Thursday, May 19, 2005

So You Want to be a Market Timer

Speculators who try to buy and sell so as to catch the important swings in the stock market averages are called market timers. Market timing became unpopular during the stock market boom led by the dot com and tech stocks in the 1990's. But I think market timing is making a comeback now and will become steadily more popular during the next 10 years.

Should you try to be a market timer? If you are reading this blog you probably have some interest in timing. But I don't believe most people have any idea how difficult it is to successfully time the swings in the market averages.

My argument is really a simple observation about how probabilities work. To be a successful market timer you have to make two consecutive good calls (i.e. two good calls in a row). You have to make a decision to sell and then another decision to buy so that your selling price is higher than your buying price. (I am ignoring issues which have to do with risk reduction since they only complicate my argument while not changing my conclusion.)

What are the odds that you can make two good calls in a row? Let's do a little thought experiment. Suppose the probability that you make a good call does not depend on whether your last call was a good one or not. As a baseline, suppose your are a novice market timer who makes good calls 50% of the time. What is the probability that you make two good calls in a row? It is just 50% of 50% - only 25% ! So a novice timer has only a 1 in 4 chance of making himself better off compared to a buy-and-hold strategy !

These are lousy odds and clearly someone who makes correct calls only 50% of the time shouldn't try to time the market. But suppose you are better than this. Suppose you make correct market calls 70% of the time. Well, the probability that you make two good calls in a row is then 70% of 70% or 49%. Still less than a 50 - 50 bet! Again, it makes no sense to try to time the market even if you make good calls 70% of the time!

Well, suppose you have improved your skills to the point where you make correct calls 80% of the time. I'll tell you right now that not many market timers will claim to achieve this level of skill. But even now, the probability of making two good calls in a row is 80% of 80% or 64%. This is still less than a 2 out of 3 chance!

I think my little thought experiment actually overestimates the success probability of market timing. I don't believe that the probability of making a good call is independent of whether or not your last call was good. In fact I believe that the fact that you made a good call last time makes it even more likely that your next call will be a bad one. People like to observe that the market is always changing the key to its lock, so the insight that leads you to make a good call probably won't work the next time you try it!

This explains why even famous market timers rarely have winning streaks more than 4 to 5 years in length. This is just the length of the average stock market cycle during the past 50 years.

So if you want to time the market you must recognize that the odds are stacked against you. Don't bet your retirement money on it!

No comments: