3 ms·
Hedge Fund Wizards: the problem with historical returns
- ctkrohn 19y agoThe author is right: this is a problem with hedge funds, mutual funds, and really any sort of financial investment. I don't think there's any real way to solve it though. The author picks an unrealistically simplistic investment strategy for which his point is clear. But in reality, you can't determine the risk and reward for more complicated strategies, and requiring hedge funds to register and report their results would do nothing to separate the scammers from the serious managers.
- iamwil 19y agoI was confused at first, as it seemed like he was pulling magic numbers out of the air, but I worked out most of it, and thought others would be interested. And also to see if I mis-understood anything. I get that we start off with $100 mil with the "2 and 20". 2% annually for funds under management, and a 20 percent incentive fee for returns that exceed the 4% benchmark. And I think here, he uses part of the $100 mil to sell $0.10 options, which if you sell 110 mil $0.10 options, you make $11 mil. So far, you have $111 mil, as long as the event doesn't happen. So he uses $110mil of the $111mil to put into an investment guaranteed to give him 4% in one year and he uses the remaining $1 mil to spend on geeks. So after one year, he'd have $110 mil * 0.04 = $4.4 mil from interest. So far, all together he has $110 mil + $4.4 mil = $114.4 mil, and he spent $1 mil on geeks. If you count just money in fund, and not money spent, it's $115.4 mil ($114.4mil + $1mil) So all together, the return is $115.4 mil (total) - $100 mil (original money), which gives us the $15.4 mil, of which $1 mil is already spent on geeks. The $2 mil comes from the "2 and 20" mentioned before, which he gets 2% of the original $100 mil for management fees. Here, I'm guessing it's the remaining amount that exceeds the benchmark of 4% of $100 mil. So $15.4mil - (4% * $100 mil) = $11.4 mil. And he gets 20% of $11.4 mil = $2.28 mil. So in total, he earned $2mil + $2.28 mil, of which he spent $1 mil on geeks so $2mil + $2.28mil - $1mil ~= $3mil I can see why he says the fund manager's not doing anything. The extra $15.4 mil return that's paying the fund manager, geeks, and investors mostly comes from betting on sure thing for options to get $11 mil of the $15.4mil and a guaranteed $4.4 mil from treasure bills, mostly on the strength of having the $100 mil in the first place. "Making $110 from $100 is hard work. Making $110mil from $110 mil is inevitable"
- iamwil 19y agoOops. I meant, "making $110mil from $100 mil is inevitable"
- marvin 19y agoMisleading headline; the article says nothing about historical returns being a poor metric of fund performance. The scenario presented is a standard example of high-risk investment which can easily (and fraudulently) be presented as ingenious investing. The problem here isn't historical returns being used as a metric for fund performance, it's transparency. Strategies like these will not yield a high return over a period of, for instance, 20 years. The only reason this works (for now) is that these funds only live for about 5 years, which is not enough to provide a good statistical sample for the events being observed. If investors were actually presented with the raw statistical data (10% chance of no returns, 90% chance of 10% returns) both manager commission and the clientelle of these funds would look completely different. Pricing deficiencies are probably present to a high degree in derivatives markets like these. But these markets are too young for investors to have any confidence in the actual probabilities of these events and how to relate to them. If one of these funds survives over a period of a decade or more and its practices look solid, that would be a sign that the managers know what they are doing. Regardless, the nature of derivatives markets is more akin to gambling and less related to the creation of wealth. As I have said before, I have a bit of a problem with the bashing of "historical returns". In many cases it is certainly the case that the returns are based on dumb luck, but saying that a good track record over a long timeframe is worthless is equivalent to saying that all fund managers understand the pricing of their market equally well. Frankly, I find this notion a bit ridiculous. If the performance of funds was truly random, there might be one or two investors in the world that beat the markets over a decade. In fact, there are a lot of such investors. Some of them operate public funds, but unless they happen to have an altruistic streak, each of them has a lot more to gain by just managing their own fortunes.
- yummyfajitas 19y ago>Misleading headline; the article says nothing about historical returns being a poor metric of fund performance Sorry about that, I thought that was an immediate and obvious implication of the article, namely that there is a problem with historical returns as a metric (not that they are a poor metric). In any case, I didn't mean to imply that historical returns are useless, just incomplete. And I thought this article did a great job of illustrating an example where that is the case.
- far33d 19y ago"If past history was all there was to the game, the richest people would be librarians." - Warren Buffett.
- 1gor 19y agoThere is nothing particularly fraudulent about the way the described hedge fund makes money. Selling implicit insurance against the unlikely event is how majority of investment industry operates. There are two types of investment strategies generally - the ones that bet on equilibrium continuing holding (or reverting to its mean) and the ones that bet that an extreme change will take place. If you take any credit risk at all (like buying a bond) -- you are in effect selling your creditor an option to default for which you'll be paid a small premium over the life of that bond. Same if you are betting that price of some mildly 'undervalued' security will come back to its historical valuation average. The example of betting on a change is Soros betting that the Pound will be dramatically devalued. The first two examples are called 'investment', the later a 'speculation'. Of course, we know why. "Investment theory" holds that extreme events are very unlikely. So betting on average is the only sensible thing to do. Which is rubbish, of course, because markets are not random and are not described by a normal distribution no-matter what "efficient markets hypothesis" tells you. Markets are dynamical systems and extreme events happen quit often So, a hedge fund manager who implicitly sells insurance (options) through his strategy is a rule in the industry, not an exception. Articles like this only get attention during times of financial crisis. During the peaceful 'mean-reverting' times (like 5 past years) unscrupulous managers absolutely crowd out the risk-conscious ones. And the investment pseudo-science actively encourages that through use of useless statistical tools like Markowitz portfolio optimisation etc.