The evidence that the stock market pays off is circumstantial, but strong. If it were a crime to make money, a jury would convict on evidence like this.

Amid all the advice on investing that sometimes works and sometimes doesn’t, one tenet remains our bedrock. Put your money in stocks for the long run and you will do OK.

From 1940 through the end of 2007, over 90% of all three or five-year stretches were winners. And there were no, zero, zip, zilch losing 10-year periods.

By rolling 10-year periods, I don’t mean just 1950 to 1959, but also 1951-1960, 1952-1961, 1953-1962 and so on. Since 1950, there have been 60 rolling 10-year periods. And no matter where you started in them, you would have made money.

S&P 500 1940 through 2007

Chart source: The Principal Group

But if you are beginning to doubt if that’s right, the results from this decade have given you plenty of reason. We started out with three back-to-back bear markets this decade: 2000, 2001, 2002. We had a strong bull in 2003. Then we got a lame period for the next two years. Finally, in 2006, the market got strong again. It stayed that way late into 2007 before it nearly fell to its open and went negative again. The calendar turned just in time. Then last year it was back down again.

At this point, nearing the end of January 2009, as Barack Obama steps up to the presidency, we are into another bad year. Two and a half good bullish years out of nine years and one month have left U.S. stocks in the red. So far this decade (since the close of 1999), the Dow has lost 30%, the S&P has shed 45% and the Nasdaq is still down 65%.

That hardly inspires confidence in the infallibility of the long run. We may be about to see the first negative 10-year period for stocks since the Great Depression.

Do you still buy stocks, and if so, what stocks?

I am buying stocks, yes, because at present valuations the market is so low that once the recession ends, the probability that stocks will rise quickly is as near to certain as you can get. The market is also likely to begin rising before the recession ends, by up to six months.

Even with this confidence, though, I am buying with the notion that I might or might not see the price rise in the next 12 months and may even see it drop before it recovers and rises. We could have a negative 10-year period, but I don’t foresee a negative 12- or 15-year period.

Interest rates on fixed income investments are now so low that many stocks pay better dividends than even junk bonds do.

What stocks do you buy for the recession? What stocks for the coming Obama years? What stocks are the best for 2009?

You’ve seen dozens of ads that purport to answer those questions. You have probably read advice like “buy utility stocks” or “master limited partnerships” or “get into medical equipment and research stocks.” Those may be fine answers as far as they go, though they aren’t the answer I would give.

I would say this: look at what a company expects to earn over the next two years. If that number is positive, buy the stock if you can get it at less than 10 times its estimated earnings for 2011.

Many companies already have 2010 earnings estimates. Not all of them have projections for 2011 yet. You can take the 2010 estimates and do your own projection for an additional year by estimating conservative growth, say 5%. Most companies will do much better than that if they are going positive at all.

This is not hard to do, either. Yahoo Finance, among many others, gives analyst estimates for stocks along with estimated 5-year growth rates. The five-year estimates are inclined to be a bit optimistic in many cases, so just use the 5% rule to be safe. Take the 2010 estimate times 1.05 to get your 2011 number. Multiply that by 10 (which would be the price of the stock at a P/E of 10), and there’s your target buying price.

Many companies are worth more than that, and you should consider them on their individual merits if you are confident in your projections. As an analyst, I will be using somewhat more tailored P/E ranges and growth rates, plus considering dividends case by case.

But if you do a quick scan and use 10 times 2011 earnings, you are so far inside the margin of safety that you can relax and feel good about doing the right thing. This is as close to a sure bet as you can get.

Source: Going Against the Evidence – Who Says Stocks Make Money?