Deep Calibration With Artificial Neural Network: A Performance Comparison on Option Pricing Models

Paper: arXiv 2303.08760

Authors: Young Shin Kim, Hyangju Kim, Jaehyung Choi

Abstract

This paper explores Artificial Neural Network (ANN) as a model-free solution for a calibration algorithm of option pricing models. We construct ANNs to calibrate parameters for two well-known GARCH-type option pricing models: Duan’s GARCH and the classical tempered stable GARCH that significantly improve upon the limitation of the Black-Scholes model but have suffered from computation complexity. To mitigate this technical difficulty, we train ANNs with a dataset generated by Monte Carlo Simulation (MCS) method and apply them to calibrate optimal parameters. The performance results indicate that the ANN approach consistently outperforms MCS and takes advantage of faster computation times once trained. The Greeks of options are also discussed.

Complexity vs Empirical Score

• Math Complexity: 6.5/10
• Empirical Rigor: 6.0/10
• Quadrant: Holy Grail — high math complexity, high empirical rigor

Why this score: The paper tackles a computationally expensive problem by substituting traditional Monte Carlo calibration with deep learning function approximators for advanced GARCH and Tempered Stable option pricing models. While the use of neural networks for calibration is an established paradigm, applying it to non-Gaussian GARCH option pricing frameworks offers practical utility and decent technical depth.

Research Flowchart

  flowchart TD
    A[Goal: Fast & Accurate Calibration of Complex GARCH Option Models] --> B[Data Generation: Simulate Option Prices via Monte Carlo MCS]
    B --> C[Dataset: Model Parameters & Corresponding Option Prices]
    C --> D[Model Training: Train Artificial Neural Network ANN]
    D --> E[Calibration & Analysis: Estimate Parameters & Compute Option Greeks]
    E --> F[Findings: ANN Outperforms MCS in Speed & Accuracy]