Tags

In the long run – yes (assuming no failed states).

Do you, guys, remember the Betteridge’s law of headlines?

How do we define a long run?

How about the whole history of the S&P 500 Index.

Since 1871 S&P 500 index has gone up 3,049% (2.4% annual return) or 1,855,212% (6.9% annual return) in real terms*.

However, it took 57 years for the S&P 500 to break firmly above its 1929 high in real terms! That’s more or less, one generation of flat returns.

And let’s look at those 148 years of returns. How were they distributed?

For 115 of those, between 1871 and 1986, the S&P 500 had a total return of about 460% (without dividends – just price change). That’s 1.5% return per year.

Then for the next 33 years, the S&P 500 returned the same 460% in total but this time, the annual return was more than 3x higher at 5.3%.

115 years of return happened in just 33 years!

What exactly happened in the early 1980s?

Let’s just say, there was no miracle.

However, the power of shareholders’ primacy lead to an explosion of US share buybacks; the economy’s financialization lead to massive M&A activity and the reform of executive compensation incentives. All this contributed to the strong returns for US stocks.  

There was, of course a positive correlation to US GDP growth and US productivity growth and US population growth, but a negative correlation to the changes in them (meaning, the lower growth rates of the US economy, productivity and population in the latter period coincided with higher stock market returns).

Were the strong returns post 1986 a payback of the decent returns prior or a pay-forward from the future earnings ahead?

My bet is on the latter given the possibility of simple mean reversion of returns probability coupled with lower real GDP growth rates and declining productivity growth rates. If negative returns on sovereign debt are any guidance, I would not be surprised if what follows is several decades of flat returns overall.

*Source for all data is Robert J. Shiller website: http://www.econ.yale.edu/~shiller/data.htm