In previous articles, we examined investing in the S&P 500 when P/Es (Cyclically Adjusted Price to Earnings Ratios or CAPE) are at extreme levels. We discovered that investing when P/Es are at (or near) extreme lows tends to generate positive returns, and conversely when P/E are at (or near) extreme highs. The average P/E since 1900 has been about 16.7, so we define extreme P/Es as below 10 or above 24.
The table below summarizes average annualized returns related to extreme P/Es. It includes returns for the preceding 20 years as well as the 20 years following the extreme P/E, namely when P/Es were below 10 or above 24. An adjustment is made for the 1980 to 2016 period, given that average P/Es were somewhat higher during this period. We believe the latter period is of greater relevance to today's investors, given changes to the structure of finance since the 1980s. We will have more to say about this in a forthcoming article.
In both cases, the data correspond with our expectations. When P/Es are below the minimum threshold (e.g. Aug 1982), returns over the preceding twenty year were weak (3.7%), while over the following twenty years, they are relatively robust (10.9%). And when P/Es are above a maximum threshold (e.g., Dec 1999), the opposite is the case (13.8% and 2.9%, respectively). This relationship holds true over both periods.
Our earlier articles built a case for using information from extreme P/Es when making an investment decision. We know that P/Es are not a particularly useful tool when used for general market timing purposes (as Robert Shiller, author of the CAPE has acknowledged), but when they approach extremes, they offer useful information.
Currently, the Cyclically Adjusted Price Earnings Ratio is equal to 29.1, well above the threshold for extremely high P/E ratios. Based on the analysis above, we would expect the S&P 500 to generate relatively modest returns over the next twenty years. We regress twenty year returns on the CAPE from 1980 to 2017. The R2 is a robust 0.62. This forecasts an annualized return of about 1.8% for investors with a 20-year horizon.
This return number is higher than the -3.5% expected return for the 1900 to 2016 period (see here). But in any case, there is little reason for individual investors or pension plans to hold out much hope for the so-called normal return of 7% to 8% per year.
Next, we explore how sharp declines in the S&P 500 might change expected returns. Before we do this, we raise expected annualized returns to a range from 3% to 5% over the next twenty years. We do not think these are unreasonable estimates, though at the high end they may be a bit optimistic.
We then explore how expected returns change, if the S&P 500 declines by 10%, 20%, 30%, 40% or 50%. In the table below, for example, if the S&P 500 declines by 20% we would expect annualized returns to increase from 5.0% to 7.4%, etc. Note that the greater the fall in the S&P, the higher the expected returns.
Assume that the price of the S&P 500 falls by 20%. Over 20 years, what is the difference between the current 5.0% return and the 7.4% return that would then be generated? The difference would amount to about 60% in return over a twenty year period.
So how do we go about making the decision? Do we invest today or wait for prices to decline? If we choose to invest today, we may experience a potentially significant drawdown that reduces capital and impairs our investment. For example, twice since 2000 (in 2000-2002 and 2007-2009) investments have declined by 40%. And that is not the end of the story, for once the S&P falls, investors tend to panic and liquidate at a loss (given our all-too-human tendency to buy high and sell low).
Alternatively, a decision to hold cash may result in our remaining on the sidelines as positive returns are generated. This may seem unlikely, given that current P/Es are above 95% of all P/Es since 1900. But it remains a real possibility.
So what is to be done? We can choose between three options, given the limitations of this two asset class world. We can (1) invest in the S&P today and hold our breath, (2) hold cash until valuations decline, or a blend of (1) and (2). With current P/Es at 29.1, if prices do not fall ("this time truly is different") then the first choice is the correct way to go. However, if prices decline sharply, then patience makes good sense. A blended (hedged) decision is also reasonable, given that we do not know the path in advance -- none of us is omniscient.
Unfortunately, there is no crystal ball -- the path is never entirely clear. We need to make judgments based on information in front of us. The decision to hold cash is not easy and for this reason most investors tend to be fully invested. That decision is understandable, especially given uncertainty about the path. But this decision should be made consciously! Implicitly betting that markets will never revert, in our opinion, is akin to arguing the trees grow to the sky or that Wile E. Coyote can defy gravity indefinitely. Again, is this time different? Yes or no? Markets are cyclical, in our view, and mean reversion is a fact of life in markets, as Jeremy Grantham and others have documented.
We see no reason why this time should be different, especially given extreme P/Es and low expected returns, rising short-term interest rates, the end of QE, highly indebted private sector (with very adverse distribution of debt that could well endanger growth in aggregate demand), an economic cycle that is long in the tooth, etc.
Some may respond that such a decision is equivalent to "market timing." Our view is that there is nothing wrong with market timing when it is based on a reasoned, conscious approach to the data. And this data suggests that the S&P 500 is overvalued. In addition, there is no reason why we need to be fully invested. If we expect markets to decline by 30% or so over the next five years, with no idea as to precisely when, we will need to generate 43% in returns just to break even. Nothing in life is certain, but utilizing tools such as extreme P/Es, to evaluate when to invest and when not to, seems to us a very reasonable decision. And we need to shake free of the notion that we must always be fully invested! Cash can be used as a strategic resource.
Another way to invest is to consider an adaptive regime-based approach to multi-asset class investment that incorporates an allocation to cash. We test this approach here.
Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.
I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it. I have no business relationship with any company whose stock is mentioned in this article.