Lately the stock market has seemed like the Muppet version of Upstairs, Downstairs--in which all the action consisted of Kermit running up some stairs, then down, then up gain. You could say this is just sound and fury, signifying nothing. But then, maybe the market is sending a message.

The best argument for continuing high stock prices goes like this: Many stocks will pay much higher dividends in the future than they do today. This is obvious for companies like Microsoft, which pay no dividends now but will someday either pay big dividends or, what amounts to the same thing, buy back their own stock at hefty prices. It is less obvious for slow-growth industries like utilities, but it is a good bet that the total cash flow to corporate stockholders-- dividends plus stock repurchases--will rise more or less in line with the growth of the economy as a whole, and that dollar GDP will rise at something like 5% per year (3% real, plus 2% inflation) for many years to come.

What does that say stocks are worth? According to a simple rule, if interest rates are 6%, a promise to pay $6 per year forever is worth $100, but a promise to pay $6 this year, $6.30 next year, $6.6015 the year after (the payout increases each year at the same rate as GDP, 5%) should be worth $600--100 times the current payment. And so if a stock really were a promise--which is to say, if there were no risk--stocks arguably should be worth something like 100 times the current cash flow to stockholders, which would mean a Dow of something like 17,000. But a stock, of course, isn't a promise; it's a hope. And so, historically, investors have demanded a large "risk premium," valuing stocks at a much lower multiple.

A number of financial analysts have noted that this historical risk premium turns out to have been excessive: In reality stocks have not been all that risky for investors with a long view. Maybe, therefore, the great surge in stock prices simply reflects a new rationality on the part of investors. Not to worry, right?

Well, let's think about this for a minute. The kind of calculation that justifies a Dow that is 100 times current dividends places a lot of weight on events in the very distant future. It says, for example, that half the value of stocks today reflects dividends and stock repurchases that will take place more than 70 years from now. That sounds silly, and it is--for two reasons.

First, while U.S. corporations probably will be paying zillions of dollars to stockholders in the year 2070, most of that cash flow will come from companies that don't yet exist and therefore have no relevance to current stock prices. Suppose that an investor 70 years ago (hmmm, that's 1929) had known about the huge profits Microsoft would someday make; that wouldn't have been a reason to buy any of the stocks that then constituted the Dow . And today's mammals are tomorrow's dinosaurs: Someday Microsoft will go the way of carriage manufacturers, and you can't invest today in the companies that will eventually eat its lunch.

Second, how reliable is that 100-times-cash-flow formula? The answer is that it's very sensitive to the assumptions. If the economy, and stock dividends, grow at 5.5% instead of 5%, the formula should be 200 times cash flow ( Dow 34,000 or so); if growth is only 4.5 and the interest rate is 6.5, it's just 50 ( Dow 8500). And since nobody knows the right numbers, the "correct" valuation of stocks is very uncertain, which means that if stocks were actually valued based on that kind of thinking, they would be extremely risky investments.

But weren't we supposed to know from history that stocks aren't all that risky? Yes, but that is because historically people have been very cautious about buying stocks, which meant both that the average return on stocks was high and that small changes in expectations about growth or interest rates didn't move valuations all that much. If people now start buying stocks based on the belief that they are not risky at all, that belief will turn into a self- defeating prophecy: The new valuations will be very unstable, swinging widely with the slightest bit of news, as the current Muppet performance suggests.

In short, people who claim that stocks are still hugely undervalued, that the Dow ought to be umpteen thousand, are trying to have it both ways. They want us to exult in the "creative destruction" of technological progress but forget about the destruction bit--which says that you can't invest today in the companies of tomorrow because they don't yet exist. They want us to believe that everything about the stock market is about to change-- except the riskiness, which will stay low no matter how speculative investors become.

None of this necessarily means that stocks are currently overvalued or even that events might not eventually justify a Dow of 36,000. But basing your investment decisions on that hope would be-- dare I say it?--a very risky bet.