Workshop on Stochastic Processes: Estimation and Hedging will be held on 2 May at Fuglesangs Allé. The workshop is organised by Bent Jesper Christensen.

Please sign up for the workshop no later than 28 April 2025: LINK

Additional information:

All interested parties are invited to the PhD defence of Anders Norlyk at 14:30 and the following reception: LINK

Program

08:45-09:00     Registration and welcome including coffee and breakfast rolls
09:00-09:45     Mikko Pakkanen, Imperial College: Hedging Derivatives by Machine Learning
09:45-10:30     Orimar Sauri, Aalborg University: Trawl Processes: Limit Theorems and Non-parametrical Inference
10.30-10:45     Coffee break
10:45-11:30     Mikkel Bennedsen, Aarhus University: Composite Likelihood Estimation of Stationary Gaussian Processes With a View Toward Stochastic Volatility
11:30-12:30      Lunch
 

Abstracts of workshop talks

Mikko Pakkenen, Imperial College: Hedging Derivatives by Machine Learning

Recent advances in machine learning (ML) have given rise to promising data-driven, and often model-free, methodologies for automating decision-making for various tasks in finance, such as trading, risk management and portfolio construction. In this talk, I will focus on the hedging of financial derivatives, which can be tackled by deep learning and reinforcement learning, although pitfalls and challenges remain. I will first introduce the basic framework of hedging by neural networks and demonstrate its applicability by computational examples. I will then address the difficulty of training ML-based hedging strategies using data that exhibit drifts and propose a solution to mitigate the problem. Finally, I will present steps towards a general, optimal hedging methodology, building on risk-averse reinforcement learning.

Orimar Sauri, Aalborg University: Trawl Processes: Limit Theorems and Non-parametrical Inference

Trawl processes is a class of continuous-time infinitely divisible stationary processes whose correlation structure is completely described by its so-called trawl function. In the first part of this talk we will discuss some probabilistic properties as well as some limit theorems for this class of processes. In the second part we will present a non-parametrical method for estimating non-linear functionals of a trawl function under an in-fill and a long-span sampling scheme. Specifically, building on the work of Sauri and Veraart (Nonparametric estimation of trawl processes: Theory and applications, ArXiv e-prints, 2023), we introduce nonparametric estimators for certain functionals involving the trawl function of interest and a non-linear test function. We show that our estimators are consistent and provide functional central limit theorems.

Mikkel Bennedsen, Aarhus University: Composite Likelihood Estimation of Stationary Gaussian Processes With a View Toward Stochastic Volatility

We develop a framework for composite likelihood inference of parametric continuous-time stationary Gaussian processes. We derive the asymptotic theory of the associated maximum composite likelihood estimator. We implement our approach on a pair of models that has been proposed to describe the random log-spot variance of financial asset returns. A simulation study shows that it delivers good performance in these settings and improves upon a method-of-moments estimation. In an application, we inspect the dynamic of an intraday measure of spot variance computed with high-frequency data from the cryptocurrency market. The empirical evidence supports a mechanism, where the short- and long-term correlation structure of stochastic volatility are decoupled in order to capture its properties at different time scales.

This workshop is sponsored by CoRE and the DNRF Chair in Econometrics