How to estimate S&P 500 downside risk

As we approach Q3 earnings season, the Street cut the bottom-up S&P 500 forward 12-month EPS estimates by an estimated -0.41% as recession anxiety begins to rise. The good news is the market is trading at a forward P/E of 15.8, which is below its 5- and 10-year averages.


In light of the heightened recession risk, what’s an appropriate valuation for the S&P 500? What’s the downside risk from here?



What’s the “right” valuation?

Forward P/E ratios don’t exist in a vacuum. Stocks have to compete with other asset classes, such as risk-free Treasuries, for investment funds. With that in mind, I went back to 1995 and looked at the last time the 10-year Treasury yield traded at the current levels.


There were two periods when the 10-year Treasury was trading at similar yield levels. The first was 2002-2003 when the market bottomed, and 2007-2011, which was the period just before and after the GFC. There are some key differences. The 2002-2003 episode was during a period when yields were falling, while the GFC episode featured yield volatility. Yields are rising today.


Nevertheless, we can see that the S&P 500 forward P/E ratio traded in a range of 15-17 during the 2002-2003 period and 13-15 during GFC, if we exclude the panicked valuation during the 2008-2009 bottom.



Here is where we stand today. FactSet reported that the S&P 500 forward P/E as of last Thursday’s close was 15.8, or a forward 12-month EPS of 237.85. Negative earnings growth varies considerably during recessions. The average decline since 2001 is -22%, which is a slightly optimistic estimate since the sample includes 2007, which was just the start of the recession, and 2015, which was a growth recession.



If we pencil in a -20% decline in forward EPS estimates and apply the 15-17 2002-2003 multiple range to the S&P 500, we get a downside potential of 12% to 22% from Friday night’s closing prices. Applying the 13-15 multiple range of the GFC experience, the downside potential is 22% to 33%.


What if there is no recession? Assuming that earnings estimates don’t change, the range is a downside potential of -16% and upside potential of 10%.


In conclusion, rising yields are putting pressure on S&P 500 valuations. If investors were to assume an average recession, the downside potential could be as high as 33% from current levels and a peak-to-trough range of -33% to -48%, which would take the S&P 500 to either the top of the pre-pandemic peak or the levels of the 2018 Christmas Eve panic. On the other hand, the market is roughly fairly valued if the economy were to sidestep a downturn.



11 thoughts on “How to estimate S&P 500 downside risk

  1. The no-recession upside calculation of 10% compared to multiple bad to horrendous downside estimates doesn’t inspire any greed to offset the fear.

  2. IMO, the only buying going on today comes down to:

    (a) Shorts locking in profits.
    (b) Underwater longs attempting to mitigate losses.

    Speaking of which, there are too many traders under water – and with each leg down they begin to concede and will accept greater losses – before a modest loss transitions into a large loss. So any rally attempts are met with selling at lower highs.

    Margin calls may create a ST bottom – but it’s unlikely to represent a LT bottom.
    trying to

  3. Three instances in 43 years. May be the bulls are hanging on to this observation by QE Rob Hanna (buy on close on Monday? Works on Black Monday and in 1979 too) :
    Breadth by multiple measures was extremely strong in mid-August, & has collapsed since then. % of $SPX stocks > 50ma has dropped from >= 90% to <= 5% within 50 days for only the 3rd time ever. Others were 10/22/79 and 10/16/87.

    1. Or may be not and 3 instances in 43 years or so may not be significant. The intraday breadth strategy did not detect any long entries today and in fact sold the Friday long position and is now short ES from earlier 10:07 ET sellshort 3712. If there were any large movement in market internals, this model should have detected it by this afternoon. The upper resistance for the ES is now at 3736 and this level is likely to be tested in the coming days.

      1. This is a case of market breadth catching up to price which gapped up overnight. It is likely there will be sufficient breadth reversal intraday to flip the strategy to long around 11:30 ET. After that, look for breakthrough of ES 3726 resistance level and any sign of selling there. The last time ES tried to gain above its higher resistance near 3921 on FOMC day saw more than 2 days of large losses in SPX.

  4. A sobering reminder of what’s potentially ahead given the slowing economy and hawkish outlook globally by central banks (though some have stopped raising rates, in particular, Brazil).
    Key assumptions are that Fed will not stop until core PCE is close to 2% as currently measured and Fed’s forecast is good, which has not been the case at all.
    (Shelter is measured by equivalent rents, not housing prices as in many European countries. If one were to look at it based on housing prices, the inflation was much higher last year and earlier this year; falling rapidly as home prices fall)
    Regardless, the path to the bottom is unlikely to be a straight down elevator.
    So, as Cam suggested on Sunday, wait for a technical bounce to exit the market or when the trend model turns bearish.
    Nonetheless, treasuries are providing an alternative.

  5. Making a call or posting about a prediction here is cheap, but trying to pick the bottom time and time again is not. This market feels more like even the people who are usually in control, are not. The bounce attempts are getting sold one after another. Cam’s bottom spotting model has usually been a reliable one, at times it may have been early by a day or two, but here it just doesn’t seem to be working. I assume this by itself should be a warning indicator.

Comments are closed.