Derivative Pricing in Practice: Stochastic Models, FFT, and Valuation Adjustments

Summary

The discussion offers brief practitioner-oriented guidance on derivative pricing methods. It describes stochastic-volatility and Lévy-process models as popular, while suggesting jump-diffusion models are less common. It confirms that Fourier transform techniques are used in practice and says Monte Carlo and lattice methods both appear in pricing work. It does not rank these methods by product, implementation, or computational trade-off.

The answer also shifts attention from pricing structured products toward valuation adjustments for vanilla derivatives, recommending valuation adjustments as a potentially relevant thesis topic. This is a concise professional perspective rather than a detailed survey: it provides no named adjustment framework, model specification, implementation examples, or empirical evidence. Its advice can guide topic selection, but practitioners’ methods will depend on the product and institution.

Key ideas

• Stochastic-volatility and Lévy-process models are described as popular, while jump diffusion is said to be less common.
• Fourier transform techniques are used in derivative pricing practice.
• Monte Carlo and lattice methods are both mentioned as practical pricing approaches.
• Valuation adjustments for vanilla products are proposed as a timely thesis direction.

Tags

# What models / methods are used in practice in derivative pricing?


# What models / methods are used in practice in derivative pricing?












I wrote my bachelor thesis about European Option Pricing under Stochastic Volatility and Jump Diffusion and am now near the end of my MSc in Quant Finance. As i want to write a "potential job"-oriented master's thesis I wanted to hear some advice. So maybe a pretty "dumb" question: do practitioners use for instance stochastic volatility jump diffusion models? And are Derivatives/Structured Products priced mainly via Monte Carlo (using variance reduction techniques and assuming there are no closed form solutions available) or do banks employ more complicated techniques like fast fourier transform? Can anyone share a practical insight?

Best regards

## Answer by Mark Joshi (score 6, accepted)

https://quant.stackexchange.com/a/26393

stochastic vol and Levy process models are popular. Jump diffusion less so.

FT techniques are definitely used.

These days most of the focus is on valuation adjustments for vanilla products rather than how to price structured products. It tends to use both MC and lattice methods. If you want to be topical, I'd advise something related to valuation adjustments.

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.