by Rob Bennett
A fellow on a discussion board who is highly skeptical of Valuation-Informed Indexing made what I had to acknowledge was a good point the other day. I am always going on about how valuations must be taken into account in investing analyses. This fellow (he posts as “GW”) noted that he has been taking some money out of his portfolio to finance a vacation. Vanguard has been willing to pay him the full stated value of his portfolio amount even though Rob Bennett says that we should be marking down stock prices to reflect the reality that today’s P/E10 value is in the low 20s and the fair-value P/E10 value is 15. Do I think that Vanguard should pay him something less than the stated value of his portfolio?
This one puts me on the spot. If I say “yes,” I sound like a crazy person. If I say “no,” I contradict my claim that valuations need to be taken into consideration when calculating safe withdrawal rates and when performing all others sorts of investment analyses. If valuations matter, they always matter. If valuations don’t matter when determining how much a mutual fund company should pay you for your shares, it cannot logically be argued that they matter in other contexts.
I say “yes,” Vanguard should discount the portfolio value for the extent that stocks are overvalued at the time shares of an index fund are sold. This will indeed make me sound like a crazy person in the ears of many. But you know what? This isn’t the first time this sort of thing has happened. And you know what else? I think it makes sense to discount the nominal amount of the portfolio to reflect the effect of overvaluation. I think that if we made that change, it would lead to all sorts of good things.
To say that stocks are overvalued is to say that the nominal values assigned to all of our portfolios are wrong. Shouldn’t Vanguard be paying the proper amount, not the misstated nominal amount? It sure seems so to me. To say that this is so is no more than simple common sense, which unfortunately we have gotten away from in much of our theorizing about how stock investing works.
There are two reasons why the idea is going to strike a lot of people as crazy. The first is that it has never been done that way. We humans are creatures of inertia. When something has been done one way for many years, we stop thinking about it and come to believe that the long-standing procedure must be right. I think we need to rethink this one. There is a reason why the idea of paying the full nominal price assigned to a portfolio has become a long-standing custom even though it defies common sense to do things this way and that reason no longer applies today.
The reason is that in the days when the custom became established there were no index funds. It is hard to ascertain the proper value of individual stocks; this is what the investors who buy the stocks are trying to do but there are of course all sorts of disagreements. So it may be that the best we can do with individual stocks is to use the nominal price (this is the price at which the last share of the particular stock sold in the last transaction that involved this particular stock). The appearance of index funds on the scene changed the environment in which we operate. It is easy to adjust the price of a broad index fund to reflect fair value through use of an effective valuation metric like P/E10. So we should do that.
Vanguard should not be paying GW the full amount of the nominal price for his index-fund shares. It should be marking those shares down for the extent of overvaluation that applies on the day they are presented.
Will it cause problems if Vanguard and other companies continue to make payments reflecting the full nominal value of shares of individual stocks? I don’t think so. If adjustments are made for index funds, the adjusted prices are the prices that will be reported in newspaper reports of the current value of the S&P and the Dow. The practice of reporting the lower prices will undercut the irrational exuberance that causes overvaluation. Thus, investors will begin assigning more accurate prices to individual stocks as well. Individual stocks will not be priced perfectly. But of course they are not priced perfectly now. This will bring us a step closer to accurate pricing, which is a step in the right direction from where we stand today.
The other argument that will be put forward for why this is a crazy idea is that P/E10 is not a perfect valuation metric. People will point out that, if we adjust portfolio values for the effect of overvaluation, we will be ascribing inaccurate prices to the indexes. That’s not good.
It’s not good. But the other side of the story is that the prices we use today are not accurate either. The prices we use today are less accurate than valuation-adjusted prices.
Consider where things stood in January 2000. The P/E10 value was 44, nearly three times the fair-value P/E10 level of 15. The Dow stood at somewhere near 12,000. Had the price of the Dow been reduced for the effect of overvaluation, it would have been reported as having been 4,000. Reporting the Dow at 4,000 would have gone a long way toward diminishing the irrational exuberance of the time. People would have known the true values of their portfolios. This would have left them far more able to determine effectively whether they could afford to buy houses, cars and vacations. We should all want people to gain the ability to plan their financial affairs effectively.
It still seems shocking, though, doesn’t it? We would be reporting the Dow to be 4,000 at a time when it was common practice to report it as being 12,000. Too strange.
The full reality is that it would not have played out that way in real life. In real life, we would have been using a valuation-adjusted figure in the years before the bull reached its top. If we had been reporting a valuation-adjusted figure all along, the Dow’s price would never have risen to anything in the neighborhood of 12,000 in the first place. If we report portfolio values accurately, investors will have the information needed to make rational investing decisions. Eugene Fama’s dream of an efficient market will be realized in those circumstances! It is our failure to report numbers accurately that has been causing investors to become so emotional.
There is no magic to the way portfolio prices are determined today. The selling price that applied on the last transaction involving a stock is reported as the new value of that stock. But that price is the price that applies only for the few shares that actually were sold at that price. Is there anyone who believes that, if all owners of that stock tried to sell their shares, they would all obtain that price? They obviously wouldn’t. If all owners of that stock tried to sell, the price would crash. So we know with certainty that the nominal price used today is not the right price.
This discussion (which I acknowledge is kind of “out there”) shows how revolutionary Shiller’s work really is. If valuations affect long-term returns, valuations must be taken into consideration in all aspects of investment analysis. Shiller changed the world of investing in fundamental ways, in ways that we are only 30 years after publication of his early findings beginning to grasp.
Rob Bennett developed a unique retirement planning calculator. His bio is here.