Sunday, July 31, 2005
Earlier on the Summer Stroll, I said that the DJIA's dividend has had over an 90% correlation to its stock price during the past 45 years. I stumbled on this bit of data in the early 1990s, and it is the primary reason I began to study dividends. If you look at other price correlations such as earnings, GDP, book value, etc., you also find correlations that are high. Earnings have about the same correlation as dividends, GDP is in the range of 80%, and book value in the 70% range. As I explained in Summer Stroll #3 my decision to use dividends as my primary indicator of value was simple: Dividends have far less volatility than earnings (about a third as volatile). That means dividends are easier to predict, which ultimately allows for the computation of a narrower "valuation window" for the DJIA. A few years ago I began studying the correlation of combinations of financial data on DJIA price. I found that adding bond yields to dividends could explain the price movement of the DJIA at over the 95% correlation level and also lowered volatility. Computing a multiple correlation on today's market using only the DJIA's current dividend and the yield from a long-term corporate bond, results in an expected price level of 11,540. Combining dividends, earnings, and interest rates results in a expected DJIA price of 11,800. Any way you slice it, the stock market, as measured by the fundamentals of the Dow Jones and interest rates, is cheap, very cheap. In my judgment, the only thing that is holding the market back is the threat of higher oil prices. High oil prices have had remarkably little effect on inflation or the economy, but they are seen as a wild card by big money. That is the reason that stocks have had such a tough time getting through the 2000 highs. My sense, however, is there is a day coming soon when the market will reconnect with its historical relationships, and stocks will spike higher toward the range of values my multiple correlations are indicating.