Monday, May 11, 2009

Stock Indexes Outpace the Economy

Historical data show that stock market indexes grow faster than the economy as measured by Gross Domestic Product (GDP). On the surface this seems impossible. The stock market is part of the economy, and while it may grow faster than the economy for short periods, it makes no sense that it could outperform GDP over the long term, but this is what seems to happen.

According to data from Angus Maddison, from 1926 to 2000, Canadian GDP per person grew by a factor of 5 above inflation. Because the population grew as well, overall GDP rose by a factor of 16 above inflation. The U.S. had a similar experience with per person GDP rising 4.3 times, and overall GDP rising just 10.3 times.

However, according to Ibbotson and Associates, over this same period of time (1926 to 2000), large stocks with reinvested dividends rose by about a factor of 300 above inflation, and small stocks rose by a factor of 700 above inflation! This means that large stocks outgrew U.S. GDP by about 4.7% per year, and small stocks outgrew GDP by about 5.9% per year.

This sustained outperformance by stock indexes over GDP growth seems impossible. So, how could it be? The short answer is that GDP is real and stock indexes are not. Extremely little money was placed in stocks indexes in 1926 and left there until 2000. Every dollar invested in large stocks in 1926 would have grown to about $3000 in 2000 (but only worth about one-tenth as much due to inflation), but very little money actually sat in the index untouched that long.  On average, all dividends get spent.

How do stock indexes outperform the average investor? Typical investors fail at market timing efforts and put large portions of their money into inferior investments like bonds and cash.
Another interesting part of this paradox is that if a significant number of investors suddenly bought into the index and left their money there for an extended period of time, and companies stopped paying dividends, stock indexes would cease to outperform the general economy because stock index returns would become real for a large proportion of wealth. It is the very fact that few people take advantage of the power of the index that makes it such a good investment.

In his 2007 letter to shareholders, Warren Buffett ridiculed those who project the Dow Jones Industrial Average (DJIA) to increase by 5.3% to 8% per year because this means that the DJIA will reach 2,000,000 to 24,000,000 by 2100. The implication is that these large figures are clearly impossible.

But, unless investor patterns change, the DJIA will continue to be just theoretical and not represent real money because few people will buy the DJIA and forget the money for 100 years. If investors continue to underperform the DJIA, it would easily reach crazy-looking heights.

I should note that Buffett was using his argument to criticize two groups:

1. Companies who make overly-rosy predictions of returns on pension funds to justify underfunding pension plans.

2. Financial advisors who sell potential clients with overly-rosy predictions.

These two groups deserve criticism. The rosy predictions are impossible because they are based on growing real money. While the DJIA could reach into the millions by 2100, the bulk of real investment dollars cannot.

We tend to think of index investors as those who are satisfied with being average. But, buy-and-hold index investors are a small minority who, on average, have outperformed the average investor by a wide margin.


  1. I am a big proponent of stock index funds as the best way for most people, including me, to get exposure to the stock market as part of their long term savings plans. And yes, re-investing the dividends is extremely important to building wealth as you explain.

    However, just to be clear, a stock market index outperforms GDP growth only if you re-invest the dividends.

    If you do not re-invest the dividends, then the growth in a stock market index should match GDP growth. In other words, the stock market is real and reflects underlying economic growth.

    This assumes P/E and P/D ratios stay the same. It also assumes the pool of common stock is niether diluted by companies raising capital but issuing new stock, or concentrated but companies using their profits to buy back outstanding shares.

    Granted these are big assumptions. Historically over the last century, S&P 500 companies have diluted the pool of stocks by a percent or two per year by raising capital(with periods where there were buy-backs). The P/E ratio has fluctuated greatly over this time period. And the P/D has been impacted by a trend to lower dividend payouts.

    Readers can learn more about this on the website. You need to go back a few years in the posts.

  2. Blitzer68: It's true that it is total returns that have outgrown GDP over the past century. Perhaps I could have reiterated this more often in my article.

    However, there is no reason why a stock market index (without dividends) couldn't outpace GDP if the index is a selected subset of stocks, and these chosen stocks consistently give above average growth relative to all stocks. Because some companies fail and new companies come into existence over time, it is possible that there is no real pot of money that consistently tracks the index over very long periods of time. This would make the index "not real" in the sense that there is no pot of money in it over say a century. Even if index outperformance is only 1% each year, this would cause the index to outgrow GDP by a factor of 2.7 over 100 years.

    So, the overall stock market is real, but a particular index may not be real.

  3. Wow, another eye-opening post. This is why I always read your new blog entries before anyone else's.

  4. Patrick: Thanks for the kind words. All it takes is a comment like this every once in a while to keep me blogging :)

    CC: If we eliminate those who sell part or all of their index portfolios every once in a while in anticipation of market losses, passive portfolios are definitely unconventional.