Thursday, October 8, 2009

Why "The Long Run" Still Matters in Choosing Stocks

Unfortunately, I was just rejected for a letter I submitted to the Economists' Voice, which was responding to a letter by Boston University financial economist Zvi Bodie.

Well, that's the breaks. At least I didn't get a form letter. Berkeley economist Aaron Edlin responded, writing, "My own take on Zvi Bodie's point suggests that you are heading at a tangent.  Not an uninteresting tangent, but nonetheless a tangent."

So anyway, in the hope that you too will find my letter not uninteresting, I am printing it below. And, in fact, I really do think it makes important points.

Letter: Why "The Long Run" Still Matters in Choosing Stocks

By Richard H. Serlin

Dear Editors:

In Zvi Bodie's letter, "Are Stocks the Best Investment for the Long Run?", he makes the argument that although the probability of a cumulative return less than the average becomes much smaller over the long run, the potential total dollar loss becomes much greater. He also mentions, "Paul Samuelson’s rebuttal of the conventional 'stocks for the long run' argument". Indeed Samuelson did prove in a 1969 paper that with some standard utility functions and assumptions typically used in economics and finance, the optimal percentage invested in risky stocks does not depend on the investor's age or time horizon. In Samuelson's model whether the investor is 22 or 62 makes no difference in the optimal percentage of stock in his portfolio.

While this model did give some important insights, like all models, it's only as good as its interpretation, and as usual, the most intelligent interpretation is not literal. First, the model assumes away potential short-term liquidity problems, like inability to pay the mortgage or children's tuition, or inability to weather a job loss or serious illness. Next, it assumes that an individual's level of risk tolerance is constant throughout life, when in fact, a healthy young person can tolerate a large loss of wealth far better than a senior.

The model also assumes that the utility an individual receives depends only on his current wealth, or income. In reality, the utility an individual gets from his current income depends greatly on what his income used to be, and on the incomes of his peers or reference group. An individual will be far happier with an income of $100,000 per year if he has spent his life at $100,000 or less, than if he has spent his life at $500,000 or more. He will also be far happier with a $100,000 income if the incomes in his reference group (family, friends, neighbors, perceived peers) are $100,000 or less, than if they are $500,000 or more (for an overview of the evidence see Frank, 2007).

Why does all of this make the time horizon and age of an investor important in deciding what proportion of his savings to invest in stock? If an investor is 60, and he invests all of his savings in stock, then if the stock market loses 50 percent in a year, as it has recently, he may not be able to afford important medical care, or otherwise may not be able to take care of himself properly. This is far less likely to be true of a healthy 25 year old. In addition, at 25, if there is a large loss in the stock market, an investor can hold off on, or slow down, increases in consumption, so that his lifetime consumption will stay steady, or better yet steadily increase. An investor of 60 having long grown accustomed to living on $100,000 per year, and with his working years coming to a close, will not be able to do this. He may have to suddenly spend the rest of his life consuming at an income far lower than what he has long grown accustomed to.

With regard to the consumption of others, if an individual is investing heavily in safe T-bills with an expected inflation adjusted return of 0.7 percent, while his reference group is investing heavily in stocks with an expected inflation adjusted return of 6.8 percent (see Siegel 2008), then over time his reference group, and indeed society as a whole, will likely grow multiples wealthier than he with the power of compound return, and this will typically greatly lower his utility (see Frank 1999 and 2007). If, however, stocks crash, they also do so for his peers, so he maintains his relative position. It's a "we're all in the same boat" situation (You may ask, if everyone invests in stock, will this substantially bring down the return of stock anyway? Not necessarily, not if the flexibility of stock simply allows firms to create more wealth than they can with bonds; see Serlin 2008.)

In the literature I have not seen these factors considered in a lifetime portfolio model. I believe this is partly due to an overvaluing of mathematical ease relative to realism, and partly due to a bias against considering positional/context/prestige factors, despite the great effect they have on utility and behavior (see Frank 2005).

Richard H. Serlin
Adjunct Professor
University of Arizona
National Personal Finance Education
Tucson, AZ, USA


Bodie, Zvi (2009) "Letter: Are Stocks the Best Investment for the Long Run?," The Economists' Voice: Vol. 6 : Iss. 3, Article 3. DOI: 10.2202/1553-3832.1488. Available at:

Frank, Robet H. (2007) Falling Behind: How Rising Inequality Harms the Middle Class. Berkeley, CA: University of California Press.

Frank, Robet H. (2005) "Positional Externalities Cause Large and Preventable Welfare Losses," American Economic Review, Vol. 95, No. 2: 137-41. May

Frank, Robert H. (1999) "Our Climb To Sublime; Hold On. We Don't Need to Go There," The Washington Post. January 24th. Available at:

Samuelson, Paul A. (1969) "Lifetime Portfolio Selection by Dynamic Stochastic Programming," The Review of Economics and Statistics, Vol. 51, No. 3: 239-246. August.

Serlin, Richard H. (2009) "Supply Side Explanation of the Equity Premium Puzzle," Working Paper. Available at:

Siegel, Jeremy J. (2008) Stocks for the Long Run. 4th Ed. New York, NY: McGraw Hill.

No comments: