EARNINGS
Private forecasters currently put a 65% probability of a recession in the U.S. and higher in Europe in 2023. David Rosenberg says it’s “a lock”, the yield curve says 80% odds and Bloomberg Economics’ recession model is at 100%, with an August debut. Goldman Sachs is a vocal outsider at 35%, more or less joined by a few other prominent brokers.
During “non-severe” recessions, earnings decline about 10% which would drag 2023 EPS from the $219.87 estimate for 2022 to the $200 range.
KKR’s own earnings model calls for a 17% drop in 2023 EPS although its “fundamental, historical and quantitative approaches all suggested around
a 10% decline in earnings.”
Corporate CEOs are pretty downbeat, suggesting weak earnings over the next few quarters:
Yet, analysts are still forecasting a 4.3% gain to $229.24 on a 8.6% revenue gain, back-end loaded. High energy prices/profits boosted total 2022 earnings. Lower energy prices/profits are now expected to hurt them through Q3’23. But who can confidently forecast energy prices these days?
Focusing on ex-E profits, we are currently in a mild cost-push earnings recession that sell-side analysts expect to end in Q2.
They are forecasting an acceleration in revenues in 23’H2 which would presumably require inflation staying above 5%… something the Fed said it will not tolerate. High interest rates will eventually bite on the economy, at least softly, hurting revenues and profits. Can’t have your cake and eat it too.
More objectively, Ed Yardeni shows that revenue growth should normally be heading towards zero, even without a recession. Nowhere near +8.6%!
Note that the December PMI came in at 48.4 from 49.0.
- The Fed is focused on bringing inflation down from its current 5-6% core range to 2-3%. Revenue growth estimates should incorporate this eventual “reality”: the Fed either succeeds, or tightens even more. I doubt Jay Powell wants to shed his Volcker frock for Arthur Burns’.
- Wages are currently rising faster than inflation/revenues (S&P 500 revenues are expected to grow 4.1% in Q4’22, 2.5% in Q1’23 and decline 0.1% in Q2’23).
- Financing costs will grow meaningfully in 2023, across the whole economy.
It thus seems wise to assume no margins expansion, at best.
That said, we may actually not have a recession (see below) in this very complex environment (see Economic Perspectives, Dec 27. 2022), so I will use a range of $200 to $230 for S&P 500 earnings to cover a wide range of possibilities.
VALUATIONS
Conventional P/E:
Twelve month-forward EPS are currently $224.78. At 3900, the forward P/E is 17.4x. On $200, the recession P/E is 19.5x. On $230, the growth P/E is 17.0x.
The good news is that the conventional P/E is back within its long-term range of 15-20 when inflation is muted (i.e. excluding 1972-1992).
- If no recession, this is a reasonably (“fairly”) valued market, taking no account of the inflation risk.
- If recession, 19.5x is in “buy high” territory, just before earnings begin to slide.
Median P/E
This metric can be useful when the weighted averages are inflated by very expensive large weight stocks like currently. I framed this Ed Yardeni chart and added the red dot as of Dec. 28 (17.7).
Note that this chart does not cover high inflation years other than the 1980s. The median P/E would likely decline closer to the bottom of the wider range if inflation stays high or a deep recession erupts.
If we only get a mild recession like in 2001 and inflation recedes towards 2-3%, “fair” value on this metric would be at 3465 (16.5 x $210) and “buy low” value at 3150 (15x).
Price to Cashflow
Based on the last 30 years of low inflation, the S&P 500 P/CF ratio is right in the middle of the range. But beware inflation.
Rule of 20 valuation:
The only model that incorporates inflation.
With inflation at 6.0%, the Rule of 20 says that the “fair” P/E is 14.0 (20 – 6). On trailing EPS ($222.41): 3115, down 20.0%.
With inflation at 4.0%, the Rule of 20 says that the “fair” P/E is 16.0. On forward EPS ($224.78): 3600, down 7.7%.
Assuming a recession, with inflation at 3.0%, the Rule of 20 says that the “fair” P/E is 17.0. On recession EPS ($200.00) = 3400, down 12.8%.
If no recession, using 2023 EPS of $230, only inflation at 3.5% begins to surface “fair value” around current levels. But strong earnings growth and low inflation are an unlikely combo this year.
An hopeful view says that inflation is actually slowing, being in the 4.5-5.0% range per the last 3-4 months. “Fair value” starts around 3400 then.
TIMING
Conventional valuation metrics show large cap stocks have corrected from bubble valuations back within their long-term range, albeit mostly within the top half of the range. On that basis, the extreme risk is behind us and gradual accumulation should be rewarded over the next few years.
