“Stocks are undervalued.”

“Stocks are overvalued.”

These are two of the most common sayings on financial TV, trumpeted by bullish and bearish pundits on a daily basis. What value do they provide to the average investor? At best, none at all, and at worst they may lead some to take deleterious action in their portfolio.

Why? For one thing, a definition of valuation is rarely provided, nor is any objective analysis of what that valuation indicator actually means for the stock market. But more importantly, time frame is totally ignored as is the question of whether these amorphous terms are indeed helpful in managing a portfolio.

Instead, these phrases are merely used to support an existing bias. If the pundit is bullish they say stocks are undervalued (or cheap) and if bearish they say stocks are overvalued (or expensive). Black and white with no gray area, as consumers of financial information are said to prefer.

But in life and markets, there is often a gray area. Valuation metrics are far from precise and there is a range of outcomes and paths that the markets can take.

Last March, on the 6th year anniversary of the Bull Market in U.S. stocks, I wrote of elevated valuations in the S&P 500:

“If trees don’t grow to the sky, then, future returns will have to suffer because past returns have been so strong.”

What was I referring to? The elevated CAPE ratio (or “Shiller P/E”), which was just above the 90th percentile historically. Traditionally, such valuations have led to below-average forward returns as can be seen in the chart below (far right side shows average 6-year returns from valuations above the 90th percentile).

Interesting, but this chart says absolutely nothing about a) the range of possible outcomes or b) path of returns.

Let’s take some time now to dig into both of these important concepts.

A Range of Outcomes

First, you’ll note in the table below that valuation tends to be a pretty good predictor, on average, of forward returns. In general, the lower the starting valuation, the higher the forward returns and vice versa.

*Note: All Data presented is Total Return (Includes Dividends)

But if you look more closely at the table, what you’ll find is that the average returns, even in the highest valuation decile (90-100%), are still positive. So while overvaluation indicates a higher probability of a below average return, it does not suggest that one should expect a negative return. Therefore, shorting the market when it is deemed to be “overvalued” does not have a positive expectancy. The best you can say when valuation is extreme is that future returns are more likely to be below average.

We can see that as well in the next table illustrating the percentage of positive forward returns. While your odds of a positive 7-year return are higher when starting from a low valuation, they are still at 84% when starting from the worst valuation decile. Only in the 2-5 year range starting from the worst valuation decile are the odds of a positive return below 50%.

Next, let’s take a look at the worst and best forward returns from various valuation deciles.

What’s interesting here is that a low valuation does not guarantee a positive return, especially in the short-term (1-3 years). Cheap can always get cheaper, as it did in 1981 before stocks bottomed a year later in 1982. What you are achieving when buying the market when it is extremely cheap is merely a higher probability of above average future returns.

On a similar note, a high starting valuation is far from guarantee of poor returns, especially in the short-run. In September 1996, stocks were “overvalued.” Over the next 3 years, the S&P 500 increased over 95%, a 25% annualized return. This was an extreme example as it preceded the frothy stage of the tech bubble but it is instructive nonetheless. Just because the market is “overvalued” does not mean it “has to” do anything. It simply means it is more likely to have a below average forward return.

Path of Returns

Valuation may give you an idea of the probability of an above/below average forward return, but it tells you nothing about the path of those returns.

The CAPE Ratio ended May 2016 at roughly 26.3, the 90th percentile historically. Let’s take a look back in history at a few other times the S&P 500 ended the month with a similar valuation.

1) September 1996. Stocks were “overvalued” in September 1996. From there, they would go on to become much more overvalued, rising for another three and a half years to their peak in March 2000. After that, a nasty bear market ensued, taking the S&P 500 down over 50% from its peak.

Summary: Stocks would end up getting to a below average 6-year return as predicted by its starting valuation, but the path (bubble, then crash) was far from predictable.

2) August 2007. Stocks were “overvalued” in August 2007. The S&P 500 would peak two months later and go on to suffer its worst decline (57%) since the Great Depression. From there, a new bull market began which eventually took the S&P 500 back above their October 2007 levels by 2013.

Summary: Stocks would end up getting to a below average 6-year return as predicted by its starting valuation, but the path (crash, then ferocious rally) was far from predictable.

3) October 2014. Stocks were “overvalued” at the end October 2014. From there, they would go higher until peaking in May 2015 only to suffer two large corrections and vertical rallies thereafter. Overall, the S&P 500 is marginally higher today than where it stood at the end of October 2014.

Summary: Stocks have thus far produced below average returns as predicted by elevated starting valuations but the path (sideways with bouts of extreme volatility) was far from predictable. Today, we have the same valuation as back in October 2014 and the future path is again unknowable.

Is Valuation Useless?

Does the fact that valuation does not tell you about a) timing, b) the exact magnitude of future returns, or c) the path of returns render it useless?

For short-term traders, the answer is yes. In the short-run the market is a voting machine and the price investors are willing to pay for a given level of earnings (the multiple) is all that matters. Overvalued can become more overvalued and undervalued can become more undervalued.

But in the long-run, the market is a weighing machine, where starting valuations tend to matter as mean reversion kicks in. This is especially true when valuations are at extremes (above the 95th percentile) as they were in March 2000.

Over the next 10 years, the S&P 500 would produce a total return of -6% with extreme volatility in between. Efficient Market Hypothesis (EMH) proponents would say, “well, what can you do?,” but back in March 2000 US large cap stocks were not the only asset class; there were many alternatives. One example: investors in long-term bonds more than doubled money over the next 10 years with significantly less risk. Tilting towards such alternatives, while certainly not a guarantee, did make some sense at the time if you didn’t fully subscribe to the EMH.

For long-term investors and asset allocators, value can be important at the very least as a reminder to rebalance your portfolio. Rebalancing by definition is taking money away from the asset class that has done relatively better (more likely to be overvalued) and putting it into something that has done relatively worse (more likely to be undervalued).

More aggressive investors can go a step further by a) tilting a portfolio away from the asset class that is at an extreme overvaluation or avoiding it altogether, and b) tilting a portfolio toward an asset class that is at an extreme undervaluation. Such action won’t guarantee anything and may in fact be harmful in the short-run, though, as we know that anything can happen over a 1-3 year period. But over the longer-term, the prudent, disciplined investor may be able to add some return or reduce some risk by respecting extreme valuations.

I say “may” here because, once again, there is no certainty in investing. There are only probabilities. I realize that’s not as entertaining as a definitive “stocks are overvalued” or “stocks are undervalued” statement accompanied with a short-term price target. But if you’ve read this far your idea of financial entertainment is probably different than most.