Estimation of a Portfolio's Value-at-Risk Using Variational Auto-Encoders

Please enable javascript in your browser.

Moghimi, Mehrdad | 2021

790 Viewed

Type of Document: M.Sc. Thesis
Language: Farsi
Document No: 54313 (44)
University: Sharif University of Technology
Department: Management and Economics
Advisor(s): Arian, Hamidreza; Talebian, Masoud
Abstract:
One of the most crucial aspects of financial risk management is risk measurement. Advanced AI-based solutions can provide the proper tools for assessing global markets, given the complexity of the global economy and the violation of typical modeling assumptions. A new strategy for quantifying stock portfolio risk based on one of the machine learning models known as Variational Autoencoders is provided in this dissertation. The suggested method is a generative model that can learn the stocks' dependency structure without relying on assumptions about stock return covariance and produce various market scenarios using cross-sectional stock return data with a higher signal-to-noise ratio. We compare the proposed model's out-of-sample findings to those of twelve existing approaches, demonstrating that it is comparable with many of the well-known models for predicting the value at risk
Keywords:
Generative Models ; Variational Autoencoder ; Artificial Intelligence ; Machine Learning ; Dependence Structure ; Financial Risk Management ; Value at Risk ; Stock Portfolio ; Portfolio Generating Function