“We found that while the approx. -20% YTD return for the SPX looks historically poor, you tend to see a bounce back in the years after similar declines. That said, it is hard to locate one that doesn’t occur during or in the immediate aftermath of a recession.
We would also highlight that negative return years tend to be followed by annual returns that are stronger than the overall average, and the worse negative returns are, the better the performance the next year tends to be. In fact, the last 3 times the SPX annual return was worse than this year’s, the SPX returned over 20% in the following year” (Jefferies)
However, note the double whammy in 1973-74 and the 3-year hell period after the tech bubble even though the Fed was aggressively cutting rates, from 6.5% to 1.7% in 12 months.
A minus 480 points pivot in 12 months and the S&P 500 nonetheless dropped 13%, and another 30% thereafter!
The S&P 500 peaked in August 2000 and only bottomed in September 2002, 46% lower, ten months after the end of the recession, 20 months after the first rate cut and 9 months after rates hit 1.7%. This while inflation was stable around 2.5%. Profits collapsed 32% even though real GDP growth was never negative on a YoY basis in 2001.
David Rosenberg on the mantra that the S&P 500 index very rarely sees back-to-back years of declines.
We are told this has happened only four times in the post-WWII era. But remember — the calendar year is just a fluke of history. The reality is that the S&P 500 hit its peak at the very start of 2022.
When you actually look at “two years” in rolling 24-month intervals, the S&P 500 has actually declined in such a timeframe no fewer than 12 times in the past, all at the hands of the Fed, in the midst and then in the aftermath of the tightening cycles; and 9 of these 12 involved an NBER-defined recession.
If you don’t care about timing, Ben Carlson has encouraging numbers for you:
I ran the numbers going back to 1950 to see what happens to forward returns for the S&P 500 if you would have bought in after it fell 25% from the highs:
You can see all of these instances saw the market fall even further, but future returns going out one, three, five, and 10 years were terrific in most instances. Every period saw positive returns but one 12-month period during the Great Financial Crisis.
Better timing would have saved 23% in 1974 (inflation), 24% in 2000-02 (bubble) and 32% in 2008-09 (bubble). Wasn’t 2021 bubbly, and 2022 inflationary?
Don’t fight the Fed?
Many pundits claim that stocks can be safely bought after the Fed pauses or cuts. Beware, they might be the same experts who secured us all before rates rose last year. Two examples:
- Why rising rates won’t derail stocks – BlackRock
- History shows stocks can weather rate hike cycle – Reuters citing various strategists
Now that “buy-the-dip” is out, “buy-the-pivot” is the new recipe.
Well, history suggests it ain’t so simple. The black circles in these next charts (log left scale) point out periods when equities fell during Fed easing periods:
While it is true that equities eventually tend to do well after Fed easing episodes, “eventually” is the key word here: most times, the Fed eases for “good” reasons. Equity markets are not always in sync with the FOMC and some lags have been very costly.
I found 15 “Fed changes of posture” since 1957.
- From the first pause to the market low, the S&P 500 troughed 9 months after, on average. But the range is -3 months to +31 months.
- From the first cut to the market low, the S&P 500 troughed 6 months after, on average. But the range is -3 months to +21 months.
Actually, the S&P 500 declined after every Fed cut but five (1966, 1980, 1984, 1989, and 1995). Equities dropped between -4.0% to -47.7% (month end data) with an average of -16.1%. If we exclude 1974 (inflation) and 2001 and 2007 (bubbles), the average is -5.9% (range: 0.0% to -19.9%).
And I found no stable correlation with valuations, inflation and profit trends that could help decide when it might be safe to jump in.
A dovish turning Fed then only tells us to reduce our underweight and get ready to buy more aggressively.
Recession or not?
Equity markets have generally proven to be good recession indicators.
But here’s an even better indicator: over the past 100 years, no bear market associated with a recession has bottomed before the recession has even begun.
Recession forecasters, stay put until you’re proven right…
But #1: are we having a recession?
If you rely on economists to forecast recessions, a Magic 8 ball might be cheaper and just as good, or as bad.
In truth, most forecasters do not have an incentive to even predict a recession in part because there may be a greater loss – reputational and other kinds – for incorrectly calling a recession than benefits from correctly calling one.
So virtually all recessions are uncalled the year before they occur, and fewer than 25% get called the year they actually occur. And they generally turn out stronger than predicted (see this IMF paper).
Yet, recession forecasts for 2023 are widespread (many were forecasting it for last year). Must be a slam dunk! … or a “type 2 error”: falsely forecasting a recession. When many take the risk, individual hazard is reduced. The safety of the gang…
Meanwhile, the Atlanta Fed GDPNow for Q4’22 is at +3.8%, nearly double the Blue Chip consensus high estimate.
