Title: Term structure of characteristic-sorted portfolios and multi-horizon investment
Authors: Serhiy Kozak - University of Maryland (United States)
Markus Pelger - Stanford University (United States)
Svetlana Bryzgalova - London Business School (United Kingdom) [presenting]
Abstract: The term structure of expected returns is explored across multiple horizons of portfolios sorted on contