Like everyone else today, I’m reading Katherine Burton’s amazing inside look at Renaissance Technologies, the greatest hedge fund in history.
Among their secrets – hiring non-finance people with science and math backgrounds, avoiding contact with Wall Street at all costs, constantly developing new edges as old ones grow stale – their ability to stick to what they know works is probably a key one. How did they learn this lesson? By “losing” a billion dollars in a few days, how else?
When rivals and former investors are asked how Renaissance can continue to make such mind-blowing returns, the response is unanimous: They run faster than anyone else. Yet all that running hasn’t always kept them on their feet when everyone else stumbled.
In August 2007, rising mortgage defaults sent several of the largest quant hedge funds, including a $30 billion giant run by Goldman Sachs, into a tailspin. Managers at these firms were forced to cut positions, worsening the carnage. Insiders say the rout cost Medallion almost $1 billion—around one-fifth of the fund—in a matter of days. Renaissance executives, wary that continued chaos would wipe out their own fund, braced to turn down their own risk dial and begin selling positions. They were on the verge of capitulating when the market rebounded; over the remainder of the year, Medallion made up the losses and more, ending 2007 with an 85.9 percent gain. The Renaissance executives had learned an important lesson: Don’t mess with the models.
Sudden, sharp drawdowns will frequently have us second-guessing our portfolios. This is only natural – in the heat of the moment, it always looks like something is wrong or that action must be taken. One of the worst things an investor (or advisor) can do is throw away the playbook in response to temporary volatility or unexpected events.
Renaissance is using leverage in their Medallion fund, so they’ve got considerations beyond performance to consider – like the survival of the company. For most investors, this is not the case, so capitulation is always a wrong move.
Don’t mess with the models.