The Motley Fool Discussion Boards

Previous Page

Investment Analysis Clubs / Liquid Lounge

URL:  http://boards.fool.com/i-dont-think-quotthe-estimation-of-25538631.aspx

Subject:  Re: HNR -- The fish Pabrai didn't throw at VIC Date:  5/29/2007  8:26 PM
Author:  educatedidiot Number:  26714 of 41316

I don't think "the estimation of probabilities is where most people go wrong" though. I don't think most investors even think in probabilistic terms.

Definitely, that's why I joked that this topic tends to be brought up more often in theoretical discussions rather than practical ones. I meant that for anyone attempting to apply a probabilistic approach, the estimation of probabilities is the problem.

Agreed again. The A-bomb scenarios, fraudulent management, etc can usually be whittled down to < 1%. Like you said, the problem is everything between 5% and 95%. That is why the odds have to be drastically in one's favor before placing a big bet. It has to be obvious, given that you have a sufficient set of facts to base a decision on. Naturally, these types of opportunities usually only present themselve under extreme circumstances (HNR, Chrysler, American Express, Geico, Washington Post, etc). They are rare, all the more reason to bet big when you come across them.

Maybe its just a matter of definition then, but to me, for the odds to be drastically in my favor, I have to have come to the conclusion that the horrible outcome is extremely unlikely to occur (i.e. less than 5%). As I said before, that's because it's usually impossible for me to really know whether something is 20% likely vs. 40% vs 60%.

You're right on when you say that estimating probabilities is where most mistakes are likely to occur (for probabilistic investors, anyway). But, in my mind, that's no reason to dismiss it as something that works in theory but not in practice. It does work in practice if done correctly.

While I can't rule out that there are certain individuals that are somehow blessed with a unique ability to assess probabilities with limited amounts of information, my experiences suggest that these individuals are exceedingly rare.

Speaking of working in theory and not in practice, your post sounds somewhat EMT-ish. Do you think markets are strong-form efficient?

Not sure what gave you that idea, but definitely not. Most markets seem to me to be highly inefficient.

I disagree. Do you have a reason that backs up why you think it is improbable that HNR's contract gets approved? If so, I may agree with you and I would love to hear your thoughts. But just saying that that may be the case because people can fool themselves into believing what they want to believe, or believing what their limited memory/experience suggests is correct, isn't a selling point for me. That's just me though.

I have no opinion on the probability of HNR's contract getting approved. I don't think that I'd be able to convince myself that its <5% likely, so for me its not worth thinking about. My post wasn't meant to be specific at all to HNR. I have no opinion on the company whatsoever. Just run through this exercise for yourself (no need to share it with the board): take your probabilities, add/subtract 20-30% to them, and formulate an argument for why the new probability distributions aren't right. Ask yourself honestly how compelling you find those arguments. If it works for you, then that's great! Absent great data or relevant historical knowledge, I just think that's a difficult thing to do.
Copyright 1996-2014 trademark and the "Fool" logo is a trademark of The Motley Fool, Inc. Contact Us