The past few weeks in our article links, we’ve included some essays and blog posts discussing the merits and flaws of Shiller’s CAPE P/E metric and market valuation in general. After having read these articles and several others that have been wrapped into the debate, we felt it might be useful to synthesize some of these thoughts and lay out some “random thoughts” of our own as to how CAPE and other valuation tools should or could fit into a broader investment worldview.
Before we begin, let’s recap the recent debate. Nobel Prize winning economist Robert Shiller has maintained a public database of US market earnings and price information going back to the late 19th century to understand valuation and earnings trends and their relationship to market moves over time. To smooth out the effects of earnings volatility through business cycles, he prefers using a P/E ratio based upon average earnings over a trailing 10-year period. CAPE stands for “cyclically adjusted PE” ratio. Recently, some have argued that, in effect, Shiller’s earnings data since the early 2000s has become less relevant compared to past data because of the change in the way US companies have to account for “goodwill.” This is a new contribution to the continuing debate over accounting inconsistencies in his database. The counter arguments state, yes, there are inconsistencies, but this doesn’t mean the indicator is meaningless. Even if one goes back and scrubs data to make it more consistent or substitutes net earnings with operating earnings, for instance, markets are still overvalued.
We find merit in both trains of thought. The Philosophical Economics blog post we included two weeks ago provides a very interesting and persuasive case as to why the accounting inconsistencies matter, especially since treatment of goodwill changed in 2001. Investor/writer/blogger Mebane Faber provided a very solid, interesting counter argument to the CAPE skeptics, which we included in last week’s articles.
Overall, this general debate has popped up because of the surprising strength of markets in the US over the past few years. Anytime markets move 30% in one year and approximately 150% off a low point, you’re inviting a good ole’ teeth-gnashing “tastes great/less filling” type argument. This comes around in some form during every market cycle. Depending on which valuation metric employed, one can make the argument that the market is overvalued or fairly valued. Typically, the fairly to under valued crowd has pointed to P/E ratios based on trailing 12-month earnings or forward earnings guidance. The overvalued crowd cites the Shiller CAPE P/E and other longer-term metrics.
We can all argue mechanics and methodology for eons and eons. The nuances of using valuation metrics get lost in the shuffle in our humble opinion. What are our (somewhat) quick thoughts on CAPE and valuation metrics in general?
- Setting the stage, there aren’t any “perfect” valuation metrics out there. Accounting rules change. The structure of economies changes over time. This doesn’t mean simple valuation metrics aren’t useful however, especially when they reach extremes.
- As Shiller and many others have pointed out over time, valuation metrics such as the CAPE P/E are terrible short-term timing tools and shouldn’t be used for such purposes. Just because the Shiller P/E is at 25x now doesn’t mean the multiple can’t go to 35x in a heartbeat. Similarly, just because the P/E is 14x doesn’t mean it can’t go to 5x quickly. Again, this doesn’t mean the metrics aren’t useful. Last year in one of our posts, we pointed out the late 1990s example where both short-termers and long-termers could have been right in the months and years following Greenspan’s “irrational exuberance” speech in 1996. It’s useful to point this out again. When Greenspan made the speech, CAPE P/E was trading at approximately 26x, not much higher than the Shiller multiple today. Short-term market cheerleaders were vindicated when markets skyrocketed over the last three years of the late 1990s creating a massive equity bubble. Long-term CAPE proponents were vindicated as the 10-year returns from 1996 onward were sub average just as the CAPE would have predicted, owing to the massive market collapse from 2000 to 2002. Both were right. It was all a matter of perspective. Over the course of that 10-year time period there were severe pleasure and pain episodes for both sides. Bottom line: using the current CAPE P/E in to make bold predictions about the coming month, quarter, or year is a fruitless, ridiculous exercise.
