Ratings could be kept. however their role in securities laws (mandated in many investment vehicles & plans) should be eliminated. Forcing firms to only purchase "certain" rated securities basically forces large market participants to purchase securities based on the opinions of others which are indirectly selected by the Govt. the last time I checked the prescience of govt was no better than the private sector.

Sell side research is bad enough, gov't sponsored/controlled sell side research is an abomination.

The ratings argument for debt securities could also be taken to the other extreme. Why not rate equities as well? Imagine the uproar if a govt. mandated entity could determine which equities you could or could not hold. This would be a soft form of industrial policy for big giant firms.

such govt. mandates are soft forms of nationalism in my books as they effectively allow certain firms cheaper access to capital at the expense of freedom of capital allocation. Fiscal enslavement in my books.

@ Jacek,
we had talked about this before, so you will not be surprised that I beg to disagree with some of your postulates:

-ratings are estimates of future events and thus --contrary to what some quants disastrously succeeded (for some time) to make us believe-- inherently judgmental and thus subjective.

-It is actually beneficial that there are different analytical approaches, as they are able to compete in the marketplace and the best approach will prevail.

-Ratings are indeed not verifiable --EX ANTE! They are, however, veriafiable --EX POST!. And that is done since 20 years. Which has shown quite clearly which agencies are better at predicting default than others. An investor who feels that a rating agency consistently issues to high (or too low) ratings is free to turn to another one --that's why plurality of methodology and competition is so important.

-I agree that more information should be disclosed --by ISSUERS! The access to non-public (often quite significant) information by the big 3 agencies has a) improved the predictive power of their ratings; but has b) raised a methodological entrance barrier to competing agencies (that would rate without mandate/payment by the issuer), thus petrifying the collusive triopoly. To deny any agency access to that information would make make ALL ratings less accurate.

-As said before, the plurality of analytical methodology might confuse some, but is an important market-discipline regulative.

- Contrary to what you suggest, ratings should and (at decent agencies) are CONSTANTLY under review; there's NO annual cycle; Yes, on-site management discussions with issuers are habitually annualy (that's what the corporate reporting cycle is), but with many more often, and there is a constant flow of information in between. WHENEVER a relevant information transpires, rating action should and will follow. Granted there are some fly-by-night agencies that review their ratings once a year (if at all), but there is reason for their obscurity.

-"Verifiability by publicly available mechanism" --That would require the presence of a truly objective algorithm --"The Truth" --something that many a hedge fund, investment bank believed in for the past 15 years´. Well, let me tell you, IT DOESN'T EXIST! A rating is..... see above. Where the big 3 agencies failed the most was where they relied (in deviation to previous practice) too heavily on quant methodology (SF, RMBS, CDOs, CDSs, joint-default-approach, etc.).

-"affordable", mechanised ratings are out there, whether by outside providers, or bank-internal (Basel II compliant). I have looked at many but found them wanting. NONE of them caught the events of the past 3 years (to be fair, few else did), but it proved that their very base, the Efficient Market Hypothesis, the Gaussian Copula and many other Quant tenets only work under certain conditions and only for short time horizons.

Don't get me wrong, as I had said before, quant modelling has its place as useful auxiliary tools; but ONLY in the context of a hybrid approach that combines them with sound expert-based, and, Yes, subjective, judgmental fundamental analysis. "Democracy" is too important an institution to be abused for a flawed concept, which I hope you'll forgive me for calling it.
Gunther

There is saying about the difference between theory and practice - "in theory there is no difference between thoery and practice, but in practice there is."

The idea of "democratizing" ratings is all well and good, in theory. As is the idea of having the purchaser of a security pay for the rating. But how well would these ideas work in practice? Especially if the security or equity in question was a new offer on which little or no information was available?

How many people are able to meaningfully evaluate the kinds of factors that should go into a bond's rating, even if they could get access to the appropriate information? Not many, I would bet.

There is another saying about crime investigation - "follow the money." if you want to know where the rating agencies loyalty lies, follow the money.

In the maritime world, there is a similar problem with what are known as the classification societies. These groups, such as the American Bureau of Shipping, Lloyd's Register, Bureau Veritas, Panama Bureau of Shipping, etc. are paid by ship owner's to issue certificates of compliance with the various regulatory schemes (e.g., Safety of Life at Sea certificates required by the International Convention on the Safety of Life at Sea and issued under the authority of flag state governments). There are good classification societies and there are bad ones. Some are so notorius that the US Coast Guard and similar port state safety administrtations actually target vessels known to be carrying certificates issued by the sleazier classification societies. I once inspected a vessel that was in such bad condition that no water actually made it to the nozzle of the BEST fire hose. It all leaked out before it got to the nozzle! But according to the inspection reports and certificates on the vessel, everything was in perfect condition. The ship also had no lifejackets and the life boat had no bottom - literally. It was truly a case of the ship having the best inspection reports money could buy.

If you want to have honesty in these kinds of third party verification systems, the third party verifier should be chosen by and paid by a second party (i.e., an intermediary such as a government entity) with the cost passed on to the first party (e.g., the ship owner, the bond offeror, etc.). If the intermediary finds out that the third party is being irresponsible, then the third party goes off the list of accepted service providers. The survival incentive then becomes one of satisfying the neutral intermediary rather than the first party that might benefit from covering up material facts.

Until the performance incentives for third party verifiers are aligned with society's interests rather than those whose creditworthiness, safety, etc. is being evaluated, we will continue to have situations such as we have seen - AAA ratings given to what should have been rated as junk bonds, clean safety certificates given to ships with no fire-fighting equipment, no lifejackets and no life boats, etc.