Another Good Year Due for the Dow

Editor's note: This is Part 1 of Jim Cramer's annual review of the prospects for the stocks that make up the Dow 30. Be sure to read Part 2, Part 3, Part 4 and Part 5.

The M, the magical M, the multiple that people will pay for earnings, remains the single-most difficult element to predict in the stock market. All the quantitative people and all the engineers and all the Nobel prize winners in the world can try to figure out earnings and where those earnings will figure in a world of both sane and insane valuations, but it's not worth squat if you get the M wrong.

That's what I conclude when I look at my 2004 predictions for the Dow Jones Industrial Average and see where I made my biggest mistakes and where I was most clairvoyant. Oh, for certain, on some I was so right, it's scary. I predicted that IBMIBM would earn $5 and the market would pay $100 for it. Nice. And that HoneywellHON would earn about $1.60 and the market would take that stock to $36. Sweet.

But who would have thought how much people would pay for the aerospace earnings cycle? I thought BoeingBA would earn $3.25 and the market would pay $48 for it. Although the year's not over yet, Boeing looks like it will come in at $2.50, yet the market likes it enough to pay $53 for the stock! Or how about CokeKO? It's gotten so out of favor that the market hates the $1.95 it earned and is going to pay only $41 for those earnings. However, when you consider that I thought Coke was going to earn $2.10 and it had a high multiple, you can understand that with those estimate cuts came the phenomenon of multiple contraction.

The whole process of prediction can be quite sobering. I was off by a third in my General MotorsGM earnings power calculation, a third too high, which led me to be wrong about GM's closing price by 25 points! Same with IntelINTC, which I predicted could earn $1.35 and have people pay $38 for it. Nope, how about $1.13 and $23. You get the estimate too high, you can see the air come out of the stock.