From FORBES.COM 8/30/2004:

In My Opinion

Time To Buy Stocks
Phil DeMuth

In Yes, You Can Time the Market, Ben Stein and I argue that buying stocks when their fundamentals say they're cheap pays off handsomely for long-term investors. Most market timers think short term, trying to read tea leaves to forecast near-term movements, and earn famously dismal returns as a result (that is, when they bother to keep records at all). In the long run,though, Ben Stein and I found that valuing the market before diving in added substantially to returns. Is right now a good timeto buy, according to our measures?
Here's a quick recap. We looked at seven classic stock market fundamentals (using the S&P 500 as our benchmark--not attempting to select individual stocks). These were: price, dividend yield, price/earnings ratio, stock earnings vs. bond yields, and the price-to-sales, price-to-cash flow and price-to-book ratios. We investigated what happened to investors who bought when the stock market was priced below it's 15-year moving average on each of these criteria, and
compared their plight to those who bought ignoring these factors (by dollar cost averaging into the market or making lump sum investments). We found that, although this "timing" made no difference in the short run, ten or 20 years later the valuation-sensitive investor fared considerably better. This seemed in stark contrast to the drumbeat we heard on Wall Street, which always seemed to be telling us that it was a great time to buy stocks.
The experiments in the book were designed to show the relevance of valuation to long-term returns, not to promote a particular buying method, per se. However, readers were more interested in cashing in on the book's insights than appreciating its contribution to academic
finance literature. For such as these, the book immediately raised several important questions. What if the seven valuation measures were giving mixed signals? What if four said "buy" and three said "don't buy"--which ones were we to believe?
Keep in mind that the period we studied was the 20th century, an extremely bullish time for stocks overall. We were happy to buy if even just one of our signals gave us the green light. Market timing worked, not by clairvoyantly forecasting every peak and trough, but simply by letting us take a pass when the S&P 500 seemed overpriced across the board.
This led to a second criticism, which is that we had no "sell" discipline. We are guilty as charged. The century opened with the S&P 500 at 7, and ended with the S&P 500 at 1,148. In this context, virtually any intervening sale was a mistake, in the sense that--by simply waiting longer--an investor would have more money still. The message of the last 100 years is that you shouldn't sell the market unless you need the money for some more pressing reason. To put it another way, we failed to discover any in-and-out method that added utility beyond buying cheap and holding forever.
The biggest turnoff, however, was that many readers assumed these measures would have kept them out of the last bull market. There is some truth to this. No buy signals were flashing from 1988 to 1994, and again from 1996 to 2001, at least on the last day of each month--the only day we looked. (We didn't look at other days.) Having sworn off the market for most (if not all) of 1988 to 1994, a market timer would have partaken heavily during 1995, and made a killing as a result. Also, we never suggested not owning stocks during the bubble. Our first principle is diversification, and this trumps all considerations of market timing. If we own no stocks and have to enter the market when it appears to be overpriced, then so be it.
What about today?
As of the end of July 2004, fully three of the seven measures were below their long-term averages: p/e, price-to-book and price-to-cash flow. This suggests that now is not a bad time to jump into the S&P 500 for long-term investors. Following the same method outlined in the book, an investor who started where our book's data left off in 2002 would have 29% more money today than someone who simply bought into the S&P 500 every month.
Does this mean that we can party like it's 1999?
Not necessarily. Consider the market's trailing price-earnings multiple. It's now about 21. That's low compared to our book's 15-year moving average of 22 (a "buy") but high compared to its century-long average of 15. The century-long average contains almost all available historically valid data, which is nice, but our shorter moving average accounts for changes in the way investors see the worth of the price-earnings ratio itself. For most of the past century, investors priced the market too cheaply, leading to big gains for stock investors. Today, the market is at a richer multiple, which is perhaps why it didn't fall from a bubble p/e of over 30 all the way down to a p/e of 7, as it did in 1980. The unpleasant corollary of this is that future returns are likely to be more guarded than they were during the salad days of the last century. Unfortunately for baby boomer's retirement dreams that check has already been cashed. For long-term investors, our metrics are saying that while it may not be the world's greatest time to buy stocks, it's not a terrible time, either. In the last century, merely sidestepping the priciest times to buy (like during the Internet/telecom bubble) made investors the most money over the long term. If the past is any guide, the light is green once more.

***

Philip DeMuth, Ph.D., is president of Conservative Wealth Management. His Web site is www. phildemuth.com. Free updates on the metrics discussed in Yes, You Can Time the Market are posted on www.yesyoucantimethemarket.com.