Python for Finance: Analyze Big Financial Data

(Elle) #1

Conclusions


This chapter presents a larger, realistic use case for the application of the DX analytics


library to the valuation of a portfolio of nontraded American options on the VSTOXX


volatility index. The chapter addresses three main tasks involved in any real-world


application:


Data gathering


Current, correct market data builds the basis of any modeling and valuation effort in


derivatives analytics; we need index data and futures data, as well as options data for


the VSTOXX.


Model calibration


To value, manage, and hedge nontraded options and derivatives in a market-


consistent fashion, one needs to calibrate the model parameters to the relevant option


market quotes (relevant with regard to maturity and strikes). Our model of choice is


the square-root diffusion, which is appropriate for modeling a volatility index; the


calibration results are quite good although the model only offers three degrees of


freedom (kappa as the mean-reversion factor, theta as the long-term volatility, and


volatility as the volatility of the volatility, or so-called “vol-vol”).


Portfolio valuation


Based on the market data and the calibrated model, a portfolio with the American put


options on the VSTOXX is modeled and major statistics (position values, Deltas, and


Vegas) are generated.


The realistic use case in this chapter shows the flexibility and the power of the DX library;


it essentially allows us to address any analytical task with regard to derivatives. The very


approach and architecture make the application largely comparable to the benchmark case


of a Black-Scholes-Merton analytical formula for European options. Once the valuation


objects are defined, you can use them similarly to an analytical formula — and this despite


the fact that underneath the surface, heavy numerical routines and algorithms are applied.

Free download pdf