Objective research to aid investing decisions
Value Allocations for Apr 2018 (Final)
Cash TLT LQD SPY
Momentum Allocations for Apr 2018 (Final)
1st ETF 2nd ETF 3rd ETF
CXO Advisory

GDP Growth and Stock Market Returns

April 13, 2018 • Posted in Economic Indicators

The U.S. Bureau of Economic Analysis (BEA) each quarter estimates economic growth via changes in Gross Domestic Product (GDP) and its Personal Consumption Expenditures (PCE), Private Domestic Investment (PDI) and government spending components. BEA releases advance, preliminary and final data about one, two and three months after quarter ends, respectively. Do these estimates of economic growth usefully predict stock market returns? To investigate, we relate economic growth metrics to future stock market index returns. Using quarterly and annual seasonally adjusted nominal GDP data from BEA National Income and Product Accounts Table 1.1.5 as available during January 1929 through December 2017 (about 87 years) and contemporaneous levels of the S&P 500 Index (since 1950 only) and the Dow Jones Industrial Average (DJIA), we find that:

The following two charts compare on log scales seasonally adjusted nominal GDP and stock index levels. The upper chart uses quarterly data and the S&P 500 Index for January 1950 through December 2017. The lower chart uses annual data and DJIA for 1929 through 2017. Both charts suggest that the stock market roughly tracks GDP growth, but with greater short-term and long-term volatility.

For precision in comparing series, we relate change in GDP to stock market return.

The next chart summarizes correlations between quarterly S&P 500 Index returns and quarterly changes in GDP and its major components for various lead-lag relationships, ranging from stock market returns lead changes in GDP and its components by eight quarters (-8) to changes in GDP and its components lead stock market returns by eight quarters (8). Results suggest that:

  • Stock market returns lead economic growth positively by one to three quarters. In other words, a strong (weak) stock market return indicates strong (weak) growth in GDP, PCE and PDI over the next few quarters.
  • Growth in economic activity leads stock market returns very weakly by one to five quarters. In other words, strong (weak) growth in GDP, PCE and PDI during a quarter very slightly suggests weak (strong) stock market returns over the next few quarters.
  • The stock market is indifferent to changes in the government spending component of GDP.

Do annual data tell the same story?

The next chart summarizes correlations between annual DJIA returns and annual changes in GDP and its major components for various lead-lag relationships, ranging from stock market returns lead changes in GDP and its components by eight years (-8) to changes in GDP and its components lead stock market returns by eight years (8). Results suggest that:

  • A strong (weak) stock market year indicates strong (weak) growth in GDP, PCE and PDI the next year.
  • Strong (weak) growth in GDP, PCE and PDI during a year very slightly suggests a weak (strong) stock market a few years hence.
  • Again, the stock market is largely indifferent to changes in the government spending component of GDP. There may be something of an opposite relationship for this component compared to the others.
  • There may be a noisy eight-year cycle in the relationships.

The sample period is modest for this kind of analysis (especially the last point).

Might there be some exploitable non-linearity in the GDP-stock market relationship?

The final two charts summarize average S&P 500 Index quarterly returns by ranked fifth (quintile) of quarterly changes in GDP and its major components since 1950. The upper (lower) chart uses stock market returns for the quarter after (two quarters after) the GDP measurement quarter. Since BEA does not release final GDP data for a quarter until the end of the next quarter, the upper chart is idealized (assuming foresight), while the lower chart represents more realistic investor use of GDP measurements.

Lack of trends and consistencies across metrics in the upper chart suggests that, even with foresight regarding final measurements, GDP growth is not useful for predicting next-quarter stock market return.

While trends are also lacking in the lower chart, results suggest that equity investors may want to be cautious during the quarter after final release of very strong GDP, PCE and PDI growth.

In summary, evidence from simple tests on available data offers some support for a belief that strong (weak) growth in GDP and its non-government components may precede a relatively weak (strong) stock market several quarters and years later.

Cautions regarding findings include:

  • Modeling and calculating GDP and its components may be crude compared to calculating stock market returns.
  • The above analyses are in-sample, using all data available across sample periods. An investor operating in real time have drawn different conclusions at different times. Series are fairly short for out-of-sample testing.
  • Testing relationships for multiple variables against the same or similar stock market return data introduces data snooping bias, such that the strongest results overstate expectations.
  • The most recent GDP data are subject to revision as new inputs accrue. Also, BEA occasionally revises GDP data (for example, to update seasonality adjustments). Such adjustments may affect the relationships quantified above. Deciding whether to use original or revised data is essentially a trade-off between real-time experience and best forward-looking model (and original data is often elusive). Using percentage changes in GDP generally mitigates this concern, since revisions have less effect on changes than on levels.
  • Sample periods are short in terms of number of complete economic cycles.
Why not subscribe to our premium content?
It costs less than a single trading commission. Learn more here.
Daily Email Updates
Login
Research Categories
Recent Research
Popular Posts
Popular Subscriber-Only Posts