As John Maynard Keynes once wrote: “The inevitable never happens. It is the unexpected always.”
And Bob Farrell’s rule #9: When all the experts and forecasts agree – something else is going to happen.
But #2: NBER-designated business peaks actually aren’t pronounced until well into recessions, and sometimes even after the recession is over.
For the 5 recessions before COVID, it took the dating committee 8.4 months post their onset, on average, to declare their start (10 months for the last 3 recessions). And if you needed to know their end date, the wait was 12.6 months on average (20 months for the 2001 recession).
So when you are told it was historically wise to buy during recessions, or some x number of months before the end of a recession, remember that these stats assume that the recessions were officially declared in a timely fashion. They never were.
TECHNICAL ANALYSIS
It’s never easy at major turning points. My approach is to heavily rely on valuations (risk management), not only of major indices but of individual equities, and on momentum, economic, profits and technical (timing).
When major indices are approaching reasonable valuation metrics, I am generally able to find an increasing number of attractively valued stocks. On the hunt, I list the most interesting companies on their business fundamentals, measure my risk/reward ratio to set entry points and watch a few key technical indicators, trying to detect changing/positive momentum to avoid falling knives and value traps.
The great Marty Zweig had a few sound advice:
- If the values don’t make sense, don’t participate.
- You’ll never know all the answers.
- The trend is you friend, don’t fight the tape.
Yes, he also famously said “Don’t fight the FED” but he tacked on a practical caveat that nobody quotes: “Less valid than #3”.
My favorite macro technicals:
- The 200-day Moving Average: “The trend is your friend”. Not yet.
- The S&P 500 Large Cap Index – 13/34–Week EMA Trend (cyclical turns): not yet.
(CMG Wealth)
CONCLUSION
I don’t seek to forecast equity markets, only to identify periods where the risk/reward equation is favorable, i.e. valuation upside potential > valuation downside risk to optimize my exposure.
The stable Rule of 20 objectively measures where we are on the valuation risk/reward scale taking inflation into account:
“20” is fair value where upside potential to 24 (+20%) = downside risk to 16 (-20%). Simple, efficient “buy low-sell high” strategy. “If the values don’t make sense, don’t participate”.
The rest is earnings behavior… although, in truth, 12-month trailing earnings declined during only 4 of the last 11 bears and, except for the GFC, they were never the biggest source of losses. Declining valuations were, by far.
This analysis begun with an assessment of earnings risk. I am ending it concluding that they currently don’t matter much since equity valuations are still too high at 23.7 on the Rule of 20 scale.
John Authers recently displayed this Société Générale chart suggesting that equity markets are back to something like “fair value.” I took the liberty to insert my own rendition of what could be a “sustainable long-term trend” (red dash line).
For 25 years, we never had to worry about inflation, always within 1.0% and 3.0% (1.6% and 2.3% since 2011).
Yet, since 1995, we have had 4 bear markets averaging -41%. Excluding the GFC, which destroyed financials’ profits, EPS declined 17% in 2000 (total drawdown 49%), were flat in the 1-month 2020 COVID bear (-34%) although they subsequently declined 15%, and rose 19% in 2022 (-25%).
Valuations matter more than profits which, after all, are generally rising over time. Earnings recessions last 6-12 months and average -10%. Missing a beat is only painful for a few quarters. Valuation corrections can hurt for years.
The crucial call this year is not whether we have a recession or not, it’s what happens to inflation. To repeat:
- With inflation at 6.0%, the Rule of 20 says that the “fair” P/E is 14.0 (20 – 6). On trailing EPS ($222.41): 3115, down 20.0%.
- With inflation at 4.0%, the Rule of 20 says that the “fair” P/E is 16.0. On forward EPS ($224.78): 3600, down 7.7%.
- Assuming a recession, with inflation at 3.0%, the Rule of 20 says that the “fair” P/E is 17.0. On recession EPS ($200.00) = 3400, down 12.8%.
A recession, which would presumably take inflation down, would be less damaging to equities than growth with high, sticky inflation. The Fed is our friend, as long as it wins!
But since “You’ll never know all the answers”, particularly in this complex environment, keep managing your valuation risk.
And use the right metric, one that takes inflation into account. In past inflation years, the conventional P/E brought investors back much too early (high) with multiples that eventually proved deceptively too high.
The more stable Rule of 20 P/E is much more dependable. It has been relatively safe to buy below 20. As mentioned above, that would be 3115 at current EPS and inflation levels (6%) and 3600 with inflation at 4% and no recession.
- 2023 probably starts weak on earnings.
- Technicals are still negative. Currently, “the trend is not you friend, don’t fight the tape.”
But selective stocks/sectors can increasingly be found cheap as this bear is getting older.