The Long View for Stocks

Ibbotson Associates is known for their collection of long-term financial returns data. Their latest yearbook isn’t out yet, but I was able to bring the data up to date by using numbers from Standard and Poor’s.

Here’s what the chart looks like of long-term stock returns from December 31, 1925 to December 31, 2014.

The blue line is stocks and dividends but is not adjusted for inflation. Starting with $1, it turns into $5,300 by the end of 2014. That’s an annualized total return of 10.12%.

The red line is the blue line adjusted for inflation. Stocks have averaged a total real return of 6.98% annualized over the last 89 years. That’s roughly stocks doubling in real value every decade.

Of course, there’s a lot of variation in that. Over the last 15 years, real stock returns have averaged less than 2% per year.

Jeremy Siegel often talks about how stocks have returned 7% in real terms over the long haul. I’ve even heard this referred to as the Siegel Constant.

Personally, I don’t think investors should expect 7% real returns going forward. The problem with this past data is that it’s based on the American Century. I’m still a long-term bull but we can’t replicate the success of 20th Century America.

Now let’s have more fun with the data. I took the red line and divided it by a 7% trend line, meaning a line that just rises by 7% per year, every year. So when we divide the two, it means that whenever the line is rising, stocks are outperforming their long-term average. When it’s falling, they’re trailing it. (Even a mildly declining line is still making money for you.) I then made it a logarithm chart so it would be easier to read. Here’s what I got:

A few thoughts: First, notice how cyclical the chart is. Stocks don’t return 7% per year consistently. Rather there are long periods of outperformance followed by long stretches of underperformance.

Second, the 1949 to 1956 bull market was one of the greatest in history. No one talks about that one. I think it’s because it never crashed.

Third, the 1929, mid-1960s and 2000 peaks are all roughly similar. The mid-1960s peak was a rolling one. Things didn’t really fall apart until the 1973-74 crash.

Similarly, the 1932, 1982 and 2009 troughs aren’t too far apart.

Fourth, 1929 to 1932 sucked. Seriously.

Fifth, as strong as the last six years have been for stocks, we’re merely back to our long-term average. Actually, we’re still a bit short of it. Stocks would have to gain about 20% in 2015 for us to be back at the mid-point.

Finally, if you use a little imagination, this chart somewhat resembles the long-term chart of P/E Ratios. This makes sense since earnings have tended to increase at a fairly steady pace over the long haul. Not quite like a 7% trend line, but not too far off, either.

Posted by Eddy Elfenbein on January 22nd, 2015 at 1:21 pm

The information in this blog post represents my own opinions and does not contain a recommendation for any particular security or investment. I or my affiliates may hold positions or other interests in securities mentioned in the Blog, please see my Disclaimer page for my full disclaimer.