Pragmatic Comparison Analysis of Alternative Option Pricing Models

TL;DR

This paper compares three stochastic volatility models for European call options, demonstrating their computational efficiency, ease of calibration, and comparable accuracy across different datasets and markets.

Contribution

It provides a pragmatic comparison of alternative stochastic volatility models, highlighting their efficiency and performance in real-world option pricing scenarios.

Findings

SV models achieve similar accuracy to existing models.

Calibration of SV models is faster and easier.

Models perform consistently across different markets.

Abstract

In this paper, we price European Call three different option pricing models, where the volatility is dynamically changing i.e. non constant. In stochastic volatility (SV) models for option pricing a closed form approximation technique is used, indicating that these models are computationally efficient and have the same level of performance as existing ones. We show that the calibration of SV models, such as Heston model and the High Order Moment based Stochastic Volatility (MSV) is often faster and easier. On 15 different datasets of index options, we show that models which incorporates stochastic volatility achieves accuracy comparable with the existing models. Further, we compare the In Sample and Out Sample pricing errors of each model on each date. Lastly, the pricing of models is compared among three different market to check model performance in different markets. Keywords: Option Pricing Model, Simulations, Index Options, Stochastic Volatility Models, Loss Function http://www.sci-int.com/pdf/638279543859822650.pdf