The Future of Risk-Adjusted Credit Pricing in Financial Institutions

Article excerpt

The current tool of choice for many institutions opting for risk-adjusted credit pricing, RAROC, doesn't reconcile the prices of loans with those of similar instruments available in the market, such as bonds, other loans, or credit derivatives. Thus, it can't assess arbitrage situations arising from relative price mismatches. The future is in improved credit valuation engines.

Financial institutions today face major hurdles in mitigating credit risks ("playing defense"), and simultaneously pursuing risk-adjusted profitability ("playing offense"). Nowhere is this harder than in large corporate lending, where, by most accounts, adequate risk-adjusted returns are hard to achieve.

To support these conflicting objectives, there has been a decade of rapid development of both in-house and commercially available credit risk and performance measurement tools. Broadly, these tools have included:

* Sophisticated, empirical default models used in establishing credit risk ratings for both public and private borrowers.

* Risk-adjusted performance measurement tools, such as risk-adjusted return on capital (RAROC), used in pricing individual loans and assessing risk-adjusted profitability.

* Portfolio management models used in measuring systematic portfolio concentrations with the assistance of correlation analysis.

In this push to implement better credit risk analytics, decision-support models for risk-adjusted pricing have received less emphasis. The reasons for this are varied, but they probably stem from the view that solving the "portfolio problem" was the next key priority after implementing better credit ratings. After all, big concentrations of risk in Texas in the 1980s, real estate in the early 1990s, and Asia most recently have gotten the big headlines.

Another reason for the slow progress on the pricing front relates to the implementation requirements. The development and maintenance of production-quality, desktop decision-support analytics for pricing loans is an enormous task. Across a large financial institution, this requires training and support for hundreds, if not thousands, of relationship and credit officers. After all, isn't it easier to train and support a small handful of portfolio management analysts working in the middle office than a much larger number of users bank-wide?

Yet another reason relates to the approach taken in negotiating loans in an increasingly competitive domestic loan market. Most participants in the large-corporate market focus primarily on the credit spread and fees offered and less on the other, more subtle structural features of the loan. Because of this, bankers often ask why they need a pricing model when the market sets the price.

Finally, and perhaps most important, secondary loan sales and other forms of market-based credit-risk transfer until recently have been severely limited. Consequently, market data on credit risk, especially loans, has been extremely scarce, making it hard to calibrate market-based, risk-adjusted credit pricing models. While the literature on credit-risk pricing models has advanced substantially over the years, the applied work has lagged far behind in large measure because of inadequate data. [1]

With loan trading, credit derivatives, CDOs and other forms of market-based credit risk transfer becoming far more prevalent, the future of risk-adjusted pricing looks much more interesting. Some of the developments that will influence the use of risk-adjusted pricing tools are summarized below.

Reasons to Implement Better Credit Pricing Models

To control the credit business and have an opportunity to make money for shareholders, financial institutions providing or investing in credit must have a way of accurately measuring the value conveyed by a loan, bond, credit derivative or any other credit contract. Credit providers who lack this critical capability are essentially trying to run a business without knowing how to measure profits. …