Abstract
The interpolation of the market implied volatility function from several observations of option prices is often required in financial practice and empirical study. However, the results from existing interpolation methods may not satisfy the property that the European call option price function is monotonically decreasing and convex with respect to the strike price. In this paper, a modified convex interpolation method (with and without smoothing) is developed to approximate the option price function while explicitly incorporating the shape restrictions. The method is optimal for minimizing the distance between the implied risk-neutral density function and a prior density function, which allows us to benefit from nonparametric methodology and empirical experience. Numerical performance shows that the method is accurate and robust. Whether or not the sample satisfies the convexity and decreasing constraints, the method always works.
Original language | English |
---|---|
Pages (from-to) | 243-266 |
Number of pages | 24 |
Journal | Journal of Optimization Theory and Applications |
Volume | 142 |
Issue number | 1 |
DOIs | |
Publication status | Published - 4 Mar 2009 |
Keywords
- Implied volatility
- Nonparametric estimation
- Option price function
- Risk-neutral density
- Shape-preserving interpolation
ASJC Scopus subject areas
- Control and Optimization
- Management Science and Operations Research
- Applied Mathematics