The world of financial modeling is often invisible to those outside of Wall Street, but recent work by Pietro Rossi, a senior financial analyst at Prometeia, is bringing novel sophistication to how risk is assessed in bond markets. Rossi, along with a team based in Bologna, Italy, has developed a new model for pricing bonds based on the likelihood of changes in credit ratings – a crucial factor for insurance companies and other investors. This approach moves beyond traditional methods that focus solely on the probability of default, offering a more nuanced understanding of potential financial outcomes.
The impetus for this innovation came from a specific challenge posed by an insurance company client. Existing models weren’t adequately equipped to handle the complexities of fluctuating credit ratings. “We realised that it was possible to build a framework where you create a stochastic scenario for transition matrices and … reproduce prices of observable bonds according to their actual ratings,” Rossi explained, according to a recent report in Risk.net. The resulting model, detailed in a January paper, is already in apply at the client’s business, demonstrating its practical application.
At the heart of Rossi’s model are “credit transition matrices.” These matrices illustrate the probabilities of a bond’s rating moving up, down, or remaining the same over a given period. The largest probabilities typically reside along the diagonal, indicating that a bond is most likely to maintain its current rating. However, the elements outside the diagonal reveal the likelihood of rating changes, a critical component for accurate pricing. By simulating scenarios using these stochastic, multi-period matrices, the model can estimate bond prices not just at maturity, but at intermediate points throughout the life of the portfolio. This allows for a more comprehensive assessment of a credit portfolio’s potential profit and loss (P&L) distribution and a detailed simulation of its rating composition.
Rossi’s work isn’t limited to credit risk. He’s also been tackling the complexities of volatility modeling, publishing a separate paper earlier this month focused on calibrating models to S&P and Vix options. This research, conducted with Fabio Baschetti of the University of Verona and Giacomo Bormetti of the University of Pavia, introduces a faster calibration technique than previously available. The team’s solution leverages neural networks to learn both S&P volatility and Vix volatility, significantly accelerating the calibration process.
The challenge Rossi and his colleagues addressed – calibrating a volatility model to simultaneously fit the S&P volatility surface and the Vix index volatility curve – has occupied quantitative finance experts for nearly a decade. Researchers like Julien Guyon, Mathieu Rosenbaum, and Jim Gatheral have all contributed to this ongoing effort. Rossi’s approach bypasses the computational bottlenecks of traditional Monte Carlo simulations, offering a substantial speed improvement. He noted that the intellectual challenge itself is a major draw for researchers in the field, potentially even more so than the practical applications.
Rossi’s work builds on the path-dependent volatility model proposed by Julien Guyon and Jordan Lekeufack, as detailed in their 2023 paper, “Volatility is (mostly) path-dependent.” The paper explores the idea that volatility isn’t a constant but is influenced by the historical path of asset prices.
Looking ahead, Rossi and his colleagues are pursuing two primary research directions. One involves a critical examination of widely used interest rate models, specifically testing the validity of the SABR model introduced by Hagan, and others. The other builds on their work with credit rating transitions, exploring whether the stochastic transition framework can be used to price Bermudan or American options on defaultable bonds using least-squares Monte Carlo methods. The results of this research could potentially lead to further publications in Risk magazine.
Rossi discussed these topics and more in a recent podcast on Risk.net’s Quantcast series. The podcast provides a deeper dive into the technical details of his work and the challenges he and his team overcame.
The development of more accurate and efficient financial models, like those pioneered by Rossi and his team, is crucial for managing risk and making informed investment decisions in an increasingly complex global financial landscape. The next step in this ongoing research will be the potential publication of new findings regarding the pricing of Bermudan and American options, further refining the tools available to financial professionals.
What are your thoughts on the evolving landscape of financial modeling? Share your comments below and join the conversation.
