Cox-Ross-Rubinstein Option Pricing Model

First introduced in 1979 by John Cox, Stephen Ross, and Mark Rubinstein in their seminal paper “Option Pricing: A Simplified Approach,” the Cox-Ross-Rubinstein (CRR) model is a foundational, discrete-time framework for valuing financial options. While the Black-Scholes-Merton model revolutionized derivative pricing using continuous calculus, the CRR model provides an intuitive, lattice-based approach.

Its ability to handle early exercise features makes the CRR framework the industry standard for pricing American-style options, complex path-dependent derivatives, and real options in corporate finance.

1. Core Mechanics & Mathematical Structure

The CRR model divides the time to expiration ( $T$ ) into $n$ discrete time steps, where each step represents an interval of $\Delta t = T / n$ .

                ┌─── Su (Up state)
                │
   S0 (Spot) ───┤
                │
                └─── Sd (Down state)

During each step $\Delta t$ , the underlying asset price $S$ can move in only one of two directions:

Upward to $S \cdot u$ with risk-neutral probability $p$
Downward to $S \cdot d$ with risk-neutral probability $1 - p$

Key Formulas

To match the mean and volatility ( $\sigma$ ) of the underlying stock price’s continuous log-normal distribution, CRR defined the parameter calibration as:

$u = e^{\sigma \sqrt{\Delta t}}$

$d = e^{-\sigma \sqrt{\Delta t}} = \frac{1}{u}$

The property $u \cdot d = 1$ makes the tree recombinant—meaning an “up” move followed by a “down” move ( $u \cdot d$ ) yields the same asset price as a “down” move followed by an “up” move ( $d \cdot u$ ). Recombination keeps computational nodes growing linearly ( $n + 1$ terminal nodes) rather than exponentially ( $2^n$ ).

Under risk-neutral valuation, the probability of an upward movement $p$ is expressed as:

$p = \frac{e^{r \Delta t} - d}{u - d}$

Where $r$ represents the continuously compounded risk-free interest rate.

2. Valuation Process: Backward Induction

Option valuation follows a multi-step forward-and-backward process:

Step 1: Asset Price Tree Construction

Working forward from time $t = 0$ to maturity $T$ , calculate the underlying asset price at every node:

$S_{j, k} = S_0 \cdot u^j \cdot d^{k-j}$

Where $k$ is the current time step ( $0 \le k \le n$ ) and $j$ is the number of upward moves ( $0 \le j \le k$ ).

Step 2: Payoff at Expiration

Calculate the option’s intrinsic value at each terminal node ( $k = n$ ):

Call Option: $C_{j, n} = \max(S_{j, n} - K, 0)$
Put Option: $P_{j, n} = \max(K - S_{j, n}, 0)$

Where $K$ is the strike price.

Step 3: Backward Propagation

Move backward step-by-step through the tree to discount expected future values back to time $t = 0$ .

For European Options:
$V_{j, k} = e^{-r \Delta t} \left[ p \cdot V_{j+1, k+1} + (1 - p) \cdot V_{j, k+1} \right]$
For American Options (Accounting for Early Exercise):
$V_{j, k} = \max\left( \text{Intrinsic Value}_{j,k}, \,\, e^{-r \Delta t} \left[ p \cdot V_{j+1, k+1} + (1 - p) \cdot V_{j, k+1} \right] \right)$

3. Real-World Corporate Example: Valuing American Put Options

A major limitation of the standard Black-Scholes model is its inability to accurately handle early exercise premium for American put options on dividend-paying stocks. The CRR model resolves this by evaluating early exercise at every node.

Strategic Application: Energy Sector Risk Management

Consider Equinor ASA, the Norwegian energy giant listed on the NYSE. Institutional desks often manage exposure to commodity price volatility using American put options.

Scenario: An energy trading desk holds an American put option on Equinor stock.
Underlying Spot Price ( $S_0$ ): USD30.00
Strike Price ( $K$ ): USD30.00
Risk-Free Rate ( $r$ ): 4.0% annualized
Volatility ( $\sigma$ ): 25%
Time to Maturity ( $T$ ): 6 months ( $0.5$ years)

       CRR 2-Step Tree Structure
       -------------------------
                 Node 2,2 (S = USD38.40, Put = USD0.00)
                /
         Node 1,1 (S = USD33.94, Put = USD0.00)
        /       \
  Node 0,0       Node 2,1 (S = USD30.00, Put = USD0.00)
 (S = USD30.00)   /
        \       /
         Node 1,0 (S = USD26.52, Put = USD3.48)  <-- Early exercise check
                \
                 Node 2,0 (S = USD23.44, Put = USD6.56)

Using a simplified 2-step CRR model ( $\Delta t = 0.25$ years):

$u = e^{0.25 \cdot \sqrt{0.25}} = 1.1331$
$d = 1 / 1.1331 = 0.8825$
$p = \frac{e^{0.04 \cdot 0.25} - 0.8825}{1.1331 - 0.8825} = 0.5091$

Calculations & Decision Points:

At Node 1,0 (S = USD26.52):
- Continuation Value (holding the option): USD3.18
- Intrinsic Value (exercising immediately): USD30.00 – USD26.52 = USD3.48
Outcome: The CRR model identifies that early exercise at Node 1,0 yields USD3.48 versus USD3.18 for holding. The model overrides the continuation value, pricing the node at USD3.48.

This ability to model optimal early exercise strategies provides financial institutions with a more accurate valuation of American options compared to standard continuous-time formulas.

4. Analytical Comparison: CRR vs. Black-Scholes-Merton

Feature	Cox-Ross-Rubinstein (CRR)	Black-Scholes-Merton (BSM)
Time Structure	Discrete time steps	Continuous time
Mathematical Basis	Binomial distribution / Lattices	Partial Differential Equations (Geometric Brownian Motion)
Option Styles	European, American, Bermudan	Strictly European (without complex modifications)
Dividends & Cash Flows	Easily handles discrete dividend payments	Best for continuous dividend yields
Convergence	As $n \to \infty$ , CRR converges directly to BSM	Closed-form exact solution for European options

The Convergence Principle

By the Central Limit Theorem, as the number of time steps $n \to \infty$ , the discrete binomial distribution of price returns converges to a continuous log-normal distribution. Consequently, the CRR option price converges to the Black-Scholes price.

5. Modern Usage in Quantitative Finance

While modern High-Frequency Trading (HFT) environments rely on advanced numerical schemes like finite difference methods or Monte Carlo simulations, the CRR framework remains widely used in institutional finance:

Exotic Derivatives Valuation: Used as a base lattice for Bermudan swaptions and barrier options with complex early-stopping boundaries.
Corporate Real Options: Utilized by corporate finance desks at firms like BP and Shell to value managerial flexibility in capital budgeting (e.g., deciding whether to expand, defer, or abandon long-term infrastructure investments).
Model Adjustments: Enhanced variants, such as the Leisen-Reimer model, build upon CRR’s lattice architecture while speeding up convergence rates to smooth out binomial oscillations.