WASHINGTON - Christine Cummings gives few outward hints that she is the guru of risk assessment at the New York Fed.
The University of Minnesota PhD speaks in plain English, avoiding the seven-syllable words that tend to leap from most economists' mouths. She's practical and down-to-earth, not someone likely to be lost in a cloud of cosigns and thetas.
Yet the New York Federal Reserve Bank's senior vice president for specialized examinations is clearly one of the country's foremost experts in the arcane yet pivotal field of risk assessment. Ms. Cummings knows what makes risk models work, and she's determined to explain it to bank examiners.
In the process, the face of many large bank examinations could change dramatically.
Risk-based models, as opposed to evaluations that focus on asset quality, could become the core of megabank exams, she said.
Ms. Cummings said many large banks are using highly sophisticated risk assessment models that accurately peg their exposure to changes in interest rates and economic conditions.
The Federal Reserve Board and the Bank for International Settlements have proposed allowing banks to set their capital requirements based on these models.
The catch is that regulators are going to have to review the models to make sure they work. While the models may be complex - they require the bank to plug risk factors into a probability equation - they demand only slightly more attention from regulators than a traditional loan file review.
Banks hire experts to build their models, just as they hire senior loan officers to oversee their credit programs. Compliance officers, internal auditors, and external auditors then review the model, just as they would examine a loan program. They should identify any flaws, she said.
"We want to make sure you have independent validation," she said.
These "checks and balances" allow banks to regulate themselves, she said.
For added safety, senior examiners with technical experience also can dissect the model. But if the other checks are working, this review won't be as important, she said.
Ms. Cummings currently is trying to squelch fears about a second Fed proposal. It would let a bank use internal models to determine its capital requirements for trading activities. If the model is off, the bank would pay the Fed a penalty.
She said bankers tell her to keep the size of the penalty low and to make the rules broad enough so they can use their current models. The Fed, which has not finalized the rule yet, is taking both of those concerns seriously, she said.
Ms. Cummings began her Fed career studying the Italian economy, steadily rising through the organization until May 1987 when she jumped ship to work in the international economics department at Morgan Guaranty Trust Co. Eight months later she was back at the New York Fed as a senior international economist.
She said the New York Fed is simply a fun place to work, combining the abstract challenges of academia with real-world problem solving.
"The Fed does some of both," she said. "How do we understand the world, and what should we do about it?"
Ms. Cummings said her challenge won't end once the field examiners begin reviewing risk models.
"The most interesting thing about my job is looking at the world and trying to figure out why it is changing and what regulatory policies we should have," she said. "It is a never-ending puzzle that we have to attempt to solve."