S&P 500 Up 5 of Last 8 Fed Rate Hike Announcement Days
In the chart above the 30 trading days before and after the last 119 Fed meetings (back to March 2008) are graphed. There are four lines, “All,” “Up,” “Down,” and “Rate Hike Days.” Up means the S&P 500 finished announcement day with a gain, down it finished with a loss or unchanged. In 119 Fed meetings, there have been just 17 rate increases. These 17 increases are represented by Rate Hike Days. Of the 17 hike days, S&P 500 was down 10 times and up 7 times with an average gain of 0.19% on all 17. On the day after the last 17 rate hike announcements, S&P 500 has declined 0.74% on average.
Eight rate increases have occurred during the most recent tightening cycle that begun a little more than 1 year ago. During the current cycle S&P 500 has been up 5 times on announcement day and down 5 times on the day after. Of the 16 days examined, all but two had moves more than 1%. This would suggest more volatility is on tap for Wednesday and Thursday.
The Fed has already raised rates twice this year and according to CME Group’s FedWatch Tool, there is currently a 100% probability that they will raise rates again tomorrow when their meeting concludes. Up until recently, the Fed was widely expected to raise raises by 0.50, but there is currently (~3 pm est) a 94.5% probability of a 0.75 increase. Regardless of the actual magnitude of the rate increase, the bond market has already done much of the Fed’s work for it already and the opportunity to “get ahead” of inflation seems long gone now.
In the chart above the 30 trading days before and after the last 113 Fed meetings (back to March 2008) are graphed. There are four lines, “All,” “Up,” “Down,” and “Rate Hike Days.” Up means the S&P 500 finished announcement day with a gain, down it finished with a loss or unchanged. In 113 Fed meetings, there have been just 11 rate increases. These 11 increases are represented by Rate Hike Days. S&P 500 trading leading up to recent hike announcements has clearly been different with a bearish bias in the 10 trading days before the announcement. Of the 11 hike days, S&P 500 was down 7 times and up 4 times with an average gain of 0.23% on all 11. This year’s rate hikes were well received by S&P 500 with gains over 2% on March 16 and May 4. On the day after the last 11 rate hike announcements, S&P 500 has declined 0.69% on average.
In one week the Fed will gather for its seventh scheduled meeting of 2019. The Fed has already cut rates twice this year and according to CME Group’s FedWatch Tool, there is currently a 92.5% probability that they will ease another 0.25% next week. In the chart below the 30 trading days before and after the last 93 Fed meetings (back to March 2008) are graphed. There are three lines, “All,” “Up,” and “Down.” Up means the S&P 500 finished announcement day with a gain, down it finished with a loss or unchanged. A green box has been drawn shading the S&P 500’s average performance five trading days before announcement day. On average the period is marked with sideways to flat trading regardless of what occurs on announcement day. Barring any major earnings or corporate disappointments, the market could spend the next five trading days in a narrow range.
Rate Cut Inbound: A Tweak, A New Cut Cycle or Merely a Placebo?
Tomorrow the FOMC will meet for the fifth time this year. Unlike previous meetings this year, it is widely anticipated that the Fed will cut its rate by at least 0.25% and possibly as much as 0.50%. Growth, measured by U.S. GDP has slowed, corporate earnings have also slowed, and inflation is running below target, but employment remains firm, risk of recession is not high, and stocks are trading near all-time highs. Ahead of this highly likely cut, historical data has been sliced and diced into numerous different variations. Below we present a straight forward look at the recent history of the S&P 500’s performance following any rate cut.
Since October 1, 1982 there have been 80 interest rate cuts by the Fed (there also was 81 increases). On average, 1-month after the cut the S&P 500 was higher 57.5% of the time with an average gain of 0.77%. At 3-months, the frequency of gains improved to 60% and the average gain swelled to 2.21%. This trend of expanding gains and frequency of gains persisted at 6 months and 1-year after the cut. Of note is since 2001, (dot-com bubble bursting and financial crisis) the impact of rate cuts has not been as strong as it was in the prior two decades.
Click here for full table in new window...
In this next chart, we have plotted the average S&P 500 performance 30 trading days before and 60 trading days after a rate cut and a rate increase. Here the initial effects of both a cut and an increase are surprisingly not all that different. In either situation S&P 500 was higher 60 trading days later. A rate cut did have the edge when it came to average performance, but the difference is not all that large.
A 0.25% cut would move Fed policy closer to market rates, a 0.50% cut even closer still. Nonetheless, how significant the impact will ultimately be is unknown. Historically speaking, rates are quite low already. Stock dividends and buybacks have been growing nicely and are likely to continue to do just that whether the Fed cuts 0.25% or 0.50% or not at all.
On Wednesday another Fed meeting will come to an end and another statement regarding monetary policy will we released. As of today, CME Group’s FedWatch Tool is showing a 93.8% chance the Fed will raise rates to a new range of 2.00 to 2.25%. There is also a small 6.2% chance that the Fed raises the range by 0.50%. If the Fed broke from recent tradition and hiked by 0.50%, the market would likely not respond well to the sudden change. Historically down announcement days have been better buying opportunities than positive announcement days.
