Deep Reinforcement Learning for Asset Allocation in US Equities

Home / Resources / Podcasts / Deep Reinforcement Learning for Asset Allocation in US Equities

Deep Reinforcement Learning for Asset Allocation in US Equities

Dan Tudball and Sonam Srivastava discuss deep reinforcement learning for US equity asset allocation.

Thursday 19 May 2022

0:00 / 0:00

Podcast Description

QuantSpeak host, Dan Tudball, is joined by Sonam Srivastava, Founder of Wright Research, to discuss the application of reinforcement learning within asset allocation, the results of her recent research, and her career journey as a quant.

Topics & Timestamps

[00:00 - 01:12] Guest Introduction and Context

[01:12 - 01:34] Opening Question: AI/ML Priorities in Trading Strategy Design

[01:34 - 06:47] Key Guidelines for AI/ML in Trading: Information Leakage, Survivorship Bias, Data Snooping, Model Parsimony, Methodology Choice

[06:47 - 09:11] Production Process: Clean Pipeline, Assumptions, Validation, Recalibration

[09:11 - 10:03] Overview: Hidden Markov Models & Reinforcement Learning for Smarter Trading

[10:03 - 15:19] Comparing AI/ML Methods with Traditional Trading Strategies

[15:19 - 16:34] Career Path: Quantitative Modeling, Finance, and Interdisciplinary Work

[16:34 - 18:33] Career Start: Entering Finance Post-Great Recession

[18:33 - 22:17] Adoption of Machine Learning in Finance: Skepticism, Evolution, and Advantages

[22:17 - 24:41] Future Research: Fine-Tuning, Policy Learning, Generative Models, Data Scarcity

[24:41 - 25:59] Generative Models for Scenario Analysis in Finance

[25:59 - 26:37] Closing.

Disclaimer

Podcasts are for informational purposes only and provided “as is” without any representation or warranty from Fitch Learning of any kind. Comments or statements expressed by speakers may not be those of the Fitch Learning. Fitch Learning is not providing advice or recommendations. Fitch Learning, its directors, officers, or employees do not accept any liability for any loss arising from the use of information.