Lessons Learned from 'March Madness'

With the NCAAs all done it seems fitting to learn something from basketball that applies to our investment activities [other than don't bet on the favorites].

If you are fully invested and then find a stock you think merits purchase you need to make a decision. Should I sell Company A to buy shares in Company B? That would all depend on how confident you are that the new shares will outperform the ones you'd be selling.

Now for our basketball lesson…

Imagine your team has a 70% foul shooter going to the line for two shots. Intuitively we feel he's likely to make both as he is a relatively high percentage free throw shooter. But is that really the case?

The odds of his making both shots would be 0.7 [the likelihood of a 70% shooter making the first shot] times 0.7 [the same percentage chance of making the independently calculated second free throw]. That means your 70% percentage turns into just a 49% chance of hitting any two consecutive foul shots.

When you need to 'sell to buy' your situation is the same. You can 'miss' by seeing the shares you sell do well or by watching the newly picked holding go down in value. Even if you're right 70% of the time you will only help your portfolio performance 49% [less than half] of the time.

Morale: You need extremely high confidence [ > 70% ] that the substituting stock is going to do better than your older holding to make a positive impact on your performance.

Psychologists have found our belief that we can predict these things is greater than our ability to do so. Frictional costs [bid/ask spreads and commissions] increase the difficulty factor.

Comments

[trance_dude]

yes, but the expected value of any given free throw is still 0.7 points, and the expected value of sending that shooter to the line for two shots is still 1.4 points.

what you said about the shooter only making both shots 49% of the time is true, but you are ignoring the two cases where he makes one and misses the other. That happens 42% of the time. He'll miss both shots 9% of the time. Add up those probabilities times the points and you get an average of 1.4 points per trip to the line.

the point being - over the long run continually sending this guy to the line is still a bad deal. a good shooter will shoot 50% from 2-point land, making the expected value only 1 point per shot. a good 3-point shooter in college might shoot 35%, also making the expected value around 1 point per shot. That's why coaches don't use the fouling strategy until time constraints become an issue.

the situation you describe might be relevant in very specific circumstances, like say the opposing team MUST make 2 consecutive free throws to tie the game. but i do not think the analogy translates well to the investing situation you described. In fact, ask any trader or blackjack player and they will tell you that being right 70% of the time would be a VERY big edge.

[me0w]

I agree this analogy doesn't translate perfectly--the way I understood it, he means you have to be 70% sure the stock you own won't do well and 70% sure the prospective stock will do well and that's where he's getting the dependent probability of 49%. But I don't know that I see the two thoughts as separate. It's more like "I'm x% confident this stock will do better than the one I have". So by definition of "being sure", if that level of confidence is >50% (plus some margin for B/A and commissions) then you would execute the trades. By that logic, unless you pick stocks like Shaq at the foul line, you can change the portfolio holdings more often than the article is arguing.

In any case, point taken, don't get trade happy unless you're confident.