Few things to ask about FoHF, hoping you can shed some light…I’m curious to find out what are the common optimisation techniques used in this space… or is there optimisation at all…
I don’t think anyone does optimisations. Optimizations maybe, but not optimisations.
Optimize => US spelling Optimise => Commonwealth (I think)
The problem with using any optimization with a FoF is how difficult it is to size positions to what the technique/model says they should be. Additionally, the rebalancing frequency is typically not very helpful to the process (buying and selling funds can take over 6 months when the cash can literally earn 0%). An additional issue with more sophisticated techniques is the data availability issue. If you get monthly NAVs, how many years are you willing to wait until you have enough data points to make a ‘meaningful’ t/z test? Not a practical choice in my opinion. You can get weekly NAVs, but getting more specific data (gross/net exposure, weights, etc.) is unusual. Hope this helps
iheartiheartmath Wrote: ------------------------------------------------------- > I don’t think anyone does optimisations. > Optimizations maybe, but not optimisations. Yeah - why on earth would the English spelling be correct?
Because everyone uses it?
wow, give Aussie English some break… it’s like realise and realize, centre and center. You know, we like to have our rule on a totally isolated island. That being said, yes I agree that rebalancing is an issue but right now I’m looking at it strategically so rebalancing should not be an issue. I agree data integrity is an issue, still try to figure out how to work around it but still not my focus yet… anyway, gotta run for now, will come back to finish the question… I was hoping to see something along the risk management side.
Based on my understanding of the FoF business I would think that whatever value there is in it lies more in the due diligence (roughly speaking) aspect of it than in “optimization” (assuming you mean mean-variance optimization). Seems pretty obvious to me that when you’re talking about individual hedge funds the inputs would be so speculative as to make MVO a pretty useless exercise (as we all know, the results of an optimization can vary drastically with changes in the inputs), although it may be worth doing to provide one perspective on a portfolio.
If by “data integrity is an issue” you mean “it’s anyone’s frickin’ guess what returns any of these #$*&@ things are going to generate in the future” then I think you’re on the right track.
Two words: Michaud.
Thanks for the response, my background is actually an asset consultant… I guess my original question was vague so I’ll try to narrow the scope and explain what I’m looking for. Firstly, by optimisation, I’m not referring to mean-variance. Agree, it’s quite useless. I was more referring to cVaR type of optimisation, there’s some researches around on this topic but not quite solve my concern about the abnormal return distribution of hedge funds. Hence, I would like to try to get few more methods or techniques and conduct some research on them. Second, I personally believe DD and SAA are equally important when it comes to FoF (more so FoHF). Like traditional investment, you can pick great stock but detracted by portfolio construction decisions. However, SAA for FoHF was much undermined for a long long time. As liquidity and transparency requirement increase post the financial crisis - this will become a focus in the next decade (I think…) Hokieomics pointed out something interesting, there is indeed some decent stats between strategies. You would be suprise how market sensitive some of these strategies are… Anyway, I guess…if there’s no common optimisation techniques, I’ll settle with my research. As an asset consultant…I learn to BS my way through anyway! 
Just to add, data integrity applies to both forward projection and historical…
takumi, with VAR measures, these are calculated based on the data that you have, and the distributions typically aren’t normal (skewed and a bit of kurtosis). You can adjust for this though. good luck with your research
Whenever a theory appears to you as the only possible one, take this as a sign that you have neither understood the theory nor the problem which it was intended to solve - Karl Popper Thanks. I feel like I’m seeing the only possible one…