Giving credit where credit receives its due: An Interview with Marco Suter

RMA Journal, The,  March, 2002  by Nicholas Hayes,  Beverly Foster

• E-mail
• Print
• Link

MS: To answer the second part first, credit risk rating is at the sole discretion of the credit officer. In my opinion, the nucleus of our system is that we will never compromise the credit rating. As you know, credit rating is an extremely important element of risk control. It is used not only to classify our counterparty risk exposures for internal and external reporting, it also determines, for example, our credit approval authority and sets the hurdle rate for profitability. Obviously, we often are pressured by the originator to assign a better credit rating to allow more favorable pricing or longer maturities. If we started changing our ratings to allow that, we obviously would jeopardize the integrity of the system.

Many of our credit policies set out the types of risks we accept for a given rating category. They also determine the frequency of credit reviews. In our domestic lending business, pricing grids are based on internal ratings to ensure that our earnings profile is in line with the risks assumed. In our international lending business, we have set maximum concentration targets by ratings category. Most important, however, our ratings are a key and core element in establishing and quantifying our expected loss, economic capital, and stress loss exposures. They determine how much of the "scarce" risk resource every single transaction consumes and are taken into consideration for the purpose of performance measurement of our business groups.

Our ratings system has 12 grades for our unimpaired counterparties and two impaired classes. Each grade is assigned a fixed group of default probabilities. When designing the rating scale, we were careful to ensure that ratings classes can be differentiated meaningfully and that the increases in default probability from one grade to the next are reasonably steep.

Given the complexity of an organization such as UBS, we have more than one ratings system. All systems, however, are linked through a master scale. Ultimately, they are all translated into default probability and expected loss. For the most part, our credit ratings measure pure counterparty default risk. In these instances, collateral or loan structures or transfer risk are taken into account separately. Two exceptions are our real estate lending activities in Switzerland and our Regulation U and Regulation T type lending businesses, for which we use transaction ratings rather than counterparty ratings. In these instances, the collateral value is directly factored into the rating.

RMAJ: What do you see as some of the greatest lessons coming from 2001? And where do you see the greatest risks and the greatest rewards in 2002? In certain industries, geographies, business lines? In the face of the almost worldwide downturn, what are you doing differently now from, say, two years ago?

MS: This past year reinforced the importance of sticking to our fundamentals. For UBS, for example, we learned anew the importance of cash-flow-based lending as opposed to enterprise-value-based lending. The stock market obviously proved this right. The other key learning point is that there must always be an awareness that surprises can occur at any time. No model would have suspected the September 11 event. But when you have this kind of tail risk, there is only one way to protect against it: Do not put too many eggs in one basket.