In the chart below the 30 trading days before and after the last 84 Fed meetings (back to March 2008) are graphed. There are three lines, “All”, “Up” and “Down.” Up means the S&P 500 finished announcement day with a gain, down it finished with a loss. Note how past down announcement days have, on average, enjoyed the best gains over the next 30 trading days.
Of the last 84 announcement days, the S&P 500 finished the day positive 48 times. Of these 48 positive days S&P 500 was down 28 times (58.3%) the next day. Of the 36 down announcement days, the following day was down 21 times (58.3%). All 84 announcement days have averaged 0.39% S&P 500 gains while the day after has been a net loser with S&P 500 declining 0.31% on average. Since the December 13, 2017 meeting, S&P 500 has declined on five of the last six announcement days with an average loss of 0.23%.
The fog started rolling in around my home last night and it looks a lot like this as I am writing. The murkiness got me thinking about how hard it is to find clarity when forecasting, especially with something as complex as inflation. I do not envy Janet Yellen and her FOMC teammates as they try to look through the fog to forecast job growth, economic growth and inflation. The major focus lately has been on inflation with the unemployment rate creeping towards historic lows.
So what does Chairman Yellen and her crew see? They have consistently told us that they expect inflation to move to long run expectations near 2% soon. But how long can you say soon before it turns into a career? Some of the messaging is clearly targeted at raising expectations in the market place. Expected inflation begets actual price rises, is the theory. But what does the data say?
The chart above shows weekly price changes in the CRB Index, a broad measure of commodity inflation. It does not cover everything but includes everything you eat, wear, build with and use to transport yourself. What is it saying? The chart shows a large run higher off of a double bottom in 2001 to a peak at the height of the financial crisis.
A swift pullback followed and a bounce to a lower high in 2011. Since then the CRB index has moved lower, slowly at first then accelerating from 2014 into 2016. The bounce set up a possible reversal with an Inverse Head and Shoulders Pattern as 2016 ended. But that was negated with the mid-2017 dip. Now back testing the prior support area from the double bottom 18 years ago, it is failing to rise again.
A few take aways: 1. Commodity inflation is at historic lows. Outside of the blip in 2016 you need to go back over 20 years to see these levels. 2. Commodity inflation is falling, not rising. 3. Expecting a snapback because it is historically low is not forecasting but guessing.
There is no evidence that inflation will rise. Lets see what the FOMC minutes say Wednesday.
Forecasting GDP in not easy. It involves multiple variables and inputs, many of which are subject to revision. If anybody could do it well, you would think it would be the Fed, right? They have an army of staffers who spend their entire working lives looking at every piece of economic data, trying to attach its significance to their projection. Immediately after the latest rate hike, market participants quickly pointed to the latest release of the Atlanta Fed’s GDPNow indicator, which is supposed to measure real-time projections for the GDP in the current quarter, as evidence the Fed made a mistake. Did they?
Let’s begin by looking at the GDPNow indicator itself. The forecast construction is quite detailed, but essentially it takes incoming economic data and estimates the impact on 13 subcomponents that comprise GDP. The subcomponents are then aggregated to provide a forward-looking GDP estimate. Official GDP figures are released with a substantial lag, reducing the usefulness to policy makers. The GDPNowcast is updated whenever there is new model data available.
The current model forecast is for Q1, 2017 GDP to come in at 0.9%, significantly below private sector estimates, and materially lower than the 3.4% model forecast just 6 weeks ago. How can the Fed raise rates when its own model predicts such a sharp drop in economic activity?
One reason is the FOMC members don’t seem to put much weight on the indicator, despite it is generated by one of the member banks. To complicate the issue, the Atlanta Fed isn’t the only one producing GDP forecasts. The New York Fed has its own model called GDPNowcast. The New York Fed model is predicting Q1, 2017 to be 3.2%, a whopping 2.3% higher than the Atlanta Fed forecast.
Something is obviously amiss. As Mike Shedlock points out in a recent blog post, the two models use different inputs. The Nowcast model takes into consideration data GDPNow doesn’t such as JOLTS job openings, permits, and regional manufacturing surveys. Even when they use the same inputs, it can have opposite effects on the direction of the forecast, which makes no intuitive sense whatsoever. That alone raises credibility issues.
Another concern with the model is its volatility. The GDPNow was forecast current quarter GDP to be 3.4% after the ISM and Construction Spending reports were released on Feb 1. Just six weeks later, it’s predicting 0.9%? Seems like a pretty massive adjustment to me, especially when the economic data, according to the Citibank Economic Surprise Index, has surprised to the upside over the same period.
The bottom line is that people who must make critical decisions on monetary policy know better than to rely on a single forecast, even if it is from a trusted source. Don’t listen to the pundits who present only one piece of evidence when arguing a point. The Fed made the right move.
