Future of Hedge Funds

[PRE]

“Why would you pay the high fees that hedge funds charge if you are able to get the same risk characteristics, in a statistical sense, by using a dynamic futures-trading strategy?”

http://www.newyorker.com/reporting/2...printable=true

Pretty good article discussing a futures-trading program (FundCreator) that was created that can mimick the risk-return relationship of the majority of hedge funds out there. One of the replicated funds mentioned in the article was Soros’ Quantum Fund NV. I don’t think it’s a question as to whether or not hedge funds will survive into the future, but I definetely believe that there’s too many of them at this point that don’t generate enough alpha compared to the risk they take on.

The one thing the article doesn’t mention, though, is that fact that FundCreator might be able to allow the average investor (ignoring the SEC for now) to pick the amount of risk he wants to take on. I know I would be interested in taking on a higher level of risk to generate a return of 20%. Anyone else?

[pig4bill]

What a ridiculous article. Half of it is just whining about hedge fund fees. That FundCreator thing is nothing but using programmed systems that are probably out there in the public domain. They are completely results-oriented, trying to compare returns and "risk" of programmed systems to hedge fund returns. That's like saying flying a 747 first class from NYC to L.A. is the same as hitchhiking because both methods end up in L.A.

[APXG]

Maybe in a few years they'll "replicate" LTCM

[Jeff W]

Not sure why there's so much negativity against synthetic hedge funds.

The fee structure of conventional hedge funds creates a divide between the best interest of the investor and the fund. Meanwhile, alpha opportunities are drying up by continued saturation of the market.

Note the article states that Goldman Sachs and Merrill Lynch launched synthetic funds recently. They're basically applying the ideas of indexing to hedge funds--it's hard to identify the active managers who can beat the market, so keep costs low(low costs are true alpha), obtain constant factor exposure and aim for average.

[kimchi]

[ QUOTE ]
What a ridiculous article. Half of it is just whining about hedge fund fees

[/ QUOTE ]

Their fees are obscene, but like so many other things, they will charge what people will pay. I doubt even a hedge fund 'tracker' with a low TER could outperform a well-diversified, low-cost portfolio.

2 and 20 fees are too much to overcome, and 3 and 30.... Theere's just too much conflict of interest between controling risk and the managers' rewards. Why would they care if the fund has a good chance of busting out within a few years or as soon as it encounters a statistical outlier if they can make $$$ in the meantime?

[DcifrThs]

i've thought about this since the post and here are some ramblings:

- alpha creation is a zero (negative) sum game. on the other side of every contract is a counterparty w/ the opposite bet.

- the fee structure does create moral hazard. the classic example of which is the fund that writes uncovered options while actual vol comes in at implied vol. as long as nothing crazy happens and markets stay calm, you earn a very steady premium. the returns look almost too good to be true. until something out of sample happens and BOOM you blow up. but hey, you just made a few years' profits + fees from your investors so you're golden.

- it is virtually impossible to mimic the best in the industry. i know because i worked at one of the top funds. if we gave a million people our return stream, the probability that even a single one of them could generate a trading system that was robust to our strategy would be close to, or effectively zero. the reason is the way the trades get researched (which generates a signal), managed (which match the trades to clients accounts), then run through a risk management. this system has to go through each one of over 177 different market segments (even more but some are bunched together) to make sure that no one market can contribute more than a given amount to the variability of the portfolio as a whole. after that, the positions & transactions get sent to trading and get executed (in a manner that minimizes transaction cost. a whole departments' career depends on creating a systemized way to bring down transaction costs). this happens 3 times a day! i don't care what statistical methodology you use, you'll never recreate that alpha.

so what would FundCreator find if it ran that fund's returns? well, we ran something similar (a tool that was created to find biases in returns- i.e. explain variability in return streams via statistical methods & combinations of other types of return streams) and found that over certain short time periods, the variability in our return stream was about 50% explained by a combination of a few currency positions. now you really think that taking those futures positions would come close to giving you the exposure that the above system generates??

no chance.

- now that is one of the best funds though. using the tool i described above, we were able to find that some "long/short" (i.e. supposedly market neutral) funds had a fairly strong long bias. they likely used leverage to produce "outsized" returns that appeared to investors to be better than the market. so maybe it's possible to produce equity based long/short type fund-esque returns via stastical replication. if so, the value is clear and the business model these guys have is obvious.

thing is though, even those are still up in the air. i wouldn't be convinced at all until i saw monthly returns over years in varying market conditions that had a very low tracking error over their target replication.

- another thing is that the return streams of many funds are not available so the software automatically has a selection bias towards the bottom of the hedge fund food chain (the best are the least likely to give out or sell their return stream). i'd like to see what streams they have and how they got them.

it was an interesting concept though (although the article could've stopped harping on the fee structure like a "woe's me" thing)

i'd be very curious to dig deep into fundcreators, though i feel i'd already have a pretty good idea what the basis is. still be interesting seeing how the program runs.

my immediate suggestion (and the best or worst marketing for them) is to start a study. take about 20 funds that they a) can get returns for, and b) claim they can replicate. then show a comparison between all 20 month by month going forward. if there are any big deviations, they're screwed. if not, thats some good publicity.

good article find though.

Barron

EDIT: the fundcreator software though does have some good selling points. funds that have an expected gross IR of .3-.6 are in deeeep water b/c fees eat up a bunch of that and a well diversified passive portoflio would outperform that fund. now can a system liek the one they've provided actually create alpha on its own above and beyond that type of fund? perhaps, but it is a study in the making like i mentioned above.

[PRE]

Very nice post Barron.

The only thing I disagreed with was your point on how the program has a selection bias towards the bottom of hedge funds. I think that only the top of the top (which are few) won't give out their returns because they have no need to. The majority of funds who might not report are those which are doing terrible, IMO. Even the funds who are doing great but are in desperate need of money will probably have to report.

