Monday, September 24, 2007

One step ahead

I read about this interesting psychology experiment:

Take a large group of people, say >10. Let each pick a number between 0 and 100. Take the average of these numbers. Whoever’s pick comes closest to 2/3 of the average wins. All participants are informed of the rules of the game beforehand.
Time and time again, the winning pick lies around 22.

Here's how it works: if everyone picks randomly, the average should be around 50 and 2/3 of that is about 33. However, seeing that, most people (i.e., the crowd) picks a number around 33, so the eventual winning number is 2/3 of 33 which is 22.

One can take this logic one step further and pick 22 * 2/3, or another step, another step, and so on. Pretty soon one would arrive at zero. Unfortunately, such “farsightedness” doesn’t win the game, since the best can be hoped for is a tie when most other participants also pick zero.

How does this experiment relate to investing and forecasting? It teaches that the essence of timing is to be one step, and only one step, ahead of the crowd. While it still doesn’t make forecasting any easier, at least it provides some explanation of why the market behaves the way it does.

The housing/MBS mess serves as a good example. The downturn in homebuilder stocks started in 2005-2006, then the subprime lenders such as New Century imploded this March, finally the subprime MBS caused a seizure in the credit markets this summer. With perfect hindsight, we now see clearly how the event in March should have warned us about the deeper problems in the mortgage market, and yet the market staged a quick recovery and robust rally following the correction in March.

Another recent example is the recent behavior of gold. In the immediate aftermath of the credit market turmoil, investors flocked to treasuries of all maturities and away from all things, including gold, that hedge funds are even suspected of holding. This was despite unprecedented liquidity injection by the central bankers and gold’s reputation as an inflation hedge. The selling culminated on Aug. 16. Since then, the attention has been shifted to a weakening dollar as the Fed started lowering interest rates aggressively in the face of a weakening economy. Gold prices soared to a quarter century high after this quick about face. In both examples, the market ignored what I considered “logical extensions” at the time; instead, it chose to react to the immediate facts/concerns at hand. To get ahead of the crowd by more than one step would incur significant losses even though the bet may eventually be proven correct.

That brings us to today. Are we really to believe that the housing ills can be cured with Fed rate cuts? (See the table mortgage of rates: rates are still higher than 6 months ago, especially jumbo rates that do not appeared to be affect by the 50 bps cut) What about the “private equity put” that not only has gone away but also is giving investment banks indigestion? No wonder Peter Grandich issued a call to short US stocks on Wednesday. (not investment advice)

Until we have convincing new highs (and preferably confirmed by the transports) I’m still operating under the assumption that we are in a relief rally as attention shifts between the credit market seizure and the weakness in the real economy. But I also thought the weak August jobs number would be a sufficient wake-up call. Instead, the market turned its focus onto the Fed without skipping a heart beat. The next make-or-break event is Q3 earnings.

As usual, we are more interested in the reaction than the actual numbers. A number of brokerages have already reported and despite Goldman’s stellar performance the broker/dealer index has once again broken its 50 dma. It should be a great cause for concern.

9/26/07 There was an error in this post: the last chart was of $bkx, the bank index, not $xbd which looks similar.