(Once upon a time, I had written a column for Economic Times, on one of the anniversaries of the Lehman Crisis)
John Moody started using three alphabets to convey credit quality in a bond. That has today become the gold standard for investors to put money across the globe, without having to study who is the originator or even know of him. The alphabets have become underlying symbols for quality and pricing. Credit rating has become a universal tool for widening the credit markets, bringing pools of money across the globe to become invested in to unknown credit, on the basis of the alphabets. A few alphabets became the bond of trust for investors.
The rating agencies generally avoided conflict of interest, though every act of theirs, right from charging a fee for a rating to indulging in advisory services, raised eyebrows across the globe. However, so long as things went ok, no one blamed them.
Rating agencies had a big role to play in creating and expanding the market for securitised debt. It started with housing loans, then went on to cover every kind of retail debt from auto loans to credit card receivables. Global liquidity expansion created a hunger for investible assets beyond the usual stuff of sovereign bonds and corporate paper.
Credit rating depends on numbers, but the final judgement is truly a subjective one. Rating is not merely about ratios and stock prices. Other models have been tried which were purely quantitative, but have not worked. Therefore it is important that the person or persons who give the final opinion, are of sufficient maturity, wisdom and experience to finally assign the final choice of alphabets that will signify the credit quality and in turn affect pricing, investments etc.,
Rating agencies did an honest job till it involved vanilla debt where the repayment of debt is dependent on a single issuer. When it comes to securitisation, there is a major difference. Securitisation works on the premise that in a pool of large number of borrowers, there is a high degree of probability that a certain portion of the aggregate loans will get paid with a high degree of certainty. Fair enough. Rating agencies made allowances for defaults, based on past statistics and adjusted it for further uncertainties and expressed their final opinion.
In the world of securitisation, the only rating which helps to sell a paper is a “AAA” (Triple ‘A’), the highest possible rating. To the investor, once this tag is attached, nothing else is delved in to. And since all investors are institutional types, there is no explanation etc., The symbols suffice.
To my mind, the biggest flaw has been in the securitisation rating. Whilst a corporate paper rating is based on future expectations , securitised paper is assigned a rating based on past experience. This worked fine so long as the papers were restricted to home/auto loans in a conservative world.
With passage of time, the rating agencies started to lose focus. Their job was to comment on credit quality. Instead, they took it upon themselves to expand the market. In this zeal, the focus was lost. To put it simply, when personal loans were being securitised, in the early days, the originator had reasonable lending norms and the underlying credit was not so bad. With passage of time, increased competition led to dilution in origination standards. Unfortunately, the rating agencies still used historic models to validate structures. This, in short is what happened and the blame for the sub prime melt down lies fair and square with the rating agencies.
Normally, a rating is theoretically said to be ‘valid till changed’. One does not expect a Triple A to become junk during its life time. In bond ratings, there has hardly been any incidence of a Triple A defaulting on maturity. However, in the sub-prime meltdown, there was hardly any paper that survived to maturity.
In this case, the underlying assets were real estate. Rating agencies assumed that prices would not crash and that there would be a vibrant market going on for real estate. The other problem was that origination standards had fallen abysmally. So much so, that these loans were referred to as ‘liar loans’. Unfortunately, the rating agencies forgot one basic thing. In the traditional securitisation loans, there is a certain statistical probability of a minimum percentage of total pool fulfilling their contractual obligations. However, you cannot take a whole bunch of junk loans and expect similar behaviour, irrespective of what cushion you put in. In this basket, the default rate tended towards 100%! Once this happens, the rating agencies start to downgrade. I have never witnessed triple A getting junked in one go. Unless there is a structured fraud (like Enron), no one expects a Triple A paper to go to junk in the first downgrade. Once the rating agencies did this, the market for sub prime just vanished. And the cookie crumbled.
To my mind, the failure was on following counts;
- Securitisation ratings are only statistical probabilities. We cannot equate these papers with single issuer papers;
- Rating agencies role is to be subjective and not objective. This calls for people with huge experience in various industries to be core employees of rating agencies. In India, when the whole thing started, we had people with considerable experience across industries. Today, that is not happening. This ultimately impacts the ability of the rating agency to understand debt in its entirety;
- Rating agencies are straying from their core business, in to areas like advisory services and equity ratings, which also results in a classic ‘conflict of interest’ situation. This must be the first thing to be set right. Rating agencies should focus on ratings. At best, they can be sellers of aggregate information. They should refrain from taking a peek at equities;
- Rating symbols for vanilla debt should not be used for securitised debt or any form of structured papers. The regulators should ensure that securitised paper are treated differently from vanilla debt. SEBI, for the mutual fund industry, has put clear limits on structured paper that a fund can own. This has perhaps saved the mutual fund industry.
- In the early days of the rating agency, we used to have a Rating Committee which had independent people with experience across different industries and with huge management bandwidth. Following the US model, this is no longer the practice, with focus on instant ratings and on-line response, thanks to competitive forces. It would perhaps be a good idea to reinstate this practice. As regards time, the world will not come to an end if a rating takes three to four weeks to get assigned.
- There is finally, the vexed issue of rating agencies being a listed company. This is something that needs to be re-examined. Listing causes dilution of standards as the focus shifts to capital market valuations.