I've been interviewing with one fund which deals with a unique type of investment that just started up last year. It generated a return of 37% and is marketed by one of the most prestigious firms on all of Wall Street. I told the portfolio manager that it must be easy getting cash now, and he said I couldn't be anymore wrong.

Also, the article doesn’t mention the fact that most wealthy people invest in fund of funds (I think), which are totally useless when you take into account the amount of fees.

Ryan

[DcifrThs]

[ QUOTE ]
Very nice post Barron.

The only thing I disagreed with was your point on how the program has a selection bias towards the bottom of hedge funds. I think that only the top of the top (which are few) won't give out their returns because they have no need to.

[/ QUOTE ]

i think that is true. i may have been biased in that opinion b/c i worked w/ a fund that was "closed." only redemptions & failed (not desired) conversions resulted in further capacity and the wait list was substantial to say the least.

[ QUOTE ]
The majority of funds who might not report are those which are doing terrible

[/ QUOTE ]

or funds that have blown up. or funds that reported but now cease to report b/c they have recently closed (a corrolary of what you mentioned). i haven't seen research on this topic and i think that we have stumbled on a great paper idea. just following a sample of "reporting" funds over years to see what is likely to result in reporting failures would be exceptionally valuable to a firm like the one in the article. the main problem i'd see is sampling since the total # of hedge funds that come into/out of existence could prove quite intractable.

[ QUOTE ]
Even the funds who are doing great but are in desperate need of money will probably have to report

[/ QUOTE ]

that group must be transitory and the proportion can't possibly be expected to stay huge (if it is even big at all). i think we are in a stage in the economic/financial/business cycle that is biasing this group upward. if you have done great for years, you probably don't need to report as money has already come in (assuming, as you are saying, that the GOAL of reporting is to bring in new money. CSFB's famed treamont index has a large # of funds but i dont' know how they report or why or how often it changes. that index dates to 1994). now, however, the # of new funds opening and doing great but not getting money has probably increased beyond what one would expect on average simply b/c the climate is so favorable (low recent volatility, the ability to leverage the carry trade to produce outsized returns, the sick performance of debt groups as spreads came in for years etc.).

from an investors perspective, the CHOICE of which fund to invest in is hard to make and the return/risk ratio, length of time earning that return/risk ratio, name recognition, and transparency of the process (i.e. NOT like goldman) all contribute to the choice to invest in that fund.

the pres & CIO of my old fund keeps ALL of his (billions in)wealth completely in a passive allocation for a few reasons.

mostly, he doesn't want to invest in the fund's flagship product b/c that would expose his wealth completely to the performance of that fund. if it blows up (which is highly unlikely but completely possible), he stands to lose far too much.

secondly, selection of good managers is freaking hard (thus fund of funds' popularity) and he can't trust his heirs to be good at it. so earning ~.6-.7 expected risk adjusted returns in the passive product is the best he can hope for.

i think those are the main reasons. others are escaping me at the moment.

[ QUOTE ]
I've been interviewing with one fund which deals with a unique type of investment that just started up last year. It generated a return of 37% and is marketed by one of the most prestigious firms on all of Wall Street. I told the portfolio manager that it must be easy getting cash now, and he said I couldn't be anymore wrong.

[/ QUOTE ]

that is understandable as the length of time the product has been around is small AND the product is "unique" (i.e. not well known and unproven). even if goldman came out w/ that it would be hard. bear is the example that comes to mind, though their product wasn't unique. just hard to value, illiquid, and/but extremely profitable if done right.

[ QUOTE ]
Also, the article doesn’t mention the fact that most wealthy people invest in fund of funds (I think), which are totally useless when you take into account the amount of fees.

Ryan

[/ QUOTE ]

agreed.

thanks for the post and please let me know if i should expand or fix anything else.

Barron

[Jeff W]

[ QUOTE ]

[ QUOTE ]
The majority of funds who might not report are those which are doing terrible

[/ QUOTE ]

or funds that have blown up. or funds that reported but now cease to report b/c they have recently closed (a corrolary of what you mentioned). i haven't seen research on this topic and i think that we have stumbled on a great paper idea. just following a sample of "reporting" funds over years to see what is likely to result in reporting failures would be exceptionally valuable to a firm like the one in the article. the main problem i'd see is sampling since the total # of hedge funds that come into/out of existence could prove quite intractable.

[/ QUOTE ]

I'm not sure exactly what you're looking for, but there is a lot of research on survivorship bias in hedge fund data, including some by Roger Ibbotson(his is highly regarded, I think) and Harry Kat(from the article in OP). Should be on SSRN.

[DcifrThs]

[ QUOTE ]
[ QUOTE ]

[ QUOTE ]
The majority of funds who might not report are those which are doing terrible

[/ QUOTE ]

or funds that have blown up. or funds that reported but now cease to report b/c they have recently closed (a corrolary of what you mentioned). i haven't seen research on this topic and i think that we have stumbled on a great paper idea. just following a sample of "reporting" funds over years to see what is likely to result in reporting failures would be exceptionally valuable to a firm like the one in the article. the main problem i'd see is sampling since the total # of hedge funds that come into/out of existence could prove quite intractable.

[/ QUOTE ]

I'm not sure exactly what you're looking for, but there is a lot of research on survivorship bias in hedge fund data, including some by Roger Ibbotson(his is highly regarded, I think) and Harry Kat(from the article in OP). Should be on SSRN.

[/ QUOTE ]

those bias studies i've read typically are aimed at estimating returns. i want to know about those who choose to report their returns. so a reporting bias is what i'm after.

i highly doubt this has been researched but it would be interesting given the discussion we just had.

Thanks,
Barron