Quantitative Finance & Derivatives Margrabe (1978) Model

Margrabe Exchange Option Pricing Lab

Value options to exchange one risky asset for another under William Margrabe's seminal 1978 framework. Analyze combined spread volatility, cross-asset Greeks, spark spreads, and Euler homogeneity.

Market Presets:
Exchange Option Value
$14.285
14.29% of benchmark asset
Combined Spread Vol (σ̂)
36.19%
Spread volatility √(σ₁² + σ₂² - 2ρσ₁σ₂)
Relative Ratio (S₁ / S₂)
1.0000
At-The-Money Parity
Correlation Greek (dC/dρ)
-$18.420
Price change per +1.00 correlation

Exchange Contract Parameters

Payoff: max(S₁ - S₂, 0)

Asset 1 (Acquired Asset S₁)

$100.00
25.0%
0.0%

Asset 2 (Delivered Asset S₂ / Strike)

$100.00
25.0%
0.0%

Cross-Asset Joint Dynamics

+0.50
-1.0 (Opposite Movement) 0.0 (Uncorrelated) +1.0 (Comovement)
1.00 yrs (365 days)

Analytical Formula Breakdown

Margrabe Option Value $14.285
Intrinsic Spread max(S₁ - S₂, 0) $0.0000
Time Value Component $14.285
Combined Volatility (σ̂) 36.19%
d₁ / d₂ Coordinates d₁: 0.1809 | d₂: -0.1809
N(d₁) / N(d₂) Probabilities N(d₁): 0.5718 | N(d₂): 0.4282
Discounted Asset Values S₁e⁻ᑫᵀ: $100.00 | S₂e⁻ᑫᵀ: $100.00
Risk-Neutral ITM Probability 42.82%
Euler Identity: S₁Δ₁ + S₂Δ₂ $14.285 (Exact match)
Margrabe exchange option value as a function of Asset 1 price (S₁), compared to intrinsic payoff and Asset 2 deliverable level.

Cross-Asset Sensitivities & Greeks

Delta 1 (Δ₁) +0.5718
∂C/∂S₁ = e⁻ᑫ¹ᵀ N(d₁) [Long Asset 1]
Delta 2 (Δ₂) -0.4282
∂C/∂S₂ = -e⁻ᑫ²ᵀ N(d₂) [Short Asset 2]
Gamma 1 (Γ₁) +0.0108
∂²C/∂S₁² = Rate of Delta 1 shift
Gamma 2 (Γ₂) +0.0108
∂²C/∂S₂² = Rate of Delta 2 shift
Cross-Gamma (Γ₁₂) -0.0108
∂²C/(∂S₁∂S₂) [Cross curvature]
Corr Greek (∂C/∂ρ) -$0.1842
Dollar change per +0.01 shift in ρ
Vega 1 (𝒱₁) $0.1963
Per 1% shift in Asset 1 volatility
Vega 2 (𝒱₂) $0.1963
Per 1% shift in Asset 2 volatility
Daily Theta (Θ) -$0.0195/day
Calendar time decay per day

Sensitivity 1: S₁ vs S₂ on Exchange Value

Evaluates option premium across joint price movements of Asset 1 and Asset 2 around base spots.

Sensitivity 2: Correlation (ρ) vs Expiry (T)

Demonstrates how negative correlation and longer time horizons amplify the option value by increasing combined volatility σ̂.

Institutional Framework: William Margrabe's (1978) Model

1. Foundation: The Option to Exchange One Asset for Another

Published in the Journal of Finance (1978), William Margrabe solved the valuation of a contract granting the holder the right to surrender asset 2 to acquire asset 1 at expiration date $T$:

Payoff(T) = max(S₁(T) - S₂(T), 0)

Unlike the standard Black-Scholes model where the strike price $K$ is a deterministic, constant cash amount, Margrabe allows the strike itself to be a stochastic asset $S_2(t)$. Both asset prices follow correlated geometric Brownian motions:

dS₁/S₁ = (μ₁ - q₁)dt + σ₁ dW₁
dS₂/S₂ = (μ₂ - q₂)dt + σ₂ dW₂, with Corr(dW₁, dW₂) = ρ dt

2. Change of Numeraire & Closed-Form Formula

By using $S_2(t)$ as the stochastic numeraire instead of the money market account, the relative price $Z(t) = S_1(t) / S_2(t)$ becomes a martingale under the $S_2$-measure with volatility equal to the combined spread volatility:

