Elevated market valuations mean greater risks? Not necessarily …

The ’90s was a fantastic decade. In particular (and I’m at risk of dating myself here), the early ’90s especially so. The theatres gave us Misery, The Fugitive, Jurassic Park, and Dazed and Confused. If you liked a good legal thriller, you likely had one of John Grisham’s books on hand. And if you were a music lover, you may have had a CD collection that included Nirvana’s Nevermind, Pearl Jam’s Ten, or if grunge wasn’t your thing, perhaps Massive Attack’s Protection, or Eric Clapton’s Unplugged. There’s a lot we can take away from the early ’90s ... but more on that later.

Stop me if you’ve heard this one: “equity valuation is at extreme levels and the markets are due for a pullback.” Or so goes the mantra of the bears. Valuation can tell us a lot about equity markets, but the propensity for a correction over the near-term isn’t one.

First, my team and I will acknowledge that yes, equity market valuations are at extreme levels. Whether that be the S&P 500 Index, the S&P/TSX Composite Index, or many others, and it would be disingenuous to claim that valuations aren’t at elevated levels. Specifically, in regards to the S&P 500 Index, we can point to the price-to-earnings (P/E), price-to-book, price-to-cash flow, market cap to GDP, or any metric you prefer, including our Rule of 20 metric, and they’d each suggest the same: equities aren’t cheap. But valuation and returns aren’t as correlated as many would believe, at least not over the short term.

Chart 1: There’s no arguing that valuation is at multi-decade highs.

S&P 500 Index
Price-to-cash flow and price-to-earnings ratios
Last 30 years through April 30, 2021

Chart 1 clearly shows what many investors are fretting about. The trailing price-to-earnings ratio of the S&P 500 Index is at a level that we haven’t seen since the peak of the dot-com boom in the late 1990s. Valuation at the current level has occurred only a handful of times in the last 50 years. However, if we look at the one-year forward returns for the S&P 500 plotted against its trailing P/E ratio, it starts to become clear that valuation is an unreliable indicator of short-term performance. It’s only when we plot returns against valuation on a 10-year forward basis that we start to see a relationship emerge.

Chart 2: Valuation is a poor predictor of short-term returns.

S&P 500 Index
Forward 1-year and 10-year return vs Trailing P/E multiple
Last 50 years

Chart 2 suggests that valuation matters to the long-term outlook. Does this suggest that investors should exit equities in anticipation of weaker 10-year forward returns? That wouldn’t be our recommendation as we know annual returns tend to be lumpy. It’s more likely that one or two bad years bring down the average as opposed to expecting 10 individual years of underperformance. Further, negative calendar returns tend to be concentrated around recessions. So, in consideration of valuation, we need to ask ourselves what the probability of a recession might be in the coming year or two. We firmly believe that given the fiscal and monetary stimulus, excess savings, pent-up consumer demand, and improving economic fundamentals, the risk of recession through 2022 is very low.

We’re not saying investors should ignore valuation. Rather, we need to put valuation in context with the broader fundamentals of the equity market.

Let’s use one of our favourites, the previously mentioned Rule of 20. It’s a simple metric that has been around for decades. It puts equity valuation in context with inflation. There’s a historical and inverse relationship between the trailing 12-month P/E ratio of the S&P 500 Index and inflation. The rule suggests that the S&P 500 Index is at fair value when the sum of the 12-month P/E ratio and 12-month inflation (Consumer Price Index) equals 20 (the long-term average is actually 20.9).

As you can see, the market is rarely at fair value. Currently, our Rule of 20 is near the highest level in 50 years. That’s not overly comforting at a time when yields and inflation are likely to continue to move higher. Higher inflation equals lower valuation. At this level for the Rule of 20 (one standard deviation above the long-term average), the one-year forward return averages -1.5%. Should we take it for granted then that we’re in an environment that may signal a negative near-term return? Well, no. Again, we need to look at the broader environment that we find ourselves in before we pass judgement.

Chart 3: Rule of 20 valuation at current levels are rare.

CPI YOY & S&P 500 trailing P/E ratio
1970 – current

There aren’t many times in history when valuation reached the levels as we have it today. In the last 50 years, we suggest the comparisons are limited to 1987, 1991, and 2000. Each of these three periods are unique.

In 1987, the S&P 500 Index jumped 7 multiple points while gaining 39% in the first eight months leading up to the crash in October of that year — but as we point out, the market finished the year up 5.2%, including dividends, even with the crash.

Following the recession and bear market of 1990, the S&P 500 Index saw a sharp P/E expansion in 1991, followed by three years of solid earnings growth while delivering positive market returns. The trailing 12-month P/E multiple expanded by 12 points (from a low of 13 to a peak of 27 in 1992). The P/E multiple receded back to the historic average over the next three years while the Index produced annual returns of 7%, 10%, and 1% between 1992 and 1994.

The year 2000 stands out in investors’ minds as a comparison to today because of the tech wreck. And while there are similarities in valuation between today and then, what followed 2000 to take the market down was first a deep valuation correction due to the excessive and euphoric tech boom (some similarities to today’s tech sector), followed by a recession and the aftermath of the 9/11 tragedy. The key difference from 2000 is that the market had yet to face the economic and earnings recession of 2001, while today, and similarly in 1991, we’re past it.

Chart 4: This market may be most similar to the period between 1990–1994.

Contribution to return by earnings growth, P/E ratio, and dividends
1971 – 2020

We believe recessions bring about most bear markets. Unless you believe we’re headed towards an economic collapse in the next year, it’s unlikely that the current market is like 2000. Therefore, the best similarities as to outcomes are 1987 (characterized by a flash crash to normalized levels — possible but unpredictable) or 1991 (a normalization of valuation through earnings growth). While we can’t completely rule out a 1987-style valuation correction, we argue that today’s market is most like 1991.

That brings up the question of earnings growth and what it could be for 2021. The macro data that we look at as a team is suggesting that earnings should be 30–35% higher than 2020 or better — that would put us well above 2019. At that level, we’d need to see the P/E multiple fall to 24 times 2021 earnings from the current level to completely cancel out the earnings growth and result in a flat market between now and the end of the year. A contraction of 8 multiple points is rare outside of recessionary bear markets. Further, during periods when earnings growth is greater than 30% on a year-over-year basis, the average P/E contraction is only 4.1 multiple points, while the average and median 12-month returns for the S&P 500 Index are 10.2% and 12.4% respectively. If earnings surprise to the upside or if we see only a partial multiple contraction, then it would be reasonable to expect returns similar to history, with upside risk.

Chart 5: Multiple expansion and earnings growth tends to be inversely correlated. We expect valuation to moderate with the forthcoming earnings growth.

Year-over-year change in S&P 500 Index earnings per share vs Change in trailing P/E multiple
Last 50 years

In some respects, the articles suggesting valuation is at extreme levels are correct. However, coming to the conclusion of a market melt-down may be too pessimistic given the earnings results we’re already seeing for the first quarter and expect through the rest of the year. We believe the worst-case scenario is a couple of years of average to below-average returns while the P/E multiple normalizes, similar to what we experienced between 1992 and 1994. But as we know, valuation is a terrible predictor of short-term performance. Given everything that we look at from an earnings perspective, it’s hard to say this market is reminiscent of 2000 while a much easier comparison would be 1991, which is why we should not be fearful of current valuations.

Going back to the 1990s can remind us of great music and great movies while also shedding some perspective on equity valuations …

Right here, right now
There is no other place I want to be
Right here, right now
Watching the world wake up from history
– “Right Here, Right Now,” written by Michael Edwards

… and Jesus Jones too.