Valuation-Informed Indexing #312
by Rob Bennett
The S&P 500 experienced a loss of 37 percent in 2008. Was that loss real? Or was it a mirage? I believe that it was mostly a mirage.
The Electron Global Fund was up 2% for September, bringing its third-quarter return to -1.7% and its year-to-date return to 8.5%. Meanwhile, the MSCI World Utilities Index was down 7.2% for September, 1.7% for the third quarter and 3.3% year to date. The S&P 500 was down 4.8% for September, up 0.2% for the third Read More
The conventional thinking is that the loss was real. Stock price changes are caused by unforeseen economic developments, according to the Buy-and-Hold Model. Investors change their assessment of the value of stocks each time they are exposed to new information. The banking crisis frightened investors. Their processing of all of the many negative information bits sent their way during that 12 months of time caused their belief about the future earning power of U.S. stocks to sour enough to justify slicing away one-third of the previous value of U.S. stocks.
Is this a realistic way to think about these matters?
The economic crisis was bad news. I recall reading an article in the Wall Street Journal in which serious people speculated that there was at least a small chance that we might be headed into a Second Great Depression. That’s scary stuff. On first consideration, it doesn’t seem too outlandish to believe that developments of sufficient negative force to bring on a Second Great Depression could cause a drop in the value of U.S. companies of 37 percent.
But there are two problems with this analysis.
First, there is always some chance that a mortgage crisis or a banking crisis will develop over the course of the next 12 months. According to the theory behind the Buy-and-Hold Model, it is only unanticipated negative economic developments that cause price drops. To the extent that investors understood that problems might develop, those losses were priced in to the market at the beginning of 2008. Only negative developments worse than the anticipated negative developments can be used to explain the 37 percent one-year loss, according to the conventional thinking.
Second, we have seen dramatic gains in the nearly eight years of time that have passed since 2008 came to a close. The annualized real S&P return (with dividends reinvested) over that time has been nearly 13 percent. That’s double the 6.5 percent average long-term return. So, according to the conventional thinking, there have been a ton of unforeseen positive economic developments during that time.
It’s quite the coincidence, isn’t it?
We see huge unforeseen negative developments in one year, followed by huge unforeseen positive developments in the following eight years. The net effect is that for the entire time period from January 2008 through July 2016, we saw an annualized real return of 5.7 percent, a number not too terribly far off from the 6.5 percent average long-term return that we would expect to see apply if there had been no particularly negative or positive economic developments at all.
My thought is that investors overshot the mark in their 2008 reassessment of the value of U.S. stocks. There really were serious economic negatives that turned up in the news in 2008. But they weren’t quite such the big deal as investors made them out to be in their first reaction to them.
And investors have overshot the mark on the up side in the years since. We were so frightened in 2008 that even average economic news would have come as a big relief. Perhaps the news was just average and the reason why we pushed stock prices up to a degree greater than average is that we were compensating for having pushed them down to an excessive extent in the overreaction of 2008.
Does it matter? We end up in the same place regardless of whether we got there by assessing things properly both in 2008 and in the years since or by overreacting to the bad news of 2008 and then by compensating for the overreaction in the following eight years.
I believe that it matters a great deal indeed.
If the Buy-and-Hold Model is correct, we don’t need to take valuations into consideration when setting our stock allocations. Prices are always set properly, so there is no way to know what the future holds. The only way to avoid price crashes is to stay out of stocks and it is not possible for most investors to finance comfortable retirements without investing heavily in stocks. So the best bet is to identify a stock allocation that makes sense for someone with your risk profile and stick with it regardless of the “noise” being generated by the market at any given time.
If the alternative Valuation-Informed Indexing Model is correct, valuations are not “noise” but important indicators of the true long-term value of stocks at any given time. Stocks become mispriced because of investor emotion and the P/E10 value reveals to us the amount of investor emotion reflected in the price of stocks at a given point of time. The P/E10 value was dramatically improved in early 2009 compared to where it stood in early 2008. The P/E10 metric was telling all investors who would listen that stocks were a good buy, that investors had overreacted in their decision to lop off one-third of the price of the market in a single year.
The description above is a simplification of the realities (as I understand them). The P/E10 value never dropped below 13 in early 2009, which is only a smidgen below the fair-value level of 15. In all prior secular bear markets, the P/E10 value has dropped to 8 or lower. So valuation-informed investors would have been reluctant to embrace stocks enthusiastically even at that time. The Valuation-Informed Indexing Model posits that investor emotion is the primary driver of stock price changes and that a drop to fair-value prices is justified even in the absence of any negative economic news. I don’t mean to suggest that using P/E10 would have told you that stock prices were going to shoot up again starting in early 2009.
The point is that there are two ways to think about what causes stock price changes that enjoy support in the peer-reviewed research. Neither model can tell you everything you want to know about how stocks are going to perform in the future. But it is my belief that the Valuation-Informed Indexing Model does the far better job of making sense of the realities that appear before us. I just don’t buy it that the value of U.S. stocks truly dropped by 37 percent in the course of that single year of bad news.
Rob Bennett’s bio is here.