Oct

31

 This is a paper by Victor Haghani of LTCM fame on bet sizing observing and analysing how people place bets on a coin flip that is biased to come up heads 60% of the time.

Ralph Vince writes:

It's a very interesting paper, and to many might be surprising. A couple of comments:
1. It assumes someone's criterion in wagering on this is to maximize what someone makes. This is certainly not the case in capital markets, where (the rather nebulous) risk-adjusted return is king, specfically: "Optimal F: Calculating the Expected Growth-Optimal Fraction for Discretely-Distributed Outcomes"

2. Even what the authors and Thorp himself claim are the amounts to wager so as to maximize expected gain, their answer is not quite aggressive enough! The amounts the refer to are asymptotic, as the number of trials ever-increases. The author himself points to a horizon of 300 plays in half an hour, and the actual optimal wager (which would, int hat time period, yield a greater return than the authors or Thorp point to) is slightly more aggressive, and can be determined from the above paper.

Not trying to toot my own horn (it needs no tooting, and besides, my horn will do a lot, but tooting it won't do) but the paper is inaccurate on these two points.

Jim Sogi writes:

Thank you for the interesting article. The other night at our band practice, the bass player's wife, who works at a public school, asked me if I was taking my money out of the market. She had heard a number of people were worried about the election and a market drop if either candidate was elected. I told her the market would probably go up, though it might jump around a bit. I thought that was interesting. Its an example of the public doing the wrong thing, for the wrong reasons. It reflects peoples fear about uncertainty about the election. It helps explain some of the market action recently. 

Rocky Humbert writes: 

Mr. Sogi's anecdote and conclusion is a textbook example of Confirmation Bias — which is the tendency to search for, interpret, favor and recall information in a way that confirms one's preexisting beliefs or hypotheses.

To wit: On what basis does Mr. Sogi conclude that the bass player's wife represents the "public" — as distinct from Mr. Sogi himself being the "public" ??!!

How the stock market will perform over the next 30 or 60 days has very little to do with the study of a coin that is heavily loaded to land on heads. At best, the stock market's performance over the next 30 days is only slightly better than 50:50. 

Alston Mabry writes: 

Just had to do a quick sim of their betting game.


Comments

Name

Email

Website

Speak your mind

1 Comment so far

  1. Arch Stanton on November 2, 2016 3:15 pm

    > To wit: On what basis does Mr. Sogi
    > conclude that the bass player’s wife
    > represents the “public” — as distinct from
    > Mr. Sogi himself being the “public” ??!!

    Mr Humbert makes a good point about narcissism, and indeed one can point to his own prior post on the election — that the sentiments underlying Brexit and Trump are related — as a related bit of evidence. Mr Humbert appears to believe he’s made an original observation, whereas in fact the Brexit/Trump nexus turns up repeatedly in day-to-day punditry, and is nearly a cliche.

Archives

Resources & Links

Search