More Money than God
I recommend Sebastian Mallaby’s book, More Money Than God, if you are at all curious about how hedge funds work. The writing style resembles Malcolm Gladwell or Michael Lewis in that a technical subject is made interesting and straightforward by anecdote-driven narrative. It made me want to join a hedge fund or at least start actively investing my own money. Mallaby also focuses on the two things every economist wonders about hedge funds- do they produce social value, and how can they make money if markets are efficient.
I had heard some of Paul Samuelson’s story before, but it is amazing that the first man to work out the math of efficient markets theory and a major proponent of index funds was also a major investor in one of the first hedge funds. Hedge funds have a way of making academics look like suckers. When the people at Samuelson’s fund found a non-random pattern in commodities markets, they used their knowledge to make huge amounts of money; twenty years later, an academic economist found the same pattern and published a paper about it, making no money and ending the pattern.
Sometimes the reason markets are not efficient and the reason hedge funds can provide social value are one and the same- illiquidity. Rather than making a premium by picking stocks (which is supposed to be impossible except by luck), they are paid a premium for providing liquidity to illiquid markets- for instance by buying a stock that is being sold en masse by an insurance company that needs to cash out to make payments. If you can identify sellers who sell for a reason other than a belief that an investment is getting less valuable, you will do well by buying from them.
Another way to make money off of inefficient markets is to identify nominal rigidities- like interest rates or currency values being fixed by governments. When a currency is pegged a government will buy and sell it at prices other than the efficient market price- this is almost the definition of a peg. But this mispricing is only valuable to traders if there is a chance that it will be corrected soon- that the peg will be abandoned. The book tells the stories of how Soros and the markets famously pushed Britain and Thailand to abandon their pegs.
Do such ‘speculative attacks’ provide social value? Sometimes they are pushing firms and governments to abandon sooner the unwise policies they would have to abandon later. However, the book also tells of many attacks on firms that were merely illiquid, not insolvent. Rather than watching fundamentals, many funds spend time watching each other and attacking the weak ones into deleveraging spirals. If a firm is leveraged 5 to 1, it can only afford to have a 20% loss, and the margin calls that come with borrowing plus investors who can pull their capital make this number even lower. This means that firms that lose a bit of money may need to sell quickly- and if they need to sell in illiquid markets or if other firms find out and bet against their positions, selling begets selling until a firm goes bust. When the market comes after your firm in particular, the correlation on all your positions can go to one. Such positive feedback loops are normally value-destroying.
One anecdote makes me worry about the accuracy of most empirical academic research. At a top hedge fund, “the data they looked at had been painstakingly swept for typing glitches and errors- it was cleaner than anything available to most finance professors. Time and time again, an eager academic would contact D.E. Shaw, claiming to have discovered a profitable anomaly in markets. Time and again, Shaw’s faculty would find that the anomaly consisted merely of misreported numbers.” If a hedge fund analyzes data and comes to an incorrect conclusion, it loses money. If a professor does the same, he probably loses nothing. It is very common for economists to critique how data is analyzed, and somewhat common to argue about which dataset to use, but it is extremely rare to check that data was properly cleaned and there were no coding errors in the analysis. For instance, right now I am working on cleaning a large dataset on gas prices, rearranging it and checking for errors, with the goal of publishing an academic paper. However, if I were going to bet money on the conclusions of the paper I would surely be more thorough.
Anyhow, read the book, for the economics and the stories. I have made it sound less interesting by focusing on the economic issues, but it is quite accessible.