Jeremy Siegel has a recent piece in the Financial Times that restates his view that stocks are the most appropriate investment for investors with a long horizon. I wonder how most of you look at this issue, especially after the recent market gyrations.

Are you still listening to Professor Siegel, or did you shred his book along with your fund statements from last year? I’d love to know how many of you agree with that view, and if your investment strategy reflects it.

Anyway, the assessment of “long run” risk in the stock market, and financial markets generally, is an issue that people in economics and finance have been examining and arguing about for decades. In the late 1960s, Nobel Prize-winner Paul Samuelson presented rigorous mathematical analysis showing that investors should not alter their portfolio allocations to emphasize riskier assets on the basis of time frame alone, given some basic and plausible assumptions about how people trade off risks and returns, and about the nature of asset returns.

Needless to say, not everyone was convinced by his argument. Many continued to advocate that the best portfolio for distant-horizon investors is the one with the highest geometric return, and finance journals are full of articles on the subject even to this day. (This argument reached an amusing crescendo in 1979, when Samuelson published an article on the subject in the Journal of Banking & Finance using only one-syllable words: “Why We Should Not Make Mean Log of Wealth Big Though Years to Act Are Long.”)

This debate has a personal angle for me as well as an intellectual one. My wife is from Japan, and we have extended family living there whom we visit from time to time. And while it is true that stock ownership is not as common among Japanese households as in American households, I am continually taken aback by the fact that broad measures of stock valuations on the Tokyo exchange are still roughly 75% below where they stood in the early 1990s. It’s not “a lost decade” they are reading about and living through in Japan. It’s two.

My own view on the subject is, quite frankly, that equity risk does not meaningfully diminish over time. I base that view on the logic of Samuelson’s analysis, history of the Japanese market, and the practical observation that basically no one is willing to write cheap long-term (20- or 30-year) insurance on the stock market into a straightforward financial contract. Such long-term arrangements are very rare in general outside of certain types of insurance productswhich do not have the reputation for being inexpensive.

I’d also refer the more technically minded readers (i.e., those of you who actually enjoy calculus) to a recent, more technical analysis that tries to take into account the issue of statistical uncertainty around parameter estimates in thinking about long-run equity returns. The analysis argues that equities are significantly riskier at longer horizons than shorter.

So, am I saying that everything you’ve heard about “investing for the long term” is bunk? Of course not.

Stocks can play a very important role in a diversified portfolio of investments, especially for younger people. But in my mind, the reason young people should consider higher equity allocations than older people is not because of the old cliché that “they have time to ride out the ups and downs.” It’s because they have a lifetime of work in front of them. This gives them far greater flexibility than older investors around altering goals, lifestyles, and their commitments to work or particular careers in circumstances in which equity returns are poor.

In other words, human capital is a very important part of the problem. So, in my view, if you don’t have other resources that can help you to offset the consequences of poor markets, heavy equity exposure involves significant risk even over decades-long periods.

That’s not to say that risk is not worth bearing. If investors are risk-averse, theory and basic logic suggest equity risk should, on average, be rewarded with a significant and positive return in excess of what one would obtain from less-risky investments. Otherwise, investors would shun stocks. It’s critical to keep in mind that the “equity risk premium” exists on average precisely—and only—because it doesn’t exist in all circumstances. In other words, no matter what the time frame, there is risk of underperformance (and, potentially, dramatic underperformance). It’s a “risk premium,” not a “return premium.”

If anything good can come out of the current financial crisis, my hope is that it’s investors having a better understanding of risk, and making decisions to hold stocks because they believe the potential rewards are worth the risk—not because they don’t think the risk is there.

Note: Links in this blog post will open new browser windows. Vanguard accepts no responsibility for content on third-party websites.