Stocks for the Long Run?

This post examines the relationship between stock prices and GDP and considers the long-term outlook for stock prices. I begin with a question: Do swings in stock prices — as as measured by a broad index like the S&P Composite — portend swings in the rate and direction of economic growth?

Here is a chart of constant-dollar GDP, indexed to its value in 1871, which I derived from estimates of constant-dollar GDP that are available at What Was U.S. GDP Then?:

Drawing on the data set for Robert Shiller’s Irrational Exuberance — a data set that Shiller keeps up to date — I graphed the real value of the S&P Composite Index and identified major turning points in the constant-dollar value of the S&P Composite:

A comparison of the two charts suggests that stock prices react strongly to transient events (e.g., shifts in the rate of inflation and consumer confidence), but that there is not a strong relationship between GDP and the daily, weekly, monthly, or quarterly gyrations of the stock market. There is a weak but statistically significant correlation between annual changes in stock prices and GDP, which is evident in the following comparison of stock prices and GDP, dating back to 1950, which I derived from historical prices of the S&P 500 (available here) and estimates of constant-dollar GDP (available here).

In words, the first three graphs suggest that stock prices, measured broadly, oscillate jaggedly around the GDP trend, in cycles of irrational exuberance and rational pessimism.

On closer inspection of the long-term trends depicted in the first two graphs, I found some evidence that major turns in stock prices mark major turns in the course of economic growth. I should emphasize that the points of coincidence between major turns in stock prices and economic growth are evident only with long hindsight; I advise against attempts to predict the near-term course of GDP by transitory changes in stock prices, and vice versa.

Specifically, I calculated the changes in stock prices and GDP that occurred between the turning points identified in the second graph above, with this result:

Market turning point (Year and month)	Duration of trend (years and months)	Real S&P change (annual rate)	Read GDP change (annual rate)




1877.06
	28.07	5.2%	4.4%
1906.01
	14.11	-7.7%	1.9%
1920.12
	8.09	20.6%	4.0%
1929.09
	2.09	-44.9%	-9.4%
1932.06
	33.07	6.5%	5.0%
1966.01
	16.07	-5.6%	2.7%
1982.08
	18.00	12.0%	3.3%
2000.08
	11.04	-3.8%	1.5%
2011.12

(December 2011 probably is not a turning point. I include the period from August 2000 to the present only to show that the trend in stock prices since August 2000 has mirrored the trend in GDP growth since that date.)

The relationship between rates of change in real GDP and stock prices seems to be robust:

A similar relationship holds even with the removal of the “outlier” (1929-1932):

Both correlations are statistically significant, despite the small sample sizes. And they are similar to the following (also significant) correlation, which is drawn from the data presented in the third chart above:

The bottom line is that stock prices can decline even as GDP rises. A long-term, downward shift in the real rate of GDP growth is likely to trigger a significant downward shift in the movement of stock prices.

In fact, unless the course of the regulatory-welfare state is reversed, a prolonged downward shift in the real rate of GDP growth is in the works — probably to about 2 percent. At that rate, expect a continuation of the present trend — stock-price “growth” of about -4 percent a year.

If you think that I am being unduly pessimistic, look at the trend for 1906-1920 (the aftermath of Progressivism’s rise) and the trend for 1966-1982 (the aftermath of the formation of the Great Society). What we have today is a vast regulatory-welfare state built on the foundations of Progressivism, the New Deal, the Great Society.

Happy New Year!

* * *

Politics and Prosperity

“Government is a disease masquerading as its own cure.”

Stocks for the Long Run?

Like this:

Related