Market valuation is always a hot topic at the start of every year. Predictions are another. And there’s always someone who combines the two, citing current valuation for what the market will do over the next year.
Let’s stick to what we know. The current U.S. valuation hovers around a CAPE of 28. You can read more about the CAPE ratio here. For the short version, 28 falls on the expensive side of things, meaning on average expensive markets tend to perform poorly over the next decade.
But expensive does not mean that a market correction or crash is imminent or guaranteed. The funny thing about averages is that it turns a large basket of numbers – added up and divided by the total number of numbers – into one number. Some people choose to focus only on the one number and ignore all the rest but we can’t.
First, let’s look at the averages:
S&P 500 Returns following CAPE over 25 | ||||
1 Year | 3 Years | 5 Years | 10 Years | |
Average | 5.83% | 0.17% | 0.27% | 4.07% |
The average speaks for itself. A high CAPE on average leads to lower returns. Adjusting your expectations over the next 3, 5, and 10 years is certainly warranted, but…
Since 1926, we’ve only seen 12 other instances where the S&P 500 started the year with a CAPE of 25 or more.
S&P 500 Returns for Years with CAPE over 25 | |||||
Year | CAPE | 1 Year | 3 Years | 5 Years | 10 Years |
1929 | 27.06 | -8.42% | -26.96% | -11.24% | -0.89% |
1997 | 28.33 | 33.36% | 27.56% | 10.70% | 8.42% |
1998 | 32.86 | 28.58% | 12.26% | -0.59% | 5.91% |
1999 | 40.57 | 21.04% | -1.03% | -0.57% | -1.38% |
2000 | 43.77 | -9.10% | -14.55% | -2.30% | -0.95% |
2001 | 36.98 | -11.89% | -4.05% | 0.54% | 1.41% |
2002 | 30.28 | -22.10% | 3.59% | 6.19% | 2.92% |
2004 | 27.66 | 10.88% | 10.44% | -2.19% | 7.40% |
2005 | 26.59 | 4.91% | 8.62% | 0.41% | 7.67% |
2006 | 26.47 | 15.79% | -8.36% | 2.29% | 7.30% |
2007 | 27.21 | 5.49% | -5.63% | -0.25% | 6.94% |
2015 | 26.61 | 1.38% | ? | ? | ? |
2017 | 28.20 | ? | ? | ? | ? |
The numbers get messy. Here are a few thoughts:
- Notice how the following 1-year returns are all over the place. Almost an equal number of years had digit gains as had losses (or near losses). The market doesn’t immediately recognize expensive.
- The 3-year returns are only slightly better. 3 of the 11 saw double-digit annual returns, while 6 of the 11 showed losses (not counting 2015’s lack of data).
- Things get less messy looking 5 and 10 years out, but outliers still exist. I’d lean on the side of caution. Adjusting expectations over the next several years is the smart, safe thing to do. Does it warrant an allocation change? That depends on your strategy.
- 1929, 1997, 1998, and 1999 were bubbles and had an impact on subsequent years. Of course, bubbles are seen in hindsight. Few people knew it at the time.
- That huge gap between 1929 and 1997. Why? 1929 seems to the be the outlier here, as the rest come in the last 20 years. I’m sure there are several reasons. One that stands out in my mind is technology. A change in the makeup of the S&P 500, notably tech stocks, impacted the valuation of the entire index. A changing tech sector introduced a different set of fundamentals from the typical businesses before it, especially on the software side (higher profit margins, little to no debt, greater economies of scale, faster growth, etc). Investors are willing to pay more because of it. Other sectors certainly played a role during those years (I’m looking at you financials and energy, and technology influenced all of them) but tech had the biggest transformation since the mid-90s.
A lot of people latched on to CAPE after Robert Shiller made a couple notable predictions. Ever since then, there’s been a steady flow of people pointing out CAPE’s imperfections. A financial model isn’t perfect? Yeah, that’s never happened before.
The better way to use market valuation models is to use more than one. The best resource I know of that does that is StarCapital’s market valuation ratios by countries and regions – CAPE, Price-to-Cash Flow, Price-to-Book, Price-to-Sales, and Dividend Yield – and has the research to back it up.
Still, all models have limits and users should understand what those are before diving in. This brings me back to something Howard Marks said in his latest memo, that coincides with the data above:
I realized recently that in my early decades in the investment business, change came so slowly that people tended to think of the environment as a fixed context in which cycles played out regularly and dependably. But starting about twenty years ago – keyed primarily by the acceleration in technological innovation – things began to change so rapidly that the fixed-backdrop view may no longer be applicable.
Now forces like technological developments, disruption, demographic change, political instability and media trends give rise to an ever-changing environment, as well as to cycles that no longer necessarily resemble those of the past.
That comes off as a this time is different pronouncement, but I see it this way. These forces Marks refers to, may just make many of the old models less relevant.
Last Call
- The Illusions Driving Up US Asset Prices – R. Shiller
- The Resilience of US Equities! – Musings on Markets
- Investors Need to Look Well Beyond the Inauguration – B. Carlson
- The Art of Holding – MicroCapClub
- How Long is Now? – S. Godin
- Embracing Bad Ideas To Get To Good Ideas – HBR
- How Curiosity can Protect the Mind from Bias – BBC
- Why Silicon Valley Needs More Physicists – Wired
- The Great Unbundling – Stratechery