Managing credit risk at a large institution like Banc of America Securities, with its global offices and multiple product offerings, is no easy task.
But Jeffrey Rosenberg, head of BoA Securities' Credit Strategy, thinks his team has developed a product that gives his firm an edge by looking at the options market - in addition to stock prices and fundamentals - to assess the future credit risk of firms it follows. It does this by using a proprietary model called Credit Option Adjusted Spread (COAS).
BoA Securities is so sure of its success that it is now taking the COAS-based product to the market in a deal with Standard & Poor's that, Rosenberg says, will take it to the "next level." S&P and BoA Securities will develop a new suite of quantitative credit-risk-analytics tools based on the model.
"The model has historically given pretty good results," he says, indicating that while "Enron was a little bit more difficult for the model to pick up," because of Sept. 11, back testing "showed a dramatic increase in Enron risk" before its collapse.
As well, he says, applying the model to Worldcom and the telecom sector, which "didn't have the noise" and were "a little bit more clear," it would have "triggered sell signals under any native-trading program," because there were "extremely large risk increases."
Investment Managers Need Good Credit
In volatile and choppy waters, staying on top of credit risk is particularly important for investment managers. The problem is that most tools are backward looking, based on historical data, rather than looking at the present. Looking forward into the crystal ball is far more difficult.
David Goldman, head of debt research at BoA Securities, part of the BoA Corporation, says, "Unprecedented credit-market volatility has challenged everyone's credit-risk assessment, and there is clearly a need for a better way to interpret market data which may provide forward-looking signals for credit. Our collaboration is designed to help us provide tools that do just that."
Risk On the Rise
Rosenberg, who also heads up COAS development, notes that, "There's been a tremendous increase in credit risk." The default rate of speculative corporate-grade bonds is at a peak not seen since the 1991 recession. If a person looks behind the numbers, he says, "It's a dramatically more severe default experience than the last time."
When it comes to investment-grade bonds, the default rate is typically near zero. For these securities, the problem comes with downgrades, such as going from triple A status to a lower level. Investment managers can lose their shirts over a few dozen basis points, so they want to stay on top of developments that could lead to a downgrade.
"Investing in credit is a loser's game," Rosenberg says. "There's limited upside and unlimited downside. You win by avoiding losses."
He says that COAS and the S&P collaboration is "not about replacing ratings and fundamental credit-risk analysis. It's augmenting with another tool for managing the credit-risk experience. It is an important development in quantitative credit-management techniques."
"We have designed our model for institutional investors around a very broad set of market-based data," says Rosenberg, one that is broader that other models currently in the market.
The key, he says, is that it not only relies on balance-sheet information and stock price, but incorporates implied volatility as well as credit spreads by looking at the options market and how that market is pricing securities going forward.
It covers the top 300 investment-grade credit issuers, which represent between 90 and 95 percent of the total-dollar volume of investment-grade financial instruments that are issued.
He says options are like an insurance policy. Looking ahead at the price that people are willing to pay for the right to buy or sell a stock provides an indication of how the market views a company going forward.
Volatility Versus Stock Price
Rosenberg notes that between October 2002 and April 2003, prior to the spring rally, the credit spread between investment-grade securities and treasuries collapsed from 280 basis points over treasuries to 140, indicating growing comfort with the risk of corporate bonds. Yet stocks didn't start rallying till March.
He says the "big decline in credit risk wasn't shown in stock-market prices, it was shown in implied volatility." The VIX index, which measures volatility, hit a high in October and by April it had dropped almost in half.
Over a long period of time, Rosenberg says, implied volatility is a better measure for credit risk than stock prices.
BoA Securities has created a Web site to support COAS that its institutional clients can access for research purposes. It provides users with a number of research options.
For example, it provides a warning indicator based on the stoplight system. Red is a dangerous investment, yellow suggests caution and green is a thumbs up. Users can access a list of "yellow-light" companies, those that have deteriorated from a green listing. As well, it has a "red-light" list, indicating which companies, if any, have deteriorated further. There is also a button that indicates companies whose credit risk is improving.
Additionally, there are weekly lists and a top-25 list of the most risky companies. Users can enter a stock symbol and get a credit-risk report and there are a variety of charts comparing equity price versus option-implied volatility and credit risk versus asset-swap spread.
William Chambers, managing director in the risk-solutions group at Standard and Poor's, says, "The marketplace has a need and desire" for both fundamental and quantitative tools for assessing credit risk. "They're different ways of looking at the same problem."
Chambers says the new COAS-based product will compete with risk vendors such as Moody's KMV, RiskMetrics and Kamakura, to name a few.
He refers to Moody's KMV as the "big fish in the pond" when it comes to risk-management tools for credit. It has a range of solutions, including CreditMonitor and CreditEdge.
As well, RiskMetrics Group has a number of credit-risk offerings, including CreditGrades, a Web-delivered model that looks at historical equity prices and balance-sheet information and CreditManager, which manages portfolio risk.
In addition, Kamakura Corporation has developed some credit-risk models. Chambers says, "Each one is slightly different in terms of the approach."
Chambers says, with the Basel II Accord, there "is going to be a lot of pressure for banks, across the board, to improve the overall quality of risk assessment. There has been a great deal of debate in the marketplace, and by regulators, about how best to use those (different) models."
Martha Sellers, managing director of client solutions for the Americas at Moody's KMV in San Francisco, says that creating forward-looking risk management is definitely important. But implied volatility isn't the best measure.
She says the model has been around since the 1970s, but it "just didn't work when implemented. It understates the risk of default for a lot of companies and is not sufficiently forward looking. It's a nice theory but it doesn't work in practice." She adds that implied volatility is a biased measure and the time horizon of options is too short. "Credit investors' horizons are longer than that."
BoA Securities, of course, disagrees and goes to great length in a 30-page documents posted at the COAS portion of its Web site to explain the model, and how it successfully caught the telecom market meltdown.
Says Rosenberg. "This system is designed to be much more sensitive to the market signals and sentiments than any other available model."