I am often asked how Valuation-Informed Indexers should go about changing their stock portfolio allocation in response to changes in valuations. Set forth below are descriptions of five strategies that are supported by the academic research.
1. The Cliff Approach
If simplicity is your most important consideration, you might want to follow a Cliff Approach. John Walter Russell did research showing that those who go with a high stock allocation when the P/E10 value is below 20 nearly always make out very well and that those who go with a high stock allocation when the P/E10 value is above 20 nearly always live to regret it. If you are looking for a rule of thumb as to how to go about setting your stock allocation in a valuation-informed way, this is it.
The Cliff Approach is my least favorite approach. It is artificial to suggest that stocks offer a great value proposition when the P/E10 level is 19 and a poor value proposition when the P/E10 value is 21. There?s nothing magic about the number ?20.? The Cliff Approach has worked historically. But there is going to be a day when it no longer works. This is a case where an excessive interest in simplicity can get you into trouble.
When Wade Pfau did his research showing that Valuation-Informed Indexing has provided far higher returns than Buy-and-Hold at greatly reduced risk for the entire 140 years for which we have data, he focused on a portfolio allocation strategy in which the investor goes with a 90 percent stock allocation when prices are insanely low, a 60 percent stock allocation when prices are moderate and a 30 percent stock allocation when prices are insanely high. So you know this one works.
It?s not necessarily the only strategy that works, however. We have only had Shiller?s research available to us for 30 years. And we only have 140 years of historical data on which to run tests. So I don?t think it is possible today to identify the one best strategy. You certainly should be thinking of doing something along the lines of 90/60/30. But I don?t think you should feel that the case for this particular strategy is so strong that it would be a mistake to follow some other approach that possesses more personal appeal.
The first reference in the literature to the Valuation-Informed Indexing concept that I am aware of was made by Warren Buffett?s mentor Benjamin Graham in his 1930s book Security Analysis. Graham suggested that it would be a good idea for investors to go with 75 percent stocks at times of low valuations, 50 percent stocks at times of moderate valuations and 25 percent stocks at times of high valuations.
These numbers are of course lower than the numbers used in the 90/60/30 strategy examined by Pfau. Why did I suggest that Wade check out an allocation strategy calling for higher stock allocations than those suggested by the Master? Perhaps I have been overly influenced by the recent bull market into feeling more enthusiasm for stocks than is warranted by a review of the entire historical record. Or perhaps Graham was overly influenced by the anti-stock feeling that was in the air in the years immediately following The 1929 crash and the Depression that it brought on. Your call.
4. The Gradualist Approach
From a purely theoretical perspective, the best allocation strategy is one in which you lower your stock allocation by a small amount (perhaps 5 percentage points) in response to each one point increase in the P/E10 value. The reality is that risk is increasing a small amount with each increase in valuations. There?s no real danger when valuations rise from the fair-value P/E10 of 15 to the slightly above fair-value P/E10 of 16. So it seems silly to make an allocation change when that happens. But the theoretically pure thing to do is to make a tiny allocation change in those circumstances.
I steered away from advocacy of The Gradualist Approach in the early years of my development of the Valuation-Informed Indexing Model. I was influenced by the intense opposition of the Buy-and-Holders to allocation shifts into trying to avoid frequent changes. I?ve come to believe that that was a mistake.
I certainly don?t think that you should make more than one allocation change in a year. But there?s no big problem with making an annual allocation shift. Buy-and-Holders plant the idea that annual shifts are a bad idea by pointing to the fact that there are costs associated with shifts. But the Buy-and-Holders themselves believe in annual rebalancing and rebalancing brings on costs too. So I don?t think the transactions costs incurred should be the deciding factor. You want to get your allocation percentage right. If there are small costs associated with doing that, those costs are worth paying.
The beauty of The Gradualist Approach is that you never have to debate whether to make a significant allocation change or not. Your allocation changes are less dramatic than those made by investors following other strategies. This means that you feel less emotion about the change. That?s good!
When valuations have been slowly moving up over the years and your stock allocation has slowly been dropping from 90 percent to 40 percent, you will have no trouble making that one more change you need to make to get the allocation down to 30 percent. That healthy emotional reality is very much in contrast to the emotional reality experienced by the investor following The Cliff Approach when he has to go from perhaps an 80 percent stock allocation to a 20 percent stock allocation in one moment of time.
Making big jumps causes you to stress out and stressing out causes you to make poor decisions. So over time I have come to feel drawn more and more to the idea of making a number of small allocation shifts (but still no more than one per year) rather than making a few big ones spaced far apart in time.