The three main government-approved credit rating agencies — Standard & Poor’s, Moody’s, and Fitch — notoriously failed in recent years in rating complex structured financial assets.
My vague impression is that the ratings agencies first became corruptible in the 1970s when two things happened:
They switched their basic business model in the early 1970s from being paid by users of their information (bond-buyers and the like) to being paid by issuers (debtors). That incentive structure created an obvious conflict of interest.
I spent years in the market research business telling consumer packaged goods manufacturers what their market share was, a business that’s fairly comparable. We would have loved to have gotten into the more lucrative financial rating business (the Wall Street mark-up is a lot higher than the Corporate America mark-up), but up at least through a 2006 reform, you couldn’t get into that market unless you were already in that market. We didn’t have particularly large conflicts of interest in the market research business, since the primary users paid us for the data. Procter & Gamble is more interested in what Crest’s market share is than anybody else is, so P&G pays its market research supplier and takes steps to make sure it’s getting accurate information.
The surprising thing is that the ratings agencies didn’t get corrupted for several more decades after these 1970s changes. But, that long period of good behavior created an assumption on the part of the markets that just because they hadn’t allowed themselves to be corrupted by their incentive structure so far, they never would.