Dealers should pay for credit ratings
News last week that a minor tweak in S&P’s corporate loan rating methodology could win them back market share in the lucrative new issue CLO market reignited the debate about the business model of issuers paying to be rated. Even if S&P’s internal controls and the Chinese wall which they insist exists between commercial and analytical considerations is robust, market participants clearly do not believe it. It is time to overhaul the way the credit ratings industry works.
Regulatory reform of the rating agency market, more or less complete across Europe and the US, has not removed the incumbents (or even weakened them much), and has not changed the fundamental business model — the issuer pays model.
A clutch of rating agencies have sprung up, all promising something a little different, but the inertia of investor mandates, as well as regulatory incentives, have kept the big two — Moody's and Standard & Poor's — on top, followed closely by Fitch and then regional or sector specialists such as Kroll, Morningstar or DBRS.
The business model has changed even less. Reforms such as the Securities and Exchange Commission’s Rule 17g5 (which required issuers to disclose data to rating agencies they did not hire) have been a damp squib. The appetite of agencies to provide ratings for free has been unsurprisingly limited, although there have been occasional public cat fights around particularly controversial structures (Goldman’s FIGSCO, for example).
But it’s hard to see how this situation is tenable in the long run, or healthy for the capital markets. Being forced to incorporate ratings into the new issue process whether or not the market has confidence in the ratings helps nobody.
The only real defence to the issuer pays model, which seems to embed an inherent conflict of interest, is that one where the investors pay would be worse.
Some investors have an incentive to puff up ratings, while others might have incentives to denigrate them. Investors that did not pay would be free riders on those that did — or else risk tripping over market abuse regulations. Issuer pays has the advantage that at least there is only one issuer, with a fairly transparent incentive in one direction only. There may well be a conflict of interest, but at least it is a well understood conflict of interest.
But there is a third option — intermediaries should pay. If ratings confer a benefit, it is to the arrangers of new issues and the traders of bonds. A greater breadth of investors and depth of liquidity in the market (happy side effects of a public rating, one would hope) make life easier for the investment banks arranging and trading the issue. They are public goods for each side of the process, too, but mainly for the intermediary.
So why not have a rating agency owned and funded by the banks? When co-operation is necessary, investment banks can do it. Markit is one example; ICMA Match, the precursor to IssueNet, is another, and plenty of the clearing houses are partly or wholly owned by their members. Italian banks own part of the Bank of Italy, big US banks are Federal Reserve member banks (so, technically, own the Fed).
If public ratings can be seen as a public good, ratings ought to be delivered like a utility. But, rightly, there has been fierce resistance to such a utility being owned by the public sector. China’s Dangong, the most prominent example, has not been a roaring success in western markets. Rating Russia one notch above the US has its downsides.
But something owned by market players would have credibility. Arranging banks will naturally want to have top ratings for their deals, but can keep one another honest, and will be carefully attuned to how credible such a rating would be with investors. The skills to arbitrage rating rules are abundant in the new issue structuring departments of investment banks. It is time they became gamekeepers as well as poachers.