Friday, February 1, 2008

“Worry-Free Investing” Book Review, Part 3

This is the third part of a review of the book “Worry-Free Investing”, by Zvi Bodie and Michael J. Clowes. This review began here.

The main theme of this book is that stocks are too risky for most investors. The authors repeatedly claim that stocks are too risky even in the long run and that most people should avoid them completely or have only a small percentage of their money in stocks. Curiously, the justification for this claim doesn’t come until Chapter 6.

The authors give some useful historical investing information for the period from 1926 to 2000, including the real returns investors received in U.S. stocks, bonds, and treasury bills (Figures 6.1 and 6.2). By “real returns” we mean the investing returns after subtracting inflation. This makes for better comparisons because we are comparing dollar amounts that have equal purchasing power.

This information was then used to simulate possible outcomes of stock investing over 30 years (Figure 6.6). These simulations were based on the assumption that the future will be similar to the past. The book gives results based on an initial investment of $100. I will base my discussion on an initial investment of $10,000 to make the numbers more meaningful. In the authors’ simulations, here is how much the stocks are worth after 30 years:

Simulation 1: $75,000
Simulation 2: $5200
Simulation 3: $200,000

For comparison, an I Bond with a 3% return above inflation would give $24,300 after 30 years. The authors then point to simulation 2 saying “we see that very bad outcomes can occur for long-time horizons.”

The outcome of Simulation 2 seemed very unlikely to me. I decided to run my own simulations based on choosing each year’s return randomly from the historical returns in Figure 6.2. Instead of only doing 3 runs, I let my PC go for a few hours, and it completed a billion runs. This is more than we need, but it gives a good picture of the possible outcomes.

One thing I learnt was that 25-year olds are about 10 times more likely to die before the end of the 30 years than they are to get stock returns as bad as in Simulation 2. In only one out of every 140 simulation runs the stocks were worth less than $5200. In 87% of runs, stocks beat I Bonds. Half the time stocks were above $86,000 compared to I Bonds at $24,300.

If we accept Simulation 2 as a meaningful result even though it happens only once out of 140 times, what about the high end of stock return possibilities? In one out 140 runs, stocks returned more than $1,060,000! The main result here is that the authors’ three simulations do not give an accurate picture of what is likely to happen.

We continue with part 4 of this book review in the next post.

1 comment:

  1. The simulations probably assume historical means and variances, normally distributed returns, and independence.

    Dollars to donuts, returns over the next 25 years will be less than historical ones. Variances could be larger or smaller. You can be absolutely sure that returns are not normally distributed, and you can be pretty sure they are not independent.