately estimate the VaR of a portfolio of risky assets?
Thanks for your reply, but if I were to use historical VaR on 20 assets (or even 6 as in your example), wouldn't it suffer from the curse of dimensionality? I thought that's why I needed to make distributional assumptions.
Apr
20
comment Do I need a copula to accurately estimate the VaR of a portfolio of risky assets?
Well no model is perfect. I am aware of the criticism that people have put too much trust in VaR and gaussian copula, but that doesn't mean they have no use. What would you suggest instead?