How Does Fed Policy React To Stock Market Declines?
The equity market sell-off since the beginning of October has led to questions around whether the Fed will maintain its current path of rate hikes. Historically, the Fed appears to have responded with more accommodative policy after stock market sell-offs, on average (Exhibit 1). This has led some to conclude that there is a Fed “put,” in which the Fed responds to large stock market declines with accommodative policy, but does not change course when faced with small declines or increases in stock prices.
The Fed might react to stock market sell-offs for two reasons. First, stock price declines can serve as a financial market signal of lower growth in the future. Second, stock price declines and FCI tightening can cause a drag on future growth, for instance through wealth effects on consumer spending.
We find that stock market sell-offs are more likely to worry the Fed if they occur in tandem with a broader tightening of financial conditions or in an environment of weak growth. Looking at the sample of FOMC meetings that follow stock market sell-offs, we find that the Fed is more likely to decrease the fed funds rate when credit spreads are widening simultaneously (Exhibit 2, left). We also find that the Fed is more accommodative when current growth is weak relative to potential (Exhibit 2, right). Such cases of stock market sell-offs correspond to times of elevated recession risk, suggesting the possibility that the Fed responds more aggressively to address downside risks. Today, in contrast, credit spreads have widened only moderately, and growth remains significantly above potential, with limited recession risk over the next year.
In fact, the Fed responds to stock market declines differently depending on whether the economy is currently in recession or not. Exhibit 3 shows that most cuts in the fed funds rate after stock market declines occurred in recessions or the immediate recovery after a recession.
The current environment most closely resembles two instances in which the Fed continued rate increases amidst a broader hiking cycle.
In May 1994, the Fed hiked 75bp despite a 6% stock market decline in the prior months, and in August 2004, the Fed continued with a 25bp hike. In both cases, the pace of growth was over 1.5pp above potential and credit spreads did not widen significantly, and the Fed continued to hike in order to slow growth.
The Fed did respond in a dovish way to an equity market sell-off in two notable instances. In September 1998, the Fed cut the policy rate by 25bp, and in early 2016, the Fed held off on further rate hikes. In both cases, growth was below potential. The 1998 cut came alongside a broader financial panic, and late 2015 and early 2016 also saw a significant widening of credit spreads and a significant worry about recession. This left little room for the Fed to tolerate a large decline in the stock market and a corresponding tightening of financial conditions.
Taken together, the evidence from these historical examples and our empirical analysis suggest large differences in the Fed’s response to stock market declines, depending on broader financial conditions as well as growth. With other financial conditions such as credit spreads still at moderate levels, and with growth running well above potential, the Fed is likely to continue with their current pace of tightening despite the decline in equity markets.
This supports our view that the Fed will hike in December, with a subjective probability of 90%. Beyond this, we expect four hikes in 2019 to a terminal funds rate of 3¼-3½%, with risks that are broadly balanced.
Goldman has been far too hawkish this entire cycle in both the US and Australia. We know US growth is going to slow sharply from mid-next year as the fiscal cliff arrives. Even more so with the oil patch now being forced into retreat which drives a lot of capex and wages growth. Add some infrastructure stimulus and you still have a decent outlook for growth but it’s not hugely inflationary.
I’m now in the camp of a couple more hikes then a pause.