Saved in:
Bibliographic Details
Main Authors: Feng, Guanhao, He, Jingyu, Polson, Nicholas G., Xu, Jianeng
Format: Preprint
Published: 2018
Subjects:
Online Access:https://arxiv.org/abs/1805.01104
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909422382481408
author Feng, Guanhao
He, Jingyu
Polson, Nicholas G.
Xu, Jianeng
author_facet Feng, Guanhao
He, Jingyu
Polson, Nicholas G.
Xu, Jianeng
contents This paper presents an augmented deep factor model that generates latent factors for cross-sectional asset pricing. The conventional security sorting on firm characteristics for constructing long-short factor portfolio weights is nonlinear modeling, while factors are treated as inputs in linear models. We provide a structural deep learning framework to generalize the complete mechanism for fitting cross-sectional returns by firm characteristics through generating risk factors -- hidden layers. Our model has an economic-guided objective function that minimizes aggregated realized pricing errors. Empirical results on high-dimensional characteristics demonstrate robust asset pricing performance and strong investment improvements by identifying important raw characteristic sources.
format Preprint
id arxiv_https___arxiv_org_abs_1805_01104
institution arXiv
publishDate 2018
record_format arxiv
spellingShingle Deep Learning in Characteristics-Sorted Factor Models
Feng, Guanhao
He, Jingyu
Polson, Nicholas G.
Xu, Jianeng
Methodology
This paper presents an augmented deep factor model that generates latent factors for cross-sectional asset pricing. The conventional security sorting on firm characteristics for constructing long-short factor portfolio weights is nonlinear modeling, while factors are treated as inputs in linear models. We provide a structural deep learning framework to generalize the complete mechanism for fitting cross-sectional returns by firm characteristics through generating risk factors -- hidden layers. Our model has an economic-guided objective function that minimizes aggregated realized pricing errors. Empirical results on high-dimensional characteristics demonstrate robust asset pricing performance and strong investment improvements by identifying important raw characteristic sources.
title Deep Learning in Characteristics-Sorted Factor Models
topic Methodology
url https://arxiv.org/abs/1805.01104