Valuation-Informed Indexing #228
by Rob Bennett
Valuation-Informed Indexing strategies are rooted in the 33 years of peer-reviewed research showing that today’s P/E10 value effectively predicts the return that stocks will provide over the upcoming 10-year time-period. It makes little sense to go with the same stock allocation when the most likely annualized 10-year return is 15 percent real (as it was in 1982) as you go with when the most likely annualized 10-year return is a negative 1 percent real (as it was in 2000).
The conventional view is that the price assigned by investors to the stock market as a whole tells us something important about the prospects of the underlying companies; positive economic developments push prices up and negative economic developments push prices down. For Valuation-Informed Indexers, it is investor emotion that has the primary influence on market prices (with economic developments having only a secondary effect through their ability to cause shifts in investor emotion). We employ regression analyses of the 140 years of historical data available to us to predict future returns and to guide our asset-allocation decisions.
Some are troubled by the assumption at the root of this analytical process. U.S. stocks have been providing a long-term average return of 6.5 percent real for 140 years now. We are presuming that this will continue to be the case. But that is of course not entirely the case. If the U.S. economy is stronger in the future than it has been in the past, the long-term return will be a bigger number and, if the U.S. economy is weaker in the future than it has been in the past, the long-term return will be a smaller number. The fair-value P/E10 number may not continue to be 15 but may become something higher or lower than that.
Is this a problem?
It’s a caveat that all following this strategy should appreciate. However, I don’t see it as being a big problem.
The U.S. economy is a stable economy and the 6.5 percent real return has applied for a long time. The odds are strong that we will not see much change in the fair-value P/E10 number. We certainly could see small changes in either an upward or downward direction. That justifies not treating research done in this area as pure science. But it does not detract from the merit of the analytical project in a terribly significant way.
People have an inclination to be far more skeptical of claims made under the Valuation-Informed Indexing Model than they are of claims made under the Buy-and-Hold Model. It’s true that the numbers that apply in the future may not be precisely the same as the numbers that we can generate by studying the past. That’s not the question that matters. We need to use some numbers to guide our asset-allocation choices. The alternative is to use the numbers used by Buy-and-Holders; that is, to treat the nominal gains or losses experienced in the market for a year as real and meaningful.
They are hardly that.
Is there anyone who truly believes that, when market prices increase by 30 percent in a year, the reason is that the value of the U.S. economy increased by one-third? Or that, when market prices decrease by 30 percent in a year, the reason is that the value of the U.S. economy decreased by one-third?
Valuation-Informed Indexers essentially assume that the market increases in value by 6.5 percent real every year, even in years when the nominal price increase is 30 percent positive or 30 percent negative. It’s entirely possible that we may have entered a new economic era in which the annual gains will be 7 percent real rather than 6.5 percent real and that the predictions that Valuation-Informed Indexers use to make allocation decisions will thus be a bit off the mark. But the Buy-and-Holders are using worse assumptions. They are assuming that 30 percent gains are real and that 30 percent losses are real. I am confident that, in a year in which market prices increased by a nominal 30 percent, the real increase in value was closer to 6.5 percent than it was to 30 percent. We don’t necessarily get things precisely right. But it is better to get things roughly right than precisely wrong.
A huge benefit of the new strategy is the discipline that it imposes on the investor. Investors following Buy-and-Hold strategies tend to be self-congratulatory when prices increase by 30 percent. They elected to follow a strategy that always calls for large stock allocations and they were rewarded (temporarily) by the market for doing so. The feedback they are receiving is telling them that they have this investing thing figured out! Investors in those circumstances don’t feel much of a need to look too closely into the question of whether those 30 percent gains are truly justified.
In contrast, Valuation-Informed Indexers start with a default belief that the market value should only increase by 6.5 percent real each year. They can of course go with some other assumption if they choose. There is no rule that says that a Valuation-Informed Indexer cannot decide for himself that the fair-value P/E10 number has risen to 17 percent or 18 percent or whatever. Still, the default belief in a 6.5 percent return grounds the process of assessing how much stock returns have increased in something real. When you are acting on a belief that stock returns cannot be known in advance, a 30 percent gain is not entirely shocking — it’s a big gain but the numbers say what the numbers say! A Valuation-Informed Indexer understands that the U.S. economy could grow sufficiently productive to support an annual gain of 7 percent or perhaps 8 percent but is highly suspicious of numbers showing a 30 percent gain. That’s a healthy skepticism!
It pays to be slow to accept upward price movements as indicative of a permanent increase in U.S. productivity. There are many who say that we should use a number higher than 15 as the fair-value P/E10 value today because we have had a longer string of insanely high P/E10 values than what we have seen at any earlier time in U.S. history. The reality is that every secular bear market in U.S. history has lasted until the P/E10 value dropped to 8 or lower. That’s 65 percent down from where we stand today. A drop of that size will cause a lowering of the average P/E10 value so that the 15 number will be viewed as being a bit on the high side rather than a bit on the low side.
We all are victims of current-day bias. We see what is happening around us as real and much longer-lasting historical realities as coincidences of bygone days. The productivity of an economy as large as the U.S. economy changes slowly and gradually. No, the fair-value P/E10 number will not necessarily always be 15. But it is unlikely that that number will change too dramatically or too rapidly.
Rob Bennett recorded a podcast titled Money Magazine Is Asked: “Just How Bad Would Things Have to Get Before You Changed Your Advice?”. His bio is here.