Reputation
998
Top tag
Next privilege 1,000 Rep.
Edit questions and answers
Badges
4 10
Newest
 Taxonomist
Impact
~12k people reached

Jul
30
comment Estimate weekly, yearly quantities from finite samples
I've read your additions on merging parallel nonoverlapping estimates. In the i.i.d. setting my gut feeling is that you're losing some information/efficiency since the boundary elementary returns participate less than the central ones, but that might still be worth it depending on your goal. Also averaging dependent values can be tricky. But what is your main goal? To capture some internal autocorrelation or the non-trivial horizon dependence even in the i.i.d. case first? Would you be fine with autocorrelation in a lognormal model?
Jul
29
revised Estimate weekly, yearly quantities from finite samples
added references
Jul
29
answered Estimate weekly, yearly quantities from finite samples
Jul
11
comment How to create a Stochastic Process through pre specified points?
This is not guarandeed to match the actual HOD as it might easily get higher earlier and then revert back.
Jul
9
comment How to deal with zeroes in returns?
Usually the choice of when not to trade is also part of a strategy, so zeros are also meaningful imho and should not be discounted for. Or is the choice of whether to trade or not exhogenous? I would also suggest looking at the literature on performance attribution to better frame your problem.
Jul
3
revised Stress testing covariance
added 141 characters in body
Jul
1
asked Covariance estimation: shrinkage, random matrix theory, what else?
Jul
1
awarded  Citizen Patrol
Jun
25
revised Principle Component Analysis vs. Cholesky Decomposition for MonteCarlo
Filled gap on PSD&Cholesky.
Jun
25
answered Principle Component Analysis vs. Cholesky Decomposition for MonteCarlo
Jun
11
answered Data Synchronization
Jun
11
comment Volatility Estimation
Is E in the first case the covariance between strategies? Are these two E-s coherent/how is it ensured?
Jun
7
comment Block Bootstrapping Relative Returns
Anyway for risk a 0 mean return is a common assumption. Not only using a drift is an arbitrary decision (just like 0 drift, but slightly less so) and the process might not be stationary, but there's $\mu_r$'s estimation error too. As a compromise you could blend in $\mu_r$ weighting it by (an estimate of) the estimate precision, not just strictly statistical precision but also sample relevance taking into account how far in the past you're measuring (200 days or weeks are different in two ways).
Jun
7
comment Block Bootstrapping Relative Returns
" You can't use the original prices, so you have to use the relative prices." Do you mean returns?
Jun
5
awarded  Informed
Jun
5
revised VaR for portfolio of funds
typo corrected
Jun
4
comment What commercial financial libraries are available to outsource implementation risk?
@SRKX: may I ask if in the meantime you found and decided to use one? If so, what were the topics needed&covered and the criteria that were satisfied?
Jun
4
comment George Soros models
Whops hadn't noticed your field... then please keep us updated if you find more on this!
Jun
4
comment VaR for portfolio of funds
Is it clear/enough like this or shall it be edited further? I'm writing in a hurry from work so it might not be clear at a first read... I kept it more on the historical modeling there, other than referring to forecasts discrepancies which emerge as a consequence anyway.
Jun
4
revised VaR for portfolio of funds
some detail on the hybrid approach