Using Two-Factor Interest Rate Models for More Accurate MBS Valuation

Beyond Parallel Shifts: The Case for Two-Factor Interest Rate Models in MBS Valuation

For years, the valuation of mortgage-backed securities (MBS) has been dominated by single-factor interest rate models. These models, which assume that all interest rates move in parallel, have been the workhorses of the industry. However, they are increasingly showing their limitations in a market where the yield curve can twist and turn in unpredictable ways. The assumption of parallel shifts is a gross oversimplification of reality, and it can lead to significant errors in MBS valuation and risk management. This is where two-factor interest rate models come in. By allowing for changes in both the level and the slope of the yield curve, these models provide a more accurate representation of the interest rate environment and lead to more accurate MBS valuations.

The Flaw in the Single-Factor Framework

Single-factor interest rate models, such as the Vasicek and Cox-Ingersoll-Ross (CIR) models, are based on the assumption that the entire yield curve can be described by a single stochastic factor, typically the short-term interest rate. This means that all interest rates are perfectly correlated and that the yield curve can only shift up or down in a parallel fashion. In reality, this is rarely the case. The yield curve is constantly changing shape, with the short end and the long end of the curve often moving in different directions. This is particularly true in times of market stress, when the yield curve can invert or steepen dramatically. A single-factor model is simply unable to capture this dynamic behavior, and this can lead to significant errors in the valuation of interest rate-sensitive securities like MBS.

The Power of Two Factors: A More Realistic View

Two-factor interest rate models, as the name suggests, use two stochastic factors to describe the evolution of the yield curve. This allows them to capture a much wider range of yield curve shapes and movements. There are a variety of two-factor models that can be used for MBS valuation, but two of the most common are the Hull-White model and the Black-Karasinski model.

The Hull-White Model

The Hull-White model is a two-factor model that is an extension of the Vasicek model. It assumes that the short-term interest rate is the sum of two components: a deterministic component that is calibrated to the current term structure of interest rates, and a stochastic component that follows a mean-reverting process. This allows the model to perfectly fit the initial yield curve, while also allowing for non-parallel shifts in the yield curve over time.

The Black-Karasinski Model

The Black-Karasinski model is another popular two-factor model. It assumes that the short-term interest rate follows a lognormal process, which means that interest rates can never become negative. This is a key advantage over the Hull-White model, which can allow for negative interest rates. The Black-Karasinski model is also a mean-reverting model, which means that interest rates will tend to revert to a long-run average over time.

The Benefits of Two-Factor Models in MBS Valuation

The use of two-factor interest rate models in MBS valuation offers a number of significant benefits:

• More Accurate Valuation: By providing a more realistic representation of the interest rate environment, two-factor models can lead to more accurate MBS valuations. This is particularly true for MBS with complex prepayment options, such as those with teaser rates or prepayment penalties.
• Improved Risk Management: Two-factor models can be used to more accurately measure and manage the interest rate risk of an MBS portfolio. This is because they can capture the risk of non-parallel shifts in the yield curve, which is a key driver of MBS performance.
• Enhanced Hedging: Two-factor models can be used to develop more effective hedging strategies for MBS portfolios. This is because they can be used to hedge against changes in both the level and the slope of the yield curve.

A Practical Example

Consider an MBS that is backed by a pool of mortgages with a low coupon rate. In a rising interest rate environment, the prepayment speed on this MBS is likely to be very low. A single-factor model, which assumes that all interest rates are rising in parallel, would likely overestimate the prepayment speed on this MBS. This would lead to an overvaluation of the MBS. A two-factor model, on the other hand, would be able to capture the fact that the long end of the yield curve is rising more slowly than the short end. This would lead to a more accurate prepayment forecast and a more accurate valuation of the MBS.

Conclusion: A Necessary Evolution

The use of two-factor interest rate models is a necessary evolution in the valuation of MBS. The assumption of parallel shifts in the yield curve is no longer tenable in today's complex and dynamic market. By adopting the power of two factors, MBS traders can gain a more accurate understanding of the risks and rewards of these complex securities. This will lead to more accurate valuations, more effective risk management, and ultimately, more profitable trading decisions.