Option Pricing Models and Lattices

Option pricing models connect a contract’s payoff with assumptions about the underlying price, volatility, rates, dividends or carry, exercise rights, and settlement. The appropriate model depends on the risk being modeled: Heston adds stochastic variance, Hull-White models short rates, and lattice or simulation methods handle exercise and payoff features that a direct formula may not.

A more complex model is not automatically better. It adds value only when the additional factors are material, supported by market evidence, implemented correctly, and controlled through calibration and validation.

Choose a Model by the Contract

Models and numerical methods are different concepts. Heston or Hull-White specifies risk-factor dynamics; a characteristic-function integral, lattice, finite-difference grid, or simulation is a way to calculate value under those dynamics.

Model-Selection Workflow

    flowchart LR
	    A["Read payoff, exercise, and settlement terms"] --> B{"What risk factors drive value?"}
	    B --> C["Equity, FX, or commodity price and volatility"]
	    B --> D["Interest-rate curve and rate volatility"]
	    C --> E{"Early exercise or path dependence?"}
	    E -->|"No"| F["Benchmark formula or stochastic-volatility model"]
	    E -->|"Yes"| G["Lattice, finite difference, or simulation"]
	    D --> H["Term-structure model plus product-specific method"]
	    F --> I["Calibrate to market evidence"]
	    G --> I
	    H --> I
	    I --> J["Validate prices, sensitivities, limits, and implementation"]

This workflow does not select one universal model. A barrier equity option may use Heston dynamics with finite differences, while a Bermudan swaption may use Hull-White dynamics on a lattice.

Evidence Required

• Complete payoff, exercise, observation, and settlement terms.
• Current spot, forward, discount, dividend, carry, and curve inputs.
• Option quotes across relevant strikes, expiries, and tenors.
• Calibration objective, quote filters, weights, parameter bounds, and tolerances.
• Numerical method, convergence tests, benchmark cases, and code version.
• Independent price verification and executable bid-ask evidence where available.
• Documented intended use, limitations, validation findings, and valuation adjustments.

Practical Comparison

Consider three contracts:

• A six-month European stock call can use Black-Scholes as a transparent benchmark.
• A five-year American-style employee or equity option requires an early-exercise method such as a lattice or finite-difference approach.
• A receiver swaption depends on an interest-rate curve and rate volatility, making a term-structure model such as Hull-White more relevant than an equity stochastic-volatility model.

The product label alone is insufficient. An exotic option may require both a stochastic process and a numerical method tailored to barriers, averages, multiple underlyings, or exercise dates.

Common Mistakes

• Selecting a model because it is sophisticated rather than because it matches the payoff and risk factors.
• Treating a fitted market surface as evidence that model parameters are true or stable.
• Confusing historical estimation with risk-neutral pricing calibration.
• Reporting theoretical value without bid-ask, liquidity, and transaction context.
• Ignoring curve construction, dividends, funding, collateral, settlement, or contract multipliers.
• Comparing model errors without using consistent units and quote timestamps.

Use current contract documents, market data, calibration records, model documentation, validation evidence, and account records for an actual valuation. This branch is educational and does not provide personalized investment, derivatives, valuation, legal, accounting, or tax advice.

In this section

Choose a subsection first. Deeper term pages live inside each subsection, which keeps large topic hubs readable.