σ̂ = √(σ₁² + σ₂² - 2ρσ₁σ₂)

Under this measure, the contract is identical to a standard European call on $Z(t)$ with a fixed strike of $1.0$, yielding the exact closed form:

C = S₁ e⁻ᑫ¹ᵀ N(d₁) - S₂ e⁻ᑫ²ᵀ N(d₂)
d₁ = [ln(S₁/S₂) + (q₂ - q₁ + ½σ̂²)T] / (σ̂√T)
d₂ = d₁ - σ̂√T

3. Euler's Homogeneity Identity & Cross-Greeks

Because the pricing function is linearly homogeneous of degree 1 with respect to $(S_1, S_2)$—i.e., $C(\lambda S_1, \lambda S_2) = \lambda C(S_1, S_2)$—Euler's homogeneous function theorem guarantees:

C = S₁ (∂C/∂S₁) + S₂ (∂C/∂S₂) = S₁ Δ₁ + S₂ Δ₂

Where the cross-asset Greeks exhibit remarkable symmetry:

  • Delta 1 (Δ₁): $e^{-q_1 T} N(d_1) > 0$ (long underlying asset 1).
  • Delta 2 (Δ₂): $-e^{-q_2 T} N(d_2) < 0$ (short deliverable asset 2).
  • Cross-Gamma (Γ₁₂): $\frac{\partial^2 C}{\partial S_1 \partial S_2} = -\frac{e^{-q_1 T} n(d_1)}{S_2 \hat{\sigma} \sqrt{T}} < 0$.
  • Correlation Greek: $\frac{\partial C}{\partial \rho} = -\frac{S_1 e^{-q_1 T} n(d_1) \sqrt{T} \sigma_1 \sigma_2}{\hat{\sigma}} < 0$.

4. Industry Applications: Spark Spreads & M&A

Merchant Power Plants (Spark Spread): The option to convert fuel into electricity is modeled as a Margrabe option where $S_1 = \text{Electricity Price}$ and $S_2 = \text{Heat Rate} \times \text{Natural Gas Price}$. A plant operates only when the spark spread is positive.

M&A Takeover Collars & Stock Swaps: Valuing stock-for-stock exchange ratios where an acquiring company's shares are swapped for target company shares at closing.

Exchange Option Mastery Quiz

Unsubmitted

Frequently Asked Questions

A Margrabe exchange option is a multi-asset derivative that grants the holder the right, but not the obligation, to surrender one risky asset (Asset 2) in exchange for another risky asset (Asset 1) at expiration, with terminal payoff max(S1(T) - S2(T), 0). Developed by William Margrabe in 1978, it generalizes Black-Scholes by allowing the strike price itself to be a stochastic asset.

The risk-free rate r cancels out because Asset 2 serves as the stochastic numeraire rather than a deterministic cash bank account. Under the equivalent martingale measure defined by numeraire S2, the relative price S1/S2 behaves like a standard Black-Scholes underlying with a fixed strike of 1.0, eliminating the need for cash discounting.

An exchange option's value decreases monotonically as correlation ρ increases. The combined spread volatility is given by sigma_hat = sqrt(sigma1^2 + sigma2^2 - 2*rho*sigma1*sigma2). When two assets are strongly correlated, they move together, compressing the variance of their spread and diminishing the probability that one dramatically outperforms the other.

Margrabe's formula is widely used to price energy spark spreads (converting natural gas into wholesale electricity), refinery crack spreads (crude oil into refined gasoline/diesel), M&A stock-for-stock takeover offers, dual-currency foreign exchange options, and equity relative performance or outperformance options.

Explore Related Quantitative Finance & Derivatives Labs

Lookback Option Pricing Lab

Model floating and fixed strike lookbacks, extrema path dependency, and Broadie-Glasserman-Kou corrections.

Compound Option Pricing Lab

Model options on options (Geske 1979), critical asset boundary S*, and staged real options valuation.

Barrier Option Pricing Lab

Model single barrier knock-in & knock-out exotic options, in-out parity, and pin risk.

Black-76 Futures Option Lab

Model options on futures, commodity forwards, interest rate caplets, and swaptions.

Bjerksund-Stensland Lab

Closed-form flat boundary approximation for American commodity and equity options.

Black-Scholes Options Lab

Evaluate classic lognormal option pricing, Greeks sensitivity surfaces, and implied volatility root-finding.