How well have stocks done over 109 years?

The year 2008 was a terrible one for investors in stocks. The markets fell so hard and fast that investors started to doubt the fundamentals of economic theory, including accepted principles on long-term investments. Pundits wailed that it was the end of the era of stocks. Or of diversification.

But were they right? One of the findings by Nobel laureates Daniel Kahneman and Amos Tversky is that people tend to leap to conclusions based on inadequate samples. Kahneman and Tversky are Israeli psychologists who won the Nobel in 2002 for breakthrough studies on behavioral economics.

In the case of the stock market, the sample should be taken over as long a period as possible. That can be done using data compiled by three British economists: Elroy Dimson, Paul Marsh and Mike Staunton, who tracked share prices in 17 markets from the start of 1900 to date.

The markets they covered constituted 90% of the global trade in stocks, in terms of market capitalization during the last 109 years, 1900 to 2008. The database has full information about every price of every share in all these 17 markets, as well as data on dividends.

Using their data, Dimson and his colleagues calculated the returns in each of the markets for each of the 109 years, and the returns of a "global index" in each of the 17 markets.

During the last 109 years, there were six shattering, protracted crises. Two of them were in the last decade.

Two of the crises were world wars. But the data shows that the wars' effect on equities was relatively small, compared with the crises of purely economic origin.

Moreover, Dimson and his colleagues noted that when trouble arrives, different markets fall by different degrees.

During the Great Depression from 1929 to 1931, the global index of the 17 markets fell by 54%. But the real drop in the U.S. market was far greater, 79%.

More recently, we find the great technology crash of the new millennium. From 2000 to 2003, the global index of stocks fell by 44%, but Germany's stock exchange fell by 65%.

In the current crisis, the global index of shares fell 54% in real terms. This time the stock market that suffered the most was Ireland's, which fell 70%.

The conclusion is that even when markets seem to be striding in tandem, there are actually great differences among them, even when they're falling.

Seen over the whole 109 years, the global index of shares generated returns of 5.2% a year, after adjusting for inflation.

The previous reading, taken at the end of 2005 (i.e., of 106 years) found that the global index of shares had returned 5.8% a year.

We can generalize and say that over more than 100 years, stocks returned 5% to 6% a year. That's very high compared with returns on government bonds, which in the last 100 years returned something under 2%.

If we weed out the countries that suffered from hyperinflation in the early 20th century, the returns on government bonds run at 2% to 2.5%, still well below the returns on stocks.

Looking at returns from a geographic perspective, we see some countries where even over 109 years, returns on stocks were dismal. Belgium is one such: From 1900 to date its average return a year has been 1.9%. Italy's no better. Germany's markets managed to return an average of 3% a year and France's generated 3.2%.

There are a number of reasons for these European nations' poor returns over the last century-plus, one being periods of high inflation. Between the two world wars, Germany suffered from hyperinflation that completely flattened any returns on stocks and bonds. Indeed, if we narrow our gaze to the bond markets of those countries, we find that yields on bonds over the 109 years were even lower: Belgium generated negative returns of 0.2% and Italy's result is minus 1.8%.

Moreover, the returns that the three British economists calculated were adjusted for inflation in dollar terms. In countries where the currencies badly weakened against the U.S. dollar (as happened with the Italian lira), dollar returns on stocks were very low, even though in terms of the local currency the returns may have been respectably high.

Over a century, returns may be quite stable. But in short periods of time, stock indexes are highly volatile. In the nine years from 2000 to 2008, the global index of stocks fell 4.5% a year in inflation-adjusted terms.

But again, not all markets behaved the same. During that time, the Irish stock market lost 9.5% a year. South Africa's indexes gained about 7.2% a year. Even during that "lost decade" for investors, people who put their money into South Africa, Norway or Australia did well.

During the decade before the "lost" one, 1990 to 1999, stock markets boomed: The global index gained 7.9% a year. But guess what. Not all markets behaved the same. Japanese investors lost 5.8% a year while Dutch investors gained 17.5% annually.

In other words, do not shrug off geographic diversification, whether the markets are in bear or bull mode.

Another lesson from these figures is that in the investment world, 10 years is not a long time. Even during the 19 years from 1990 to 2008, investor received real returns of 1.8%, which paid them no better than government bonds.

The main reason returns were so low during that time is Japan, where stocks spiraled down by 5.8% a year.

But if we extend our measurement period to 29 years, from 1980 to 2008, the returns on the global index rise to 5.7%, which is very close to the long-term average over a century.

Over 30 years, we find that it paid to be in stocks despite the horrible financial crisis toward the end.

To drive home the point, it takes 30 years for an investment returning 2.5% a year to double in terms of the investor's purchasing power. But if the return is 5.5% a year in 30 years, the investor's purchasing power has quintupled.

Reinvesting the dividends

How actually do we make positive returns on stocks? Because of the sharp fluctuation of stock markets, we tend to assume that we make money when share prices rise and that's all.

Analysis over 109 years shows a different dynamic at work.

It turns out that in all stock markets, without exception, the main component of returns was reinvestment of dividends.

The increase in share price contributed just 1% a year to the inflation-adjusted return on the global index. The other 4.2% originated from dividend yields.

Moreover, if we take the United States out of the equation (where annual real capital gains were exceptionally high), we find the median capital gain of the other markets was zero. In other words, it was all about dividends.

History may or may not repeat itself, but a sample of 109 years is more indicative than a sample of 10 or 20. We can lay down a few basic rules for the long-term investor.