Dynamic Initial Margin via Chebyshev Tensors

TL;DR

This paper introduces two Chebyshev tensor-based methods for efficiently computing dynamic sensitivities in Monte Carlo simulations to determine initial margin, outperforming existing regression and neural network approaches in accuracy and speed.

Contribution

The paper presents novel Chebyshev tensor techniques for dynamic sensitivity calculation, improving accuracy and efficiency over traditional and neural network methods.

Findings

Chebyshev methods achieve higher accuracy than benchmarks.

The methods are faster and easier to implement.

Performance surpasses regression and deep neural network approaches.

Abstract

We present two methods, based on Chebyshev tensors, to compute dynamic sensitivities of financial instruments within a Monte Carlo simulation. These methods are implemented and run in a Monte Carlo engine to compute Dynamic Initial Margin as defined by ISDA (SIMM). We show that the levels of accuracy, speed and implementation efforts obtained, compared to the benchmark (DIM obtained calling pricing functions such as are found in risk engines), are better than those obtained by alternative methods presented in the literature, such as regressions (\cite{Zhu Chan}) and Deep Neural Nets (\cite{DNNs IM}).