Python Quants Tutorial 12 - Derivative Analytics - Calibrating an Opti | Refinitiv Developers

In this second part of the Derivative Analytics tutorial we build upon the ease with which the Eikon Data API allows you to work with Options Chains by showing how to calibrate an options pricing model. We will show that this model is capable of approximating prices observed in the real world. We combine the Merton (1976) characteristic function with the Lewis (2001) integration function and finally evaluate the integral via numerical quadrature. To conclude, we compare the model-based prices to those observed in the markets and draw some conclusions. #Eikon #API #Quant #Python #MachineLearning #DataDevelopers #Refinitiv

- Retrieving options data based on chain RICs
- Implementing Merton (1976) Jump-Diffusion Model in Python to calculate a modelled option valuation
- Calibration of the model, based on the Root Mean Squared Error (RMSE), analysis of the results

[giacomobianchi]

Hey! Interesting video! But I have a question:
- Generally, we are not sure that during the optimization that we have a convex optimization problem, so why the decision about using it? It could happen that we find a local minimum in this way, right?

Extra mathematics question: It could happen (because we don't know how is it done) that the objective function could be not differentiable, so taking tools based on optimization with gradient (i.e. methods for convex optimization) could not make sense right? Is the only alternative brute force?

[yousi]

In calibration part, you take option market price = option['CF-Close'], so where is the price of stock in data, could you explaine it to me, PLEASE?