ANNANDALE, Va. (MarketWatch) -- It is one of the ironies of stock-market timing that it is easier to forecast where the market will be in several years than where it will be in several days.
And, according to a valuation model from a research firm with an excellent long-term record, the stock market is likely to be significantly higher in several years' time -- regardless of whether the final low of the last year's bear market has been seen.
The firm in question is Ford Equity Research of San Diego. This firm is on my radar screen because it publishes a newsletter entitled Ford Equity Research Investment Review. And its market-timing model deserves to be on your radar screen because it has a good long-term record.
Ford bases its model on an analysis of individual stocks' valuations. For each of several thousand issues, Ford calculates a so-called price-to-value ratio, which "is computed by dividing the price of a company's stock by the value derived from a proprietary intrinsic value model."
According to Ford, "A [price-to-value ratio] greater than 1.00 indicates that a company is overpriced while a [ratio] less than 1.00 implies that a stock is trading below the level justified by its earnings, quality rating, dividends, projected growth rate, and prevailing interest rates." After calculating a price-to-value ratio for each of the several thousand stocks that it monitors, Ford calculates an overall average.
Since 1970, when Ford started calculating the ratio, this average's record high level was 1.81, which it hit at the end of September 1987, three weeks prior to the worst crash in U.S. stock market history. Its second highest level, 1.79, was registered at the end of February 2000, just a couple of weeks prior to the bursting of the Internet bubble and the beginning of the 2000-2002 bear market.
Today, this average stands at 0.68. That's the lowest since December 1974, when it got slightly lower, to 0.61. Other than that December 1974 reading, the current level of this average is the lowest since 1970, when Ford began calculating it.
In addition, the current level is well below the four-decade average, which stands at 1.17.
This would certainly appear to be good news for the stock market.
To test whether Ford's model has statistical significance, I fed the monthly values for the average price-to-value ratio into my PC's statistical package, along with data on how the stock market performed. As expected, I found an inverse correlation between where the Ford average stood and the stock market's performance over the subsequent one-, three-, and five-year periods.
That is, higher price-to-value ratios were more often than not followed by lower market returns, and vice versa. And these correlations were significant at the 95% confidence level that statisticians often use to judge whether a correlation is genuine.
Can the Ford data be used to forecast anything shorter term?
Ford's analysts pointed out in a recent special report prepared for their clients that the average price-to-value ratio of U.S. stocks "dropped below 1.0 [only] seven previous times - April 1973, April 1976, April 1980, September 1990, August 1993, August 1998, and June 2002. In each of these cases, the market made an initial low within the first three months" after breaking below that 1.0 level.
In the current bear market, the 1.0 level wasn't broken until October. As a result, Ford's analysts conclude that, "We should expect to see that the market has bottomed by year-end."
No comments:
Post a Comment