« Andres Piedrahita, the Lifestyle Heuristic, and the Scientific Method | Main | Richard Bookstaber on Large Hedge Funds »

April 01, 2009


Feed You can follow this conversation by subscribing to the comment feed for this post.


Unless you mean a conditional probability theorist who has read "General Theory", I think you mean Bayesian.


I disagree with your point on sunk gains. As I was training to manage money I was initially on your side of the argument. It seems to make absolute sense that the future return is all that matters. But as I've progressed, and seen a lot of smart people get a lot of things wrong, I have come to focus very much on the unknowability of future events. You can do your fundamental work well, but some exogeneous event may render that irrelevant. What really matters for your return is the margin of safety - it improves the odds that things will work out as you hope, and you play those odds. If you are sitting on substantial gains, I think it is very likely that your margin of safety has deteriorated, and the risk/ reward has shifted, if not actively against you, then less in your favour. In almost every case I would, at the least, lower the position, preferring a concrete gain to a notional future profit. There are also opportunity costs to factor into this discussion, and perhaps I am not displaying the sort of confidence you would like to see from a top tier value manager, but I look around me and wonder whether confidence is all it's cracked up to be....


Yes, that's a correct use of the word "Bay(n)esian", and I think your point re: gains after a serious bear market is well taken. I don't know anything about Tanous, but if it's true that one of the two pillars of his argument is essentially "historically outsize short-term gains are rare, so why would they happen today", full stop, then that strikes me as an example of shoddy inductive reasoning.

The comments to this entry are closed.

I'm a Seeking Alpha Contributor:

  • Seeking Alpha Certified

Enter your email address:

Delivered by FeedBurner

Blog powered by Typepad