The Arbitrage-Free Frontier: Understanding Bond Pricing and Market Completeness

The pricing of financial instruments, particularly bonds, seems straightforward on the surface. However, the underlying mathematics and the conditions for a "complete" market are surprisingly complex. Tutorial 7 from Andrew Cairns’ derivatives pricing course highlights these intricacies, delving into bond pricing models, arbitrage-free conditions, and the behavior of forward rates. Understanding these concepts is vital for anyone seeking to build robust investment strategies, especially when considering fixed income.

The core premise revolves around the idea that markets should be "complete." A complete market allows investors to perfectly hedge any contingent claim. This isn't about eliminating risk entirely, but rather ensuring that any payoff can be replicated with a combination of existing assets. The tutorial's first question tackles this directly, examining a two-bond market and evaluating its completeness.

The theoretical underpinning of market completeness is rooted in the concept of a risk-free portfolio. If an investor can construct a portfolio whose payoff is identical to any other potential payoff, then the market is considered complete. This implies the existence of a replicating portfolio – a combination of assets that mirrors the payoff of the derivative in question. Conversely, incompleteness arises when such replication isn't possible, meaning certain risks cannot be hedged.

Defining Market Completeness Through Derivative Illustration

The tutorial’s example of two bonds maturing at times 1 and 2 with prices of 0.9 and 0.81 respectively, and subsequent outcome-dependent prices, is a useful illustration. To determine if this market is complete, we must consider whether a derivative's payoff can be replicated using these bonds. If a derivative exists whose payoff cannot be replicated, the market is incomplete.

Imagine a derivative with a payoff of 0.87 if ω1 occurs, 0.91 if ω2 occurs, and 0.89 if ω3 occurs. Can we construct a portfolio of the two bonds to replicate this? The ability to do so depends on whether the weights assigned to each bond allow for this precise payoff across all outcomes. If not, the market is incomplete, meaning a risk premium would be required to trade the derivative.

This lack of completeness has significant implications for pricing. Incomplete markets often exhibit unexplained risk premia, as investors demand compensation for risks they cannot hedge. Conversely, a complete market theoretically eliminates these premia, as all risks are priced and hedgable.

The Dynamics of Arbitrage-Free Zero-Coupon Bond Pricing

The second part of the tutorial introduces a more complex scenario involving the evolution of spot rates. The model assumes a specific form for the risk-free rate, r(t), and provides a framework for how it evolves over time. It further postulates two possible paths for the spot rate curve at time 1, each defined by curves u(s) and d(s). The key question here is how to determine d(s) in terms of d(2) and u(s), ensuring the model remains arbitrage-free.

Arbitrage-free conditions are the bedrock of any consistent pricing model. They dictate that no riskless profit can be generated by simultaneously buying and selling assets. The tutorial’s framework uses the spot rate curve to define the evolution of bond prices, and the arbitrage-free condition demands that these prices be consistent.

The derivation of d(s) involves intricate mathematical relationships based on the model’s assumptions. Crucially, the model reveals that as s approaches infinity, d(s) must also approach infinity. This implies that the long-term spot rate is influenced by the initial conditions and the defined curves u(s) and d(s).

When Arbitrage-Free Conditions Break Down: A Numerical Impossibility

The tutorial then presents a scenario where d(s) is fixed at 0.01 for all s ≥ 2, and u(2) = 0.01. This seemingly innocuous modification leads to a critical conclusion: it becomes impossible to derive values for u(s) for all s that maintain arbitrage-free pricing.

This demonstrates a fundamental constraint on model design. Imposing arbitrary constraints on the spot rate curves can violate the underlying principles of arbitrage-free pricing. The model’s internal consistency is disrupted, revealing that the initial assumptions are not compatible. It serves as a stark reminder that models are only as good as their underlying assumptions.

This highlights a critical aspect of financial modeling: any imposed restriction must be carefully considered to avoid creating arbitrage opportunities. In real-world scenarios, this translates to a need for rigorous model validation and sensitivity analysis.

The Stochastic Drift of the Risk-Free Rate

The fourth section shifts focus to the stochastic behavior of the risk-free rate itself, modeled as a stochastic differential equation (SDE). The equation incorporates a drift term, µ(t, r(t)), and a diffusion term, σ(t, r(t)), which govern the rate’s movement over time. The derivation of the cash account, B(t), using this SDE is a standard result in financial mathematics.

The formula B(t) = B(0) exp(∫t0 r(s)ds) demonstrates how the cash account grows exponentially over time, compounded by the risk-free rate. This is a foundational concept in derivative pricing, as the value of any future cash flow must be discounted back to the present using this risk-free rate.

Understanding the SDE for the risk-free rate is crucial for pricing interest-rate derivatives. The drift and diffusion terms encapsulate the expectations and volatility of interest rates, respectively, and these parameters significantly influence derivative prices.

The Impact of a Jump in the Risk-Free Rate: A Test for Arbitrage

The tutorial then introduces a model with a jump in the risk-free rate at time 1, creating a scenario where r(t) is constant before time 1 and then jumps to a random value, ϵ, after time 1. This jump introduces a discontinuity into the interest rate process, and the model examines whether arbitrage opportunities arise.

The derived equation P(t, T) = EQ[exp(−∫tT r(u)du) | Ft] defines the risk-neutral probability of a future event, given the information available at time t. The proof that no arbitrage opportunity exists at time 0 is a rigorous demonstration of model consistency. It establishes that the model’s pricing mechanism is self-consistent, even in the presence of a discontinuous jump in the risk-free rate.

This type of jump diffusion process is often used to model real-world phenomena, such as unexpected policy changes or economic shocks. The absence of arbitrage opportunities in this model reinforces the importance of carefully considering the impact of such events on asset prices.

Forward Rate Curve Dynamics and the Influence of Distribution Shape

The final section explores the shape of the forward rate curve, f(0, T), under different assumptions about the distribution of a random variable, U, which dictates the risk-free rate after a certain time. The tutorial examines two scenarios: one where U follows an exponential distribution, and another where U follows a different distribution.

The crucial distinction lies in how the distribution shape influences the forward rate curve. In the exponential distribution scenario, the forward rate curve exhibits a characteristic shape dictated by the exponential decay. When U follows a different distribution, the shape of f(0, T) changes accordingly.

The difference in the forward rate curves in parts (a) and (b) highlights the sensitivity of pricing models to distributional assumptions. This underscores the importance of carefully selecting the appropriate distribution to accurately reflect the underlying economic dynamics. The implications extend to pricing interest rate swaps, caps, and floors, all of which are heavily reliant on accurate forward rate curve modeling.