What doesn't make sense? I'm just saying change the bet so that it lasts one year. Optionally, you could repeat the bet to get results to closer match expectation. So like you said, you wouldn't have to worry as much about funds getting too big. With this logic Buffet might think the break even point might be something like 6-7 years. Does this make sense that he wouldn't take the bet if it only lasted 3 years?

If you do it once, you dramatically favor investments with an asymmetric risk-return profile. If yo do it multiple times, thereby allowing market-timing on the part of the bettor, not the hedge fund managers, you need to give Buffett the same option.

Why shouldn't you have to worry about this?

Ok, I see what you mean.

Perhaps the bet is more to show that "the average investor is better off in an index fund" as opposed to "you can't pick hedge fund managers that will capture alpha net of fees." At first I thought it as the second one, which I disagree with, as opposed to the first one which I do agree with.

To test the second one, you can have the person pick new hedge funds every year and keep a running scoreboard to adjust for assymmetric returns.

Yes I would BET against a lot of the FoFs cause the double fee taking at each level (hf level and the Fof level) mean you have a lot of vig to beat the SP500

If they were smart they would take him on in a weekly/monthly sharpe ratio contest

Can one pick an individual FoF for this or does it have to be a group of 'em

also a poor measure for hedge funds. just look at LTCM.

Did anybody actually read the article? It clearly states there's a 2% management fee, plus a 20% cut on the profit for each hedge fund, plus a 5% cut on the rest of the profit for this fund of 5 hedge funds. So that's an annual fee of 2% no matter what plus 24% of profit if any. Any thoughts?

Bit of a nitpick here:

I think the mistake that LTCM made was that they assumed that the distribution of underlying price changes was Normal when it was not. So, it's not that standard deviation is an inherantly bad measure but that you need to get the shape of the underlying distribution correct.

The mistake is that the shape of the underlying distribution can't be predicted by human beings. LTCM felt they were safe barring market conditions that only happen once every thousand years. But they were large participants in these markets using huge amounts of leverage and their own presence greatly amplified existing conditions and created the thousand year event.

What exactly are they trying to measure? Doesn't this determine the metric they should use and not vice versa?

You're looking at returns over a 10 year period...

but the realized standard deviation over a ten year period doesn't necessarily tell you anything about the tails of the underlying distribution. it's a useless measure if you're looking at, for example, a "hedge" fund that is heavily short (downside) volatility and the ten year period doesn't include a significant market crash.

it was this combined with a few other factors that doomed ltcm:

1) belief in their infallibility (extended investments into M&A arb where far different skills/expertise was needed)

2) sick use of leverage as available opportunities narrowed.

3) unwillingness to scale back exposure in the face of clear signs that one or two days could break the fund.

Barron

I don't think this really was a thousand year event. Just another of the many examples of how massive leverage adds to risk and potential disaster. They didn't understand the risk. "Instead of When Genius Failed" the book should have been called "When Genius was Really Dumb"

liquidity was so scarce at that point that an attempt to scale back any significant amount would likely have busted the fund anyway (hence the need for the bailout)

imo the only real mistake the "geniuses" made was having too much of their own capital in the fund. they could have had a pretty sick freeroll. instead they went busto. a little bit of utility analysis would likely have been in order.

How much money did they lose?

spoken like a guy running a hedge fund.


This post convinces me that Buffett is oh so right again!

*slaps forehead*

Exactly right. There is no such thing as a thousand year event. It's also why historic behavior of the markets isn't very meaningful for making critical assumptions about future behavior. It can take only a single firm with large amounts of leveraged capital to do stupid things and prices can be driven to levels never before contemplated.

Agreed, Buffet actually welcomes volatility if it allows him to buy a business he is interested in at more attractive prices.

Not Lolo,

what are his chances of buying a buisness with a return of .400?

impossible i assume

lol

As investors go Buffet is a career .500 plus hitter.

It's actually the opposite I wouldn't bet against him.

Never applies to other less probable events!

APPEAL TO AUTHORITY LOGICAL FALLACY OVERLOAD IN BFI

uhmmm thats true for every thread in this forum

Repeating that using the S&P is ******ed unless Protege invests in equity-only funds (which is almost certainly not the case). A good FoF should have a big edge on the S&P, and while not knowing much about Protege, they seem to know what's up.

Also, the double-fees argument is true but always overstated. Yes, the FoF charges fees for the alpha they generate in addition to the normal hedge fund fees. Good FoFs add way more value than they charge in fees, just like how good hedge funds generate more alpha than they take in in fees, and just like how a good mutual fund will outperform even after expense ratios. You pay the fee for the service that a good FoF provides. The people usually most vocal about the double-fee argument are usually the ones with the least understanding of how an FoF provides value to investors.

I also really don't understand Buffet's hardon for efficient markets. Obviously he believes he has some innate skill in beating the markets - is it so hard to believe that a skilled FoF manager has the same skill?