Saved in:
Bibliographic Details
Main Authors: Cheng, Ziteng, Coache, Anthony, Jaimungal, Sebastian
Format: Preprint
Published: 2023
Subjects:
Online Access:https://arxiv.org/abs/2308.08427
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914218044817408
author Cheng, Ziteng
Coache, Anthony
Jaimungal, Sebastian
author_facet Cheng, Ziteng
Coache, Anthony
Jaimungal, Sebastian
contents We investigate a framework for robo-advisors to estimate non-expert clients' risk aversion using adaptive binary-choice questionnaires. We model risk aversion using cost functions and spectral risk measures in a static setting. We prove the finite-sample identifiability and, for properly designed questions, obtain a convergence rate of $\sqrt{N}$ up to a logarithmic factor, where $N$ is the number of questions. We introduce the notion of distinguishing power and demonstrate, through simulated experiments, that designing questions by maximizing distinguishing power achieves satisfactory accuracy in learning risk aversion with fewer than 50 questions. We also provide a preliminary investigation of an infinite-horizon setting with an additional discount factor for dynamic risk aversion, establishing qualitative identifiability in this case.
format Preprint
id arxiv_https___arxiv_org_abs_2308_08427
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Eliciting Risk Aversion with Inverse Reinforcement Learning via Interactive Questioning
Cheng, Ziteng
Coache, Anthony
Jaimungal, Sebastian
Machine Learning
We investigate a framework for robo-advisors to estimate non-expert clients' risk aversion using adaptive binary-choice questionnaires. We model risk aversion using cost functions and spectral risk measures in a static setting. We prove the finite-sample identifiability and, for properly designed questions, obtain a convergence rate of $\sqrt{N}$ up to a logarithmic factor, where $N$ is the number of questions. We introduce the notion of distinguishing power and demonstrate, through simulated experiments, that designing questions by maximizing distinguishing power achieves satisfactory accuracy in learning risk aversion with fewer than 50 questions. We also provide a preliminary investigation of an infinite-horizon setting with an additional discount factor for dynamic risk aversion, establishing qualitative identifiability in this case.
title Eliciting Risk Aversion with Inverse Reinforcement Learning via Interactive Questioning
topic Machine Learning
url https://arxiv.org/abs/2308.08427