12 Comments

All I am saying is that you should trying to keep the amount of risk you run invariant through your life. If you manage your risk only in relation of your saving, you will inevitably over-expose yourself to the periods of time went your saving are at a maximum.You really have 2 forms of wealth: your explicit wealth and your ability to earn income in the future. If you invest based only on the first you'll run too much risk when you are in your fifties most likely...Imagine you live 100 years: work from 20 to 60 and retire from 60 to 100. You always earn the same wage and save the same percentage. At 60 your saving will likely reach their maximum. Even at 20% equity, you'd be running more risk than on year 1 on 100% equity (since you'd have 40 x 20% = 8 time more equity investment).

Expand full comment

> The diversification you get from investing mostly in equities when young is due to the fact that you tend to have less money when young.

I don't think this is correct.  If you have Y dollars when you're young and X dollars when you're old, then you could always invest the Y dollars as if you were old (i.e. bond-heavy) and X-Y of the X dollars as if you were young (i.e. equity-heavy).

Expand full comment

> the article clearly talks about choosing a type of an investment, if you have only 10 years to live you cannot choose to invest over 40 years so there would be no question at all

The full problem is "what balance should be made between bonds and equities, given one's level of risk aversion".  The conventional wisdom is that young people, who can diversify better across time, are better off with more equities as a fraction of their investment than an older person *with the same risk aversion*

Expand full comment

"the likely future returns to investing in equities" is meaningless without something for comparison. Events that wipe out equities almost always wipe out bonds, too, so such a comparison is not biased and so the equity premium is quite mysterious. Cash is often wiped out by such events, too, and even sometimes human capital. 

Comparing investing in one country to investing in many countries is a real comparison, though I don't know why he leaves out eastern bloc markets.

Expand full comment

Price/dividend ratios don't wander off to infinity, so either prices or dividends must mean-revert to keep the ratio balanced.  Consequently, even if the statistical evidence for mean-reversion in prices is weak, dividends do not seem to exhibit any mean-reversion at all, so the reversion must be coming through prices.

Additionally, I'm not sure why survivorship bias would be a problem in studying stock return predictability — are markets where returns are forecastable more likely to survive for a long sample?

Expand full comment

"The correct comparison will be between a young person (who invests over, say, 4 decades) and an old person (who invests over one decade)."Is it? the article clearly talks about choosing a type of an investment, if you have only 10 years to live you cannot choose to invest over 40 years so there would be no question at all. It seems to be about the correct choice for people who have many years of life left.

And then there is this:

"Contrary to convention wisdom, even young savers need to diversity across different assets types and countries in order to get that effect and be confident of retiring in comfort: [...]"

Expand full comment

I think this study misses the point.The diversification you get from investing mostly in equities when young is due to the fact that you tend to have less money when young.Assuming that returns across time are roughly uncorrelated, you want to have the same amount of risk in each year of your life (as this time diversification will lower risk without much affecting return). So when young you probably want to leverage as your available assets are small compared to your lifetime earnings and when old you want to de-leverage as your assets are now relatively large when compared with your life-time earnings. Ideally you'd need to calculate your lifetime earnings and risk the same percentage each year.The length of the investment is a minor concern. Not increasing the amount of risk you run compared to your life earnings isn't...

Expand full comment

(Bad formatting because of DISQUS's bad interface.  My comment starts at "I don't think...")

Expand full comment

>  from a purely statistical viewpoint it really shouldn't matter whether I invest $100 in a one year investment every year for 10 years or I invest $100 once in a 10 year investment.

>So isn't this "enduring question in finance" just the result of a misinterpretation of statistics?I don't think so.  The two options you describe both stretch over 10 years.  The correct comparison will be between a young person (who invests over, say, 4 decades) and an old person (who invests over one decade).

Expand full comment

So, we're measuring risk by variance in stock prices, while recognizing that the real risk is in things like "war[s], government interference, and financial collapse, some stock markets disappear[ing] altogether"

Before making further conclusions about portfolio management, wouldn't it make sense to take a second look at our definition risk?

Related: has anyone done for finance what psychologists did for economics? That issue of risk sounds as dicey as homo economicus.

Expand full comment

Young people have streams of future regular earnings that are much larger than that of old people.  So they start from a more diversified position, and are in a better position to absorb risk if the expected returns justify it.

The multi-trillion dollar question, of course, is whether today's stock market has sufficiently high expected returns...

Expand full comment

"A lesser known result is that a broader and representative sample of stock histories shows that investing over long time horizons doesn’t reduce the variability of your return."Is that so weird? I mean inflation will make long term investments seem better, but from a purely statistical viewpoint it really shouldn't matter whether I invest $100 in a one year investment every year for 10 years or I invest $100 once in a 10 year investment. Long term investments in technology and infrastructure may be better for society overall, but a) most private investment is in speculative hot air and land/noble metals, not technology and infrastructure, and b) the original investor probably won't receive the additional gains to society in monetary form and hence they won't be included in his accounts. So isn't this "enduring question in finance" just the result of a misinterpretation of statistics?

Expand full comment