Peter Bernstein on Amaranth: Better to be Wrong Than Too Right

Peter “Against the Gods” Bernstein has delivered the most choice quote so far on the Amaranth hedge fund meltdown:

“What happened to that hedge fund shows that when you’re really right, you always overstay the position and that’s when you get murdered. It’s better not to be right so much.”

It’s a lovely quote, and so spot-on. Diane Vaughan called it “normalization of deviance” in her classic book on the sociology of risk, but the gist is the same: Humans who are rewarded too often for taking ill-understood risks (or at least are not penalized) normalize the risk, and then take on even more risk. One day, of course, the whole edifice almost inevitably comes crashing down, sometimes literally, as in the case of the low-temperature launch of the space shuttle Challenger, which is Vaughan’s subject.


  1. Bernstein is definitely a great read. Some hedge funds are very dangerous have a look here:
    my own pet risk study is systmic risk induced by major participants (big bank/funds) model autocorrelation.
    The concern involves a moderate exogenous shock to a credit or equity market that gets exacerbated by too many homogenous risk models forcing risk capital to flee from the market or participant. Too much look alike VaR modelling can cause a harmonic effect. The basel 2 capital accords could induce tighter harmonics in bank risk modelling and exposure profiles.
    If anyone knows of a good study on correlation variances in prop trading (banks and funds) over the last 20 years. I would be interested to see it. Ideally returns are positive with low correlation. Increasing degrees of correlation would be a concern as it would indicate greater homogeneity potential either in exposure or response to small shocks.

  2. Very true. I’ve taken a shot with this story as well (link below), based on my hedge fund experience. I paraphase Maounis’ July monthly letter comment on natural gas as: “We probably should cut our natural gas exposure, but we think we can make some money in the short run. So we’ll stay the course”
    It’s easy to get married to a position (and your brilliance) when you’ve had some success. Unfortunately, the market won’t let you dictate its course.

  3. I think the uneducated just call it “pressing your luck.”

  4. This is exactly why I run a quant fund. Strip out a magnitude of layered complexity (positions) and remove the emotions.
    Don’t ever let anyone tell you that algorithms aren’t your friends. Caveat = US DOD algorithms if you are looking at them from the wrong end of the barrel.

  5. what exactly is the “price” for being wrong? NOTHING.
    that is why you will continue to see them press the risk. it PAYS to blow up. you make a fortune in the run up and get to keep if things go bad. heads I win tails you LOSE. you can guess where this is going to lead.