New research shows that equity investors still win, over the very long term, but there can be long periods of underperformance.
Shares are risky, and we all knew that even before the credit crisis beat it into us. And because they're risky, we pay less for them and expect that over the long term we'll be more richly rewarded for owning them, benefiting from the equity risk premium.
It makes intuitive sense that this should be the case, and returns over the decades seem to support this argument. But how about returns over the centuries? A couple of hundred years is an insignificant period in terms of our evolution, so is it any less relevant to consider how humans weighed the elements of risk and reward in the nineteenth century?
In a recent article in IndexUniverse, Rob Arnott, who pioneered the idea of fundamental tracking, has gone back as far as 1802 in his analysis of returns on the US markets. The results will raise a few eyebrows:
- Over this 207-year period, risky shares outperformed risk-free (for the sake of argument) government bonds 150-fold;
- A dollar invested in equities at the start of the period would now be worth $4m, while a dollar in government bonds would be worth $27,000, with returns re-invested in each case; and
- This outperformance equates to an excess return of 2.5% per year for shares over government bonds -- considerably less than the 5% or so often assumed to be case.
But what may be more interesting is the number of long periods over which equities underperformed:
- From 1803 to 1857, shares underperformed government bonds, and from that point it took another 14 years for them to catch up. That's a 68-year period during which shares were lagging behind;
- The cumulative returns from government bonds were ahead of those from stocks again from 1929 to 1949, and from 1969 to 2009; and
- These periods include times when shares were growing faster than government bonds, but in which they were still catching up after the relative underperformance of previous years. For 173 of the past 207 years, shares have been either falling relative to gilts, or clawing back those losses.
Ignoring dividends and the comparison with government bonds, and looking purely at share prices and inflation, the crash from 1929 to 1932 actually brought inflation-adjusted share prices back below the level of 1802. That's a 130-year period after which share prices were the same in real terms.
The same happened in 1982, when inflation-adjusted share prices fell to a level first seen at the peak of 1905.
So what conclusions do I draw from this?
- To reiterate: shares are risky. If you're not comfortable with that risk, then perhaps they're not the best investment for you;
- Historically, investors have been compensated for that risk, to the tune of 2.5% per annum over the very long term. It's less compensation than many would have expected, but in my view it's worth having;
- An ability to time the markets would be handy, but in the absence of a fully-functional crystal ball we can still profit from regular investing and the passage of time; and
- Dividends matter. Even after those long periods when inflation-adjusted stock prices made no progress, investors in businesses still pocketed their share of the profits.
So while the data is interesting and perhaps surprising, it serves to reinforce many of the Foolish principles. It's also quite frightening to contemplate the length of time that an equity investor can be out of pocket, but as Peter Lynch says, "the key to making money in shares is not to get scared out of them".
> If you haven't become scared of shares, you can invest for as little as £10 per trade with The Motley Fool Sharedealing Service. It's free to open an account.
