Unpredictability of Stock Prices
To the Editor:
Charles Rotblut’s interview with Burton Malkiel [“Stock Price Movements Are Unpredictable,” March 2011 AAII Journal]causes me to wonder if Malkiel’s random walk theory should also be applied to portfolio strategies. That is, perhaps a truly diversified portfolio should not only include Professor Malkiel’s approach, but also separate sub-portfolios made up of varying disciplines that might include trend, relative strength, or some other technical-based strategies, with each sub-portfolio diversified among the suggested sectors. After all, the portfolio make-up proposed in Professor Malkiel’s book may or may not be the correct strategy over the next 10 or 20 years if asset class correlations and relative returns veer toward previously undefined territory. Why not hedge by using two or three approaches with reasonable risk-adjusted returns that have low correlations? My guess would be that, for most individual investors, today’s liquidity, low costs, and educational resources (such as AAII) make a variety of various strategies sensible alternatives to using one methodology that relies primarily on periodic rebalancing into asset classes that may or may not perform in a manner that is consistent with their historical risk, returns and correlations.
To the Editor:
Burton Malkiel is a giant in the financial world, and “A Random Walk Down Wall Street” has been an enduring classic for almost 40 years (with revisions). He is right about almost everything—technical analysis is indeed nonsense, and there are extremely few people (Warren Buffet is one) who can use fundamental analysis with great effect. However, he ignores the basic statistical method of regression. It is fairly easy to build a regression model with a few very significant variables that explains or predicts more than 60% of the monthly variations from trend in average stock prices. Unpredictable would be 0%. It is complicated to figure out how to make money with such a model, but it can be done. I am not sure how many people do, but I suspect it is more than a few.
Stock prices are predictable with regression, probably because the basic causal forces are few, straightforward and consistent. I agree with Dr. Malkiel that interest rates are not predictable, probably because there are too many intertwined factors involved. The economy is similar, and even sophisticated econometric models often fail.
However, in all fairness to Dr. Malkiel, the necessary data was not available until around 30 years ago, and the period from 1983 to 2000 saw such a powerful upward movement in stock prices that it was dangerous to be out of the market. And the most recent bear market was caused by the collapse of the housing boom and the derivative crash, so it was not predicted by stock market models. That leaves the 2000– bear market as the most likely time a regression model would have produced profits. Not a strong practical case for market timing as a strategy, even though we can argue that stock prices are, in fact, statistically predictable.
To the Editor:
I enjoyed your interview with Dr. Malkiel. As a (soon retiring; I’ve done my bit) academic who has worked with statistical theory and analysis, I suggest that you refer to him not merely as “a professor of economics at Princeton University” but also as “the professor of rational finance, appreciated the world over and heeded, if you know what’s good for your portfolio.”
While reading your interview, one of my mental images was a graph of the perverse twists of the random walk formula that he described to a tee.
A month ago, as I read the 10th edition of his book, I alternately broke out in applause or chuckled in recognition at the vagaries of our all-too-human minds, which are often stoked with superstition.
Don R. Warren
To the Editor:
Paula Hogan’s article (“The Role of Inflation-Indexed Annuities,” March 2011 AAII Journal) recommends that an annuity be held in a tax-sheltered account. My question is: Do payments from such an annuity count toward the RMD (required minimum distribution) for those of us over 70½ years of age?
Paula Hogan responds:
Yes, when an immediate annuity is funded with IRA funds, the lifetime payments meet the RMD requirements for the funds invested in the annuity. You would still have RMD requirements for the funds remaining in the portfolio IRA.