Concentrate!
Katherine Burton offers a fair warning to investors in concentrated hedge funds: a fund with fewer holdings will have more volatility.
Her star example is Edward Lampert's ESL, which is, indeed, down 27% from last year. Now, if Lampert's investors are leveraged 3:1 against their investment in the fund, this is, indeed, very bad news. Of course, if they borrowed to invest in the fund before last year, they've long since solidified their financial situation thanks to his 30% annual returns. Burton apparently believes that if Lampert would only stick to his top 50 stocks rather than his top 5, he'd have survived the recent rout. True, but it ignores the fact that even Lampert would have trouble showing such great returns from 1988 to 2006 if he'd divided his resources so thinly (that's not even taking into account ESL's incredibly rigorous due diligence. By the time they actually buy a stock, they probably know more about the company than the CEO).
Burton also cites the case of SRM Global, down 70% thanks to concentrated bets in Countrywide and Northern Rock, among others. That's truly damaging, but: surely Burton is aware that Countrywide and Northern Rock each owned hugely diversified portfolios consisting of thousands of different mortgages and mortgage-backed securities. If diversification is such a panacea, why did SRM do so badly diversifying by proxy? And was most of Countrywide owned by concentrated portfolios -- or were most of the losses out-of-the-headlines declines due to money blindly allocated to the financial sector rather than carefully invested in a single business?
Careful thought will show that while Lampert's portfolio may be concentrated, it isn't unusually so: large mutual funds and hedge funds with thousand-stock rosters of holdings are clearly not analyzing each company individually, so they must be purchasing stocks in order to execute a strategy (e.g. 'Buy energy stocks as long as the dollar is dropping and Americans insist on commuting by car,' or 'Expect companies' valuations to converge on the industry average'). Is a portfolio based on four or five of these theses, executed through trading hundreds of stocks, really less concentrated than a Lampert-style list of half a dozen individual businesses with individual reasons to excel? Perhaps from one day to the next, the five-thesis manager will have more steady returns than the five-stock guy. And in a terrible year, someone who owns a thousand average stocks will be closer to the median return than someone with a few extraordinary holdings. But over a lifetime, there is almost certainly an inverse correlation between the quality of one's decisions and the number of decisions made.
Burton should know: the next few years may be boring years for a journalist who writes mostly about hedge funds. But Burton's name wouldn't be recognizable if she just wrote a little bit about everything. The author of this should know better.