US securities analysts have been encouraged not to rate every company they cover as a Buy. That may not be working even as earnings fall.

According to the FT, "Equity research departments around the world have become much more bearish since the start of the year, but US analysts remain markedly more bullish on stocks than peers elsewhere." Research quoted by the paper shows that only 6.7% of stocks covered by U.S. stock researchers rate a Sell.

The news is disturbing because positive ratings are one of the things that keep investors in stocks and analysts who are slow to cut their price targets and modify opinions are likely to cost shareholder money.

U.S. stock researchers have long believed that lowering ratings gives them less access to management. That is a poor excuse for being overly bullish on shares.

The investing public and press mock analysts who put Sell ratings on stocks after they have dropped 90%. It appears that those actions are built into the U.S. equity research system.

Douglas A. McIntyre is an editor at 24/7 Wall St.

Fortune's thinks that Wall Street analysts are congenitally wired to be optimistic about earnings. (Fortune shares a parent, Time Warner Inc. (NYSE: TWX) with BloggingStocks). He thinks analysts are optimistic because they think the companies they cover are special. I disagree. In my view, analysts are optimistic about 2008 because they have to help sell stocks and banking business to keep their jobs. And selling stocks and closing banking business depends on forecasting higher earnings.

I have worked with Colvin on a few occasions -- for example, he interviewed me on Wall $treet Week with Fortune in 2004 -- and I find him smart and disciplined. But in my view, he misses the boat in his explanation of why analysts, such as those at Merrill Lynch & Co. (NYSE: MER). expect S&P 500 profit growth of 14% for 2008. The fact is that nobody knows the future so analysts are free to predict a future which happens to help them remain employed.

In my view, the place to look is how analysts get paid. What Colvin does not examine is that analysts are part of an investment bank's overhead. They don't charge clients for their research reports. Instead, brokers and investment bankers give the reports to clients to get them to transact business with the bank. If the reports help close a deal, part of the resulting fees are used to pay the analysts' share of overhead.