Monday's market presaged nearly one million government workers being told that they are temporarily but indefinitely without jobs or paychecks. Tuesday's market evaluated the magnitude of the effects of the government shutdown on the economy and business prospects, and shrugged. While government leaders work towards restarting the government, I thought it would be useful to examine what happened to the stock market in the previous 17 government shutdowns.
Obtaining a list of previous government shutdowns, as well as historical prices of the S&P 500 (SPY), I compared the closing price of the last trading day before each shutdown to the closing price of the first trading day after each shutdown. First I thought that government dysfunction is proportional to the length of shutdown and impact on economy, and therefore reflect in the stock market. Figure 1 shows the length of government shutdown plotted against the percentage change in the S&P 500. Although there is a small relationship between the shutdown length and market decline, the relationship is weak and likely not useful to investors.
Next I considered only the relative performance of the S&P 500 during the shutdown. That is, I took the proportion of the yearly change of the S&P 500 for each shutdown, and compared the actual market change in that time. Those data are plotted against shutdown length in Figure 2. Although the relationship is somewhat stronger than the uncorrected data, it is still a weak and likely not useful for prediction.
Seeing a lot of scatter in the data for shorter shutdowns, I next examined whether there exists a stronger relationship between shutdown length and market change for shutdowns exceeding 5 days. The results are plotted in Figure 3. This figure uses the corrected percentage change in the S&P as in Figure 2. However, there is still no strong correlation.
Lastly, let's consider whether the market performance during a government shutdown is predictive of subsequent market performance. For Figure 4, I compared the change in the S&P 500 during the government shutdown to the change in the S&P 500 during the subsequent 1-shutdown length (if a shutdown were 22 days, I compared the market performance in the 22 day shutdown to the market performance during the following 22 days). In this case, a reasonably good linear relationship exists-with an R2 of 0.47, nearly half of the market variance after a government shutdown can be explained by the market performance during the government shutdown. And the direction of the market is predictive as well - when the market declined during a government shutdown, subsequent sessions resulted in the market falling in eight out of ten cases. When the S&P 500 increased during a government shutdown, the S&P continued to increase after the shutdown ended in all seven cases.
These figures convey three important lessons to learn about the relationship between government shutdowns and broad market performance as shown by the S&P 500. First, no strong relationship exists between the length of a federal government shutdown and the market's performance during the shutdown. Second, the market declines in a majority of both long and short government shutdowns. For short shutdowns (≤ 5 days), the market declined in five out of eight cases, and for long shutdowns (> 5 days), the market declined in six out of nine cases. Third, and perhaps most interestingly, market performance during a government shutdown is predictive of market performance after a market shutdown. Investors may be able to profit after the government reopens if the market increases during the shutdown.