Thursday, November 10, 2005

Stock Market Models--Which One to Believe?

I have been mystified all year that large cap stocks have not fared well. I know there have been headwinds: oil prices, terrorism, natural disasters, political intrigue, inflation, and interest rate hikes by the Fed. But, my work has indicated that the surprisingly good economic and earnings growth should have been able to overcome the headwinds. Corporate Earnings for the S&P 500 are likely to be near 14% above last year, almost twice what many analysts were forecasting at the beginning of the year. Dividends will likely rise 12%, well above projections; and GDP is likely to grow near 3.7%, again, well above the estimates at the beginning of the year. I cannot remember the last time such good economic and earnings news was so totally ignored. My most reliable valuation model, which looks at the historical relationships between price, dividends, and interest rates, currently points to a fair value for the Dow Jones of 11,800, over a thousand points higher than the Dow's current level. This model has been able to explain nearly 95% of the annual movements of the Dow over the last 45 years. A second tool, our dividend discount model, which computes the present value of future earnings and dividend growth, says the intrinsic value of the DJ30 is over 12,000. Both of these models are time tested and have seldom been wrong for lengthy periods of time. But they have been overly optimistic in 2005. In attempting to understand why large cap stocks have utterly ignored their excellent fundamentals, I fell back on one of my old models, the P/E model. I stumbled across this simple relationship between P/E and inflation in the early 1990s. It has not been as precise as the other two models, thus, I don't spend a lot of time with it. This model does not pay much attention to projected earnings, dividends, or interest rates. I have tested them in the model but they do not improve the model's correlation with actual price to earnings ratios. The formula uses only the Consumer Price Index and a 3% premium. Here's how it works. To find the appropriate P/E for today's market you add 3% to the current rate of inflation. The current CPI core rate of inflation is 2.3%, year over year. 3% +2.3% = 5.3% This produces an expected 5.3% earnings yield (E/P) for the today's market. To determine the predicted P/E ratio, we divide the expected earnings yield into 1. 1/5.3% = 18.7X P/E (Projected) The formula says the Dow Jones should be trading at 18.7 times trailing 12-month earnings. With last the last 12 months DJ earnings at $650, that produces a price of 12,155 for the Dow. That's great, but it does not help us understand why stocks are selling at only about 16X earnings. Then it hit me. I did all the original research on the P/E model using the actual average annual CPI, not the core CPI. The core CPI, which excludes food and energy, is the measure of inflation that is most used on Wall Street because it is less volatile;it is also the inflation indicator that the Federal Reserve watches most closely. As I was thinking about this, it occurred to me that the relentless rise in oil prices may well have tipped the scale in favor of investors using the actual CPI instead of the core CPI, because they may have come to believe that oil prices and inflation are only going higher. They may have also abandoned using the core CPI because the difference between it and the actual CPI is as wide as it has been in decades. Over the last 12 months, the actual CPI has averaged 3.4%. If we insert this figure into the data, we get the following result. 3%+3.4% = 6.4% 1/6.4% = 15.6X P/E (Projected) 15.6 X 650 = 10,140 Yikes, that is not very encouraging, yet the DJ 30 did touch 10,156 in mid-October just after the hurricane-induced spike in inflation. It was also at about that time that investors became more worried about the economy and earnings because of the possibility of rising interest rates. Thus, the terrible storm along the Gulf Coast have created a kind of perfect storm in the financial markets. The Katrina, et al, spiked oil prices and inflation and which dampened prospects for economic growth and profits. Even though I wrote here and elsewhere that oil supplies were fine and that the economy would weather the storm without great effect, apparently the stock market was not buying my argument(it seldom does in the short run). But, in recent weeks as economic and corporate growth data has shown only a modest impact from the storm, the market has shaken off much of its lethargy and begun to rise. So what do we believe, the simple, old fashion P/E model that uses actual CPI and says stocks should be having a tough year, or the P/E model that uses core CPI and says stocks should be 12%-20% higher? The world thinks the value of a stock is what it is selling for today. I do not believe that, and I can show you proof after proof that, during times of crisis, the stock market almost always goes the wrong direction at first, before recovering it senses and more accurately pricing earnings and dividend growth. With oil prices now at $57.50, well off peak their peak of $70+ per barrel, the CPI should moderate dramatically in the coming months. I predict that by the middle of 2006, the CPI will be below 3% on a average year over year basis. The reason is simple, the Fed's target for inflation is about 2%, and moderating oil prices will help them get there. For your information, I have provided below a chart showing the actual P/E vs. the level projected by my simple PE Model. I think you will agree that the model has done a surprisingly good job of capturing the trend of the actual PE. It will change again next week when new CPI data come out. I'll report here what it looks like then.