Has the S&P 500 run too far, too fast over the last few months? It certainly feels that way, with the index up 11% over the last 65 trading days (roughly 3 calendar months). That’s more like a whole year’s typical return, not just a quarter’s performance.
To assess how unusual an 11% 3-month return really is, we pulled the S&P 500 daily return data back to 1990 and ran rolling 65-trading-day returns. Here is what we found:
#1: We’re certainly in an unusual spot.
- Since 1990 3-month returns have averaged 2.2%. Compound that into a year and you get 9.1%, so that makes sense.
- The standard deviation of those quarterly returns is 7.3 percentage points. That means any 3-month return between -5.1% and +9.5% is in the meaty part of a normal return distribution.
Upshot: yes, the current 11% return over the last 3 months is well outside the normal bounds for quarterly S&P 500 performance.
#2: A look at just the last 15 years (2006 to present) shows the S&P 500 has rarely rallied much more than 10% (the 1 standard deviation upside level) unless it is snapping back from a serious prior downdraft in the recent past.
- The chart below shows that 10% is usually a hard cap on 3-month returns, either before or after the Financial Crisis.
- The only real exceptions are mid 2009, late 2010, late 2012, and early 2019. In each instance the S&P had recently pulled back more than 10%.
Upshot: the current 11% 3-month return does not come after a 10% pullback, because at its worst levels in early October the S&P was only down 3% in the previous 30 days.
#3: You have to go back to the 1990 – 2005 experience to find a period when the S&P managed to rally materially further after a +10% 3-month advance without the requisite prior near-term +10% decline.
- Mid 1997 was the first instance, with a peak 25% 3-month return.
- It happened again in Q1 1998, with a 20% 3-month return.
- Important: annualized S&P earnings growth was accelerating during both periods. From Q1 to Q3 1997, it went from 7% to 11%. From 1998 to 1999 earnings growth accelerated from 1% to 20%, so Q1 1998 was looking forward to that improvement.
Upshot: supersized short-term rallies are possible, but they need a strong fundamental tailwind in the form of higher expected earnings.
Pulling all this together:
- Now you know why there’s so much chatter about a near-term pullback for US stocks. The rolling 3-month return math is clear: we’re at the level where stocks typically stall out or drop.
- Since Fed policy should be stable this year and long rates are creeping higher in anticipation of better global growth, we’re not set up for an equity rally based on lower discount rates (as in 2015 – 2017).
- That leaves accelerating earnings growth expectations as the only supporting pillar for any further rally. That’s how it worked in the 1990s, and that’s how it has to work today. Yes, analysts have 9-10% growth in their models for 2020, up from 0% last year. But markets have to be dead-certain that will actually come to pass, and that 2021 can continue the trend.
Bottom line: upcoming earnings reports and management guidance have to be great, not just good, in order to keep the S&P 500 from succumbing to the weight of statistical gravity. The strength of the tape says they will be, supported by strengthening global economic data. We’re in the bullish camp, so you know where we stand on this. Still, we get it… A lot has to go right to see further near term gains. If your own outlook calls for less-than-stellar earnings outlooks, now is a reasonable time to reduce risk and wait for a better entry point.