What Is Jump Diffusion and How Does It Model Price Behaviour?
Jump diffusion is a class of financial model that describes asset price movements as a combination of two processes: a continuous diffusion component — the gradual, random price fluctuations that occur over time — and a discontinuous jump component that accounts for sudden, large price moves. These jumps represent the kind of abrupt market shifts triggered by unexpected events such as earnings shocks, policy announcements, geopolitical incidents, or liquidity crises.
Continuous vs. Discontinuous Price Movement
| Feature |
Continuous (Diffusion) |
Discontinuous (Jump) |
| Nature of movement |
Gradual, incremental changes |
Sudden, large discrete moves |
| Cause |
Day-to-day market noise and flow |
News events, crises, announcements |
| Modelled by |
Brownian motion (Wiener process) |
Poisson process with random magnitude |
| Frequency |
Constant |
Random, relatively rare |
How the Jump Diffusion Model Works
The most widely referenced jump diffusion model is the Merton (1976) framework, which combines a geometric Brownian motion for continuous price drift with a compound Poisson process that introduces random jumps of random magnitude at random times. The model requires three additional parameters beyond standard diffusion: the expected number of jumps per year (jump intensity), the average jump size, and the volatility of jump sizes.
Jump Diffusion vs. the Black-Scholes Model
The standard Black-Scholes options pricing model assumes that asset prices follow a continuous path with constant volatility — an assumption that fails to account for the observed ‘fat tails’ and volatility smiles in real options markets. Jump diffusion models improve on this by explicitly modelling the probability of sudden large moves, producing option prices that better reflect the market’s observed pricing of deep out-of-the-money options and short-dated contracts where jump risk is most significant.
Applications in Finance
Jump diffusion models are applied across several areas of financial practice:
-
Options pricing: more accurate valuation of contracts sensitive to gap risk and tail events
-
Credit risk modelling: capturing the possibility of a sudden default event rather than a gradual deterioration
-
Equity derivatives: modelling individual stock behaviour around earnings announcements
-
Foreign exchange: accounting for policy-driven currency gaps and central bank interventions
Limitations of Jump Diffusion Models
Jump diffusion models are more realistic than pure diffusion approaches, but they introduce additional complexity and parameter estimation challenges. Calibrating jump intensity and jump size distribution accurately requires sufficient historical data on rare events — which are, by definition, infrequent. The models also assume that jump parameters remain stable over time, which may not hold during extended periods of structural market change or regime shift.
Risk Warning:Trading CFDs involves significant risk and may not be suitable for all investors. Losses can exceed your initial deposit. Please ensure you fully understand the risks involved before trading.
Back
