The Evolution of Short Rates and the Shape of the Term Structure

Welcome to the World of Short Rates!

Hello future FRM! Today we are diving into a crucial chapter of the Market Risk section: The Evolution of Short Rates and the Shape of the Term Structure. If you have ever wondered how we predict where interest rates are going or how we price complex bonds, this is where the magic happens. Don't worry if stochastic calculus sounds intimidating—we are going to break it down into simple, logical steps that anyone can follow. Let’s get started!

What is the Short Rate?

Before we look at the models, we need to understand what we are modeling. The short rate (often denoted as \( r \)) is the interest rate applicable for an infinitesimally short period of time. Think of it as the "instantaneous" interest rate.

Analogy: Imagine you are driving a car. Your speedometer shows your speed at that exact second—that is like the short rate. Your total trip time and average speed are more like the yield on a 10-year bond. In this chapter, we want to see how that "speedometer" changes over time.

Quick Review: We use the short rate to build the entire term structure (the yield curve). If we know how the short rate behaves, we can price bonds and derivatives of all maturities.

Two Philosophies: Equilibrium vs. No-Arbitrage

In the FRM curriculum, short-rate models are generally divided into two "schools of thought":

1. Equilibrium Models: These start with assumptions about economic variables. They usually produce a term structure as an output. Famous examples include the Vasicek and CIR models. They are great for long-term forecasting but might not perfectly match today’s market prices.

2. No-Arbitrage Models: These models take today's market yield curve as an input. They are designed to be perfectly consistent with current market prices. The Ho-Lee model is a classic example here. They are heavily used for pricing derivatives because they don't allow for "free lunch" (arbitrage) opportunities against current prices.

The Building Blocks: Drift and Volatility

Almost all short-rate models follow a similar structure called a Stochastic Differential Equation (SDE). It looks like this:

\( dr = [\text{Drift}] dt + [\text{Volatility}] dw \)

Drift (\( dt \)): This is the "expected" or predictable part of the rate change over time.
Volatility (\( dw \)): This is the "random" part. The \( dw \) represents a Wiener process (random walk) which provides the "shocks" to the system.

Model 1: The Simplest Model (Random Walk)

The most basic model assumes the short rate follows a simple random walk with no drift.

\( dr = \sigma dw \)

The Problem: In this model, the interest rate is just as likely to go up as it is to go down. This is unrealistic because it implies interest rates could easily become negative or grow to infinity without any "pull" back to reality.

Model 2: The Ho-Lee Model (Adding Drift)

To make the model more realistic, the Ho-Lee model adds a time-dependent drift.

\( dr = \theta(t) dt + \sigma dw \)

In this equation, \( \theta(t) \) is a function that ensures the model fits today's initial term structure.
Key Takeaway: Ho-Lee is a no-arbitrage model. It captures the drift needed to match market bond prices perfectly. However, its main weakness is that interest rates can still theoretically become negative, and it lacks "mean reversion."

Model 3: The Vasicek Model (Mean Reversion)

The Vasicek model is a major step forward because it introduces mean reversion. This is the idea that if interest rates get too high or too low, they will eventually be pulled back to a "long-run average."

\( dr = k(\theta - r) dt + \sigma dw \)

Components:
- \( \theta \): The long-run mean (the level the rate wants to return to).
- \( k \): The speed of reversion (how hard the "rubber band" pulls).
- \( r \): The current short rate.

Analogy: Imagine a rubber band attached to a ball. The ball is the interest rate, and the anchor point is \( \theta \). If the ball moves far away, the rubber band (\( k \)) pulls it back toward the anchor.

Common Mistake to Avoid: Don't confuse the \( \theta \) in Vasicek (a constant level) with the \( \theta(t) \) in Ho-Lee (a time-varying function). In Vasicek, if \( r > \theta \), the drift becomes negative, pushing the rate down. If \( r < \theta \), the drift becomes positive, pushing the rate up.

Model 4: The Cox-Ingersoll-Ross (CIR) Model

The CIR model is very similar to Vasicek, but it fixes one major problem: volatility. In Vasicek, volatility (\( \sigma \)) is constant. In CIR, volatility depends on the level of the interest rate.

\( dr = k(\theta - r) dt + \sigma \sqrt{r} dw \)

Why the \( \sqrt{r} \) matters:
1. As the interest rate (\( r \)) approaches zero, the volatility (\( \sigma \sqrt{r} \)) also approaches zero.
2. This makes it much harder for interest rates to ever become negative, which is more economically "logical."

Quick Review: Both Vasicek and CIR are equilibrium models. They both feature mean reversion. The primary difference is that CIR has "interest-rate-dependent volatility."

Summary of Models

Ho-Lee: No-arbitrage, no mean reversion, can go negative.
Vasicek: Equilibrium, has mean reversion, constant volatility, can go negative.
CIR: Equilibrium, has mean reversion, volatility depends on \( r \), stays positive.

The Shape of the Term Structure

How do these models affect the shape of the yield curve? The yield curve is influenced by three main factors within these models:

1. Current Rate (\( r \)): If the current short rate is very low, the curve is likely to be upward sloping as mean reversion pulls it up.
2. Mean Reversion (\( k \) and \( \theta \)): The stronger the pull toward the long-run mean, the faster the long end of the curve levels off.
3. The Risk Premium: Investors usually demand a "premium" for holding longer-term bonds. Models often incorporate a "market price of risk" to adjust the drift so that it reflects actual market prices rather than just the physical expectation of rates.

Did you know? In a mean-reverting model, the long-term yield (forward rate) eventually stays flat at a certain level. This is why you often see yield curves "flatten out" at the 20 or 30-year mark.

Key Terms to Remember

Drift: The average change in the interest rate over time.
Stochastic Process: A process involving a random variable (like interest rates).
Mean Reversion: The tendency of a variable to return to a long-term average level.
No-Arbitrage: A condition where the model is calibrated to match current market prices exactly.

Final Tips for the Exam

1. Identify the Model: If you see a square root of \( r \) in the volatility term, it is CIR. If you see \( k(\theta - r) \), it is a mean-reverting model.
2. Understand Mean Reversion: Remember that higher speed of reversion (\( k \)) means the short rate returns to \( \theta \) faster, which makes the long end of the yield curve less sensitive to current short-rate changes.
3. Don't Panic on the Math: Focus on the logic of the equations. The FRM exam often tests your understanding of the properties of these models rather than just your ability to solve complex calculus.

You've got this! Understanding how these rates evolve is the first step to mastering market risk management. Keep pushing forward!