- …Which brings us to general perspective. Just because US markets are moderately overvalued currently according to a Shiller P/E (across a range of earnings types) doesn’t mean that markets are going to hell in a hand basket and that we need to crawl in bunkers with canned food somewhere. Instead of using valuation frameworks to think in stark black/white, bull/bear, elation/depression, invest everything/run for the hills terms, we need to bring more nuance to the table. There are better ways to conceptualize valuation. For instance, many people invest in equity markets with some longer-term goal in mind, whether that relates to retirement goals, education goals, or something else important to the investor’s life. It’s essential to approach these goals as realistically as possible. At the current levels above 20x, it’s not necessarily important to think of the level of future returns and obsess over them, or to obsess over what markets are going to do next week or next month or next year, but to take into consideration generally that there’s a solid probability that returns are going to be average, if not below average, at best over the coming decade. While the current valuation level in the US doesn’t mean the world is coming to an end, it means it’s probably not an effective strategy to count on markets going up 15% to 20% a year over the next 10 years to make up for a lack of saving or planning. If markets go up 20% per year, great. If not, having the proper perspective going in means that the investor is prepared to meet goals without stress. From an institutional portfolio management perspective, elevated valuations don’t force us to immediately sell everything and run away. Overvaluation does alert us, however, to be on the lookout for potential volatility and turmoil. Certainly, if markets are trading at extremes, like the 40x witnessed in early 2000, or the 8x observed in the 1980s, investors should take note and act in a more aggressive fashion from the sell or buy sides. In the end, during times like these with moderate over or undervaluation values, attacking and/or obsessing over the messenger, i.e. the valuation model, is a distraction and waste of time. Using them to help frame expectations appropriately and plan realistically makes sense.
- …Which leads us to some thoughts on the use of specific market predictions based on valuation. We see these often and include forecasts for future returns in our monthly chart book. They are usually based on statistical analyses evaluating the relationship between fundamental metrics and future performance. Statistically speaking, the Shiller P/E and other metrics have a solid record in terms of forecasting future annualized returns. Though forecasts are helpful as an additional framework to guide the investing mindset, oftentimes, forecasts and predictions are misconstrued, misrepresented, or misinterpreted, though. The investing public takes numbers and predictions like these and runs with them as if they’re straight-line gospel, forgetting that the path to those returns over a decade can prove a wild ride. Let’s say a valuation model with a decent statistical record predicts that annualized returns over the next 5 years will be 5%. Sounds benign enough. At the end of 5 years, $100 dollars would be worth $127.63. Slow and steady, right? If markets went up 100% in year 1, down 50% in year 2, up 75% in year 3, down 31% in year 4, and up 5.7% in year 5, you get to the same performance number. The model would have been correct in this case, but that would have been one heckuva ride for an investor. Meanwhile, pundits from both sides in the valuation debate would have found ways in those numbers to support their respective causes. Bringing this back to the current period, US markets are somewhat overvalued currently based on these longer-term metrics. If the market goes up another 30% next year, does it mean the long-term valuation forecasts are invalid or crazy? Absolutely not. It’d probably mean the comeuppance would be more dramatic somewhere down the line. If markets go up 30% next year, will the bulls sling rocks and arrows saying the metrics are useless? Of course they will. That’s the nature of the game. Bottom line: we humans tend to condemn or congratulate too much based on very short-term data. Though very difficult, investors need to look past the clutter of short-term figures
- Next, let’s not ignore the notion of relative value. In our rush to debate whether or not the US S&P 500 is trading at 20x or 23x or 25x, lets not forget that there are indices and stocks here and abroad trading at 5x or 8x or 10x. When it comes to these types of debates, we tend to be myopic. Whether the US is fairly valued, moderately overvalued, or very overvalued depending on who you talk to, there are numerous indices and individual stocks around the world that are significantly undervalued. Just as it’s “always 5 o’clock somewhere,” there are almost always undervalued investment opportunities here or abroad for those willing to use a disciplined approach and pay attention (and willing to ignore headlines). Bottom line: metrics like the Shiller P/E aren’t perfect, but they’ve been pretty decent indicators over time when thinking in terms of relative valuation. Getting hung up in the “deep in the weeds” methodology discussions or obsessing over whether or not the US is overvalued distracts us from understanding and taking advantage of all sorts of opportunities here and abroad that show up as undervalued across a range of metrics.
- As a final thought, it’s always eyebrow-raising when these types of debates over valuation pop up; it’s a little qualitative nugget to put in the back pocket to remind us to stay aware. In our experience, people find all sorts of ways to attack traditional valuation indicators like the Shiller P/E when markets are moving higher and higher. Generally, we hear that the indicators are outdated and that new methodologies are more capable of guiding us through “new economies” or “new paradigms.” In the late 1990s, the Shiller P/E and others were maligned. Analysts shifted to other metrics like price per clicks or eyeballs. As the housing bubble reached it’s peak, we heard all sorts of reasons why metrics such as price to rent or price to disposable income were no longer relevant. For the record, we don’t think the markets are in any sort of bubbly situation like the tech or housing bubbles, at least not yet. It does make the hair on the back of the neck stand up, however, when we all of a sudden see numerous articles appear on our screens dismissing longer-term valuation metrics outright and touting notoriously underwhelming metrics like forward price to earnings ratios. Again, there are flaws embedded in many traditional fundamental indicators, but their long-term predictive ability has been demonstrated many times.
