From Ben Stein, who says that maybe you can time the market! From realmoney.com:
Maybe You Can Time the Market
Soon I shall be shuffling off this little column and leaving you to the tender mercies of other observers. But before I do, I want to share with you in a very general way the debunking of a myth about the validity of "buy and hold" and the supposed inability of investors to time the market.
One of the great selling tools of stockbroker's is that it is impossible to time the market -- that is, to buy at a specific time when you get better investment results than buying at any other time. The supposed rule is that we never know when to buy and when to sell, so we might as well just keep buying all the time.
If true, this is good news for brokers. It also happens to fly in the face of all common sense. Why would there be a good time to buy and sell real estate, commodities, collectible's and bonds, but not a good time to buy and sell stock? Surely stocks are cheap at some times and expensive at other times.
Reviewing History
Well, I have just seen the preliminary results of a study by my investment guru pal Phil DeMuth that I think pokes a major hole in that "you can't time the market" baloney.
Dr. DeMuth took all of the 10-year periods in the Dow Jones Industrial Average since the end of World War II. He assumed that an investor made a buy of $100 worth of the Dow from the beginning of that period until the end of 2001. The investor, as you might think, made an excellent return.
Then Dr. DeMuth made a different assumption. He computed the trailing price-to-earnings ratio of the Dow for each 10-year period before each month. If the P/E in that month were one standard deviation less than in the prior trailing period, he bought $200 worth of stock. In the other months, he bought nothing. (A standard deviation is a range within which two-thirds of all the reported results fall. Thus, buying at lower than one standard deviation below the trailing P/E would mean buying when the P/E was unusually low.) The results gave that investor more than twice the percentage gains that the dollar-cost-averaging guy received.
If the investor then cleverly sold $200 worth of stock in every month in which the P/E was greater than one standard deviation above the 10-year trailing P/E, his percentage return would more than double again.
It's Not Rocket Science
I don't want to claim too much. It may be that future results will vary. And I do not want to claim too much credit. I did suggest to Dr. DeMuth that he do this research, although he chose the exact parameters of each regression. But the basic idea -- that stocks are sometimes more of a bargain than at other times -- is not rocket science, except to people whose livelihood depends on constantly selling stocks.
John Bogle, late of Vanguard, made similar calculations 10 years ago. He found that when stocks were paying a larger dividend than usual as a percentage of price, their return over the next decade tended to be above average. When the return of stocks as measured by dividends was below average, returns over the next decade tended to be below average. (Great differences in yields yielded great differences in total return over the succeeding decade.) And of course, Warren Buffett says he makes similar calculations, although I wonder.
Today, with earnings yields and dividend yields at uniquely low levels, a conclusion is suggested: Stocks in general are still too darn high to offer a good return over the next many years.
Luckily, there are bonds, real estate and cash.
Come see me at Systematic Investing group #board-966 lets talk formula plans.
