Valuation-Informed Indexing #203
by Rob Bennett
Valuation-Informed Indexers believe that investors must change their stock allocations in response to big valuation shifts to have any hope of keeping their risk profiles roughly stable. We use P/E10 (the price of a broad index over the average of the last 10 years of earnings) as our valuation metric. P/E10 is superior to P/E1, the better-known metric, because it compares today’s price to the earnings level that applies over an extended time-period rather than to the earnings level for a single year, which is too influenced by temporary economic ups or downs.
Would P/E12 work as well or better? How about P/E8? There is obviously no magic to using “10” as the denominator of the fraction. Shiller just picked up on the suggestion planted back in the 1930s by Benjamin Graham. P/E10 has been tested and it works. So it is gradually winning support. No doubt Graham chose it in part because “10” is a round number.
We are still in the early days of development of the Valuation-Informed Indexing model. So we certainly should encourage research into the question of whether P/E12 or P/E8 or perhaps P/E15 would do a better job. Getting locked in to P/E10 just because that is what sounded right to Benjamin Graham back in the early 1930s would obviously be a mistake.
That said, there has now been a good deal of research published that examines valuation-related questions using the P/E10 metric. We lose much of the benefit of that research if we switch to another metric. So, while we should not be wedded to P/E10, we should require another metric to outperform it in a significant way before giving thought to dethroning it as the king of of valuation metrics among valuation-informed investors.
The thought that drives people to ask this question is not whether P/E10 is the perfect valuation metric or not. The evidence is that P/E10 is a very powerful tool. But of course everyone who believes in it understands that there may be a day when we will discover something better. The driver here is that the Buy-and-Hold Model has been dominant for so long that many investors are skeptical whether any valuation metric can help them invest more effectively and thus wonder whether the 140 years of historical data showing the power of P/E10 might be pointing to a temporary phenomenon.
It would be great if we had more data. It would be wonderful if we had 1,400 years of data rather than 140 years of data. But I very much doubt whether something that has held true for 140 years isn’t at least somewhat true. P/E10 may not be a perfect metric. But it has been shown to be a very good one and we should accept that that much is so.
There’s another issue that is a bit more tricky. When people use P/E10 to set their stock allocations, they are assuming that the U.S. market will continue to deliver a long-term return of 6.5 percent real or something in that neighborhood. It’s not necessarily so. It could be that we are about to enter a new Golden Age in which our economy will be so productive that returns will be higher even for stock purchases made at historically high P/E10 levels. And it is possible that we are about to enter a new Dark Age when purchases made at historically low P/E10 levels will deliver poor results all the same.
It’s so. We don’t know how productive the economy will be in future days. So we don’t know whether using the P/E10 metric to decide on our stock allocations will work as well in the next stretch of time as it has for the 140 years of stock market history available to us.
The other side of the story, however, is that a less productive economy will hurt Buy-and-Holders as well. And Buy-and-Holders too will be underinvested in stocks in the event that we see levels of productivity that few anticipate. The reality that the U.S. economy may be more or less productive in the future than it has ever been in the past is not an argument for staying at the same stock allocation at all times so much as it is an argument for being flexible in one’s expectations of future returns.
Many people have an inclination to see Buy-and-Hold strategies as the default choice. My view is that it is the strategies that are supported by the academic research that should be the default choice. People should choose Valuation-Informed Indexing strategies because they are the strategies supported by the academic research.
But they should keep in mind that our economy may be more or less productive in the future than it has been in the past and adjust their stock allocations according to their personal beliefs as to where things are headed. Investors who believe that the economy is going to be more productive in the future than it has ever been in the past can adjust the allocation choices indicated by their use of the P/E10 metric accordingly upward . And investors who believe that the economy is going to be less productive in the future than it has ever been in the past can adjust the allocation choices indicated by their use of the P/E10 metric accordingly downward.
The proper comparison is not between using P/E10 to determine one’s stock allocation and using a perfect valuation metric to determine one’s stock allocation. The proper comparison is between sticking with the same allocation at all times and using a valuation metric to better inform one’s allocation choices. Only after an investor has concluded that his stock allocation should vary with changes in valuations does he need to decide on a valuation metric to use.
The best metric available to us today, based on the evidence available to us today, is P/E10. But we should always remain on the lookout for something even better.