Markets are in a skittish mood, and in my experience, skittish markets are noisy. Because we are donning our tin hats, we listen to every half-sensible macro explanation, ponder it, then listen to some other bit of doom-laden narrative and sit tight in cash. That, friends, is the way of the market timer.
As an investment observer I would love to say that you should ignore those voices and noise and only ever think long-term, but that would leave me with dangerously little to pontificate about! So this week, I thought I would instead share with you my brutally simple way of thinking through stock market valuations. And show why we have more pain to come.
Let’s be honest. The $64tn question at the moment is whether we are closer to the trough than we were to the peak. I could throw at you all sorts of complex models involving regime changes, inflation and so on, but over the many decades of following US markets, I have settled on a simple heuristic. The market is one giant short-to-medium-term valuation machine that simply weighs up the likely incoming earnings per share, works out a relatively sensible multiple (comparing it to risk-free alternatives) and then in that weird collective hive mind decides whether stuff is cheap or expensive. And the prime driver for this exercise in the art of sticking one’s finger in the air and playing around with a few, actually quite simple, numbers is the benchmark S&P 500 index, tracked by more ETFs than I have had hot dinners.
I am now convinced that in our constantly adaptive markets (to borrow Professor Andrew Lo’s heuristic) we use the S&P 500 as our benchmark for valuation – so if we think it is in aggregate cheap, we feel optimistic and momentum rallies. And vice versa. So, although bottom-up analysts would like to tell you that individual stock valuations determine what investors regard as ‘reasonable,’ I’m of the view that aggregate index valuations are actually what investors care about, especially as there’s usually a vast discrepancy between what bottom-up analysts expect for earnings and what top-down types plug in. The S&P 500 aggregate valuation ‘sets the mood’ for the wider global markets.
So, let’s lay out a few basic facts. We are in a bear market, of course. And as recent notes from Hartford Funds remind us, there are some useful reference points for bear markets. These include the following:
- Stocks lose 36% on average in a bear market. By contrast, stocks gain 114% on average during a bull market.
- There have been 26 bear markets in the S&P 500 Index since 1928.
- Bear markets tend to be short-lived. The average length of a bear market is 289 days, or about 9.6 months.
- Every 3.6 years: That’s the long-term average frequency between bear markets.
- A bear market doesn’t necessarily indicate an economic recession. There have been 26 bear markets since 1929, but only 15 recessions during that time.
With these familiar factoids out of the way, let’s return to current markets and the S&P 500 itself. Current predictions for the cumulative earnings of all 500 companies in the benchmark index, according to Ed Yardeni, is 228 per share for the aggregate S&P 500 for FY 2022 and 249 for FY 2023. As an aside, Yardeni is much more sceptical and plugs in 216 for FY 2022 and 235 for FY 2023. I’ve seen other estimates for FY 2022 EPS of as much as 248 but for now let’s use Yardeni’s numbers. For the record, current 12 month EPS for the S&P 500 at March 22 was a measly 204.
Those forward consensus estimates for earnings already have some fairly heroic assumptions built in. If we look at Factset’s most recent earnings round-up we find that ‘analysts believe net profit margins for the S&P 500 will be higher than Q2 2022 for the rest of this year. As of today, the estimated net profit margins for Q3 2022 and Q4 2022 are 12.9% and 12.7%, respectively. For CY 2022, analysts are projecting earnings growth of 9.9% and revenue growth of 10.7%.’
Now I don’t want to be a party pooper, but do you not think that collectively these bottom-up analysts have failed to get the message? We may already be in a recession at worst or a slow-down at best, and virtually everyone including my aged mother thinks that everything is grim at the moment. I mean, if I were cynical, I might even assume that that it is precisely what the Federal Reserve thinks at the moment, and wants us to think as well.
And just in case you believe I’m the only person thinking these slightly bearish thoughts I’d point you in the direction of Vincent Deluard from StoneX, who observes that bottom-up analysts expect S&P 500 index earnings growth to accelerate to 11.1% in the fourth quarter. In fact, these crazy numbers are based on cash-flow models which predict that margins will expand in 2023 due to a disinflationary boom. Take the specific example of the Consumer Staples sector, where analysts expect margins to increase up to an all-time of 14.3% next year. This rather contrasts with the fact that just 14% of S&P 500 companies issued a positive guidance for this quarter and 44% issued negative guidance, the most bearish outlook since Covid started. I think the word we are all searching for here is ‘disjuncture’.
Now, we can of course peer back into the record books and see what happened in previous years when aggregate earnings per share for the S&P 500 slowed down or actually fell. I’ve gone back and tabulated total earnings per share for the S&P 500 since 1945 and the composite price-to-earnings ratio, also since 1945. In addition, I extracted the years in which either measure declines, ie when earnings per share for the S&P 500 declines and years in which the PE ratio for the index declines. I could have gone further back to the 19th century but I’m not entirely sure that that is a useful exercise.
So. What does this very simplistic data exercise tell us?
Over the previous 76 years, we’ve had 27 years in which the S&P 500 EPS declines in value. The simple arithmetic mean is a decline of 15.4% in that year with a standard deviation at 16.7, which implies that the one-standard-deviation event is a 32% decline. The median decline was 11.4%.
Next, moving to the years in which the PE ratio declined, we have 39 years where the PE declined – thus proving that not all S&P PE declines are caused by an earnings recession (sometimes markets just get overextended). The simple average decline in the PE ratio is 16.4%, with one standard deviation at 13.7 (implying a one-standard-deviation outcome of -30%). The median decline is 12%.
Let’s translate these to the market at the end of March 2022 when earnings per share was 204. Currently the PE on the S&P 500 is 20. Given these numbers, I think it’s fair to assume that a 15% decline in earnings per share is possible by the end of 2022 – taking the S&P 500 earnings per share down to around 170, while the PE multiple might decline to around 17. Alternatively I could use Yardeni’s own FY 2022 EPS estimate (216) and after knocking of 15% earnings reduction, and I’d end up at 185-ish. Maybe, just maybe, the consensus for FY 2022 is right – though I severely doubt it – and after a 15% reduction we end up around 195. To add an external opinion to the mix I would also point to Deluard at StoneX, whose own simple macro model predicts that ‘EPS should decline by 5% in the next 12 months [to $189 per share] due to rising borrowing costs, oil prices and the strong USD. Buybacks, which soared to a record $263bn this past quarter, would be the first victim of the coming profit squeeze.’ Conveniently, 189 is comfortably in my 170/195 range.
If we apply a PE of 17 to these reduced earnings – between 170 and 195 – we end up with a range between 2890 and 3315 for the S&P 500. We’ll come back to that in a second after we’ve taken a slight detour on what constitutes a ‘reasonable’ valuation.
That 17 times earnings is smack back in the middle of most investment banks’ estimates for long-term forward PE ratios, including JPMorgan analysts. They recently released a fascinating piece of analysis which suggested that the long-term average forward PE ratio for the past 25 years was, drum roll please, 16.85 times earnings. And just in case the crowd is wrong, let’s also take what I call the common sense test. My hunch is that in zero-bounded interest rate environments, PEs can probably expand out to a range of between 20 and 25, largely because deflation helps keep earnings growth steady and upwards, pushing profit margins ever higher. Once we head back into a more inflationary environment where risk is greater as is volatility of earnings (stop-start economies), then a range of between 15 and 20 seems more logical. Or perhaps, more precisely, 17 times earnings.
Armed with these numbers – EPS for the S&P 500 ranging between 170 and 195 – and PEs likely to be in the 15 to 20 range, we can go back to our very simple model. Which, to repeat, spits out a range between 2800 and 3350 for the S&P 500 very roughly. And where are we today, after the recent falls? As I write the S&P is at 3944, representing a 17.8% decline from its peak at the start of the year (roughly 190 days ago).
I shall leave it to the reader to draw their own conclusions from this exercise, but I think my simplistic way of looking at markets and specifically the S&P 500 is shouting one message at you – we’re not there yet. That is to say, there’s more pain to come as the bottom-up EPS numbers play catch up with the increasingly obvious slowdown. And a decline past 3350 (another 15% decline from current levels) would give us a peak-to-trough decline of over 30% which approximates to the historic norm.
Allow me to add one last coda. I’d echo the recent observations of Michael Hartnett, the chief investment strategist at Bank of America, who reckons the S&P 500 at 3,600 would be grounds to ‘nibble,’ at 3,300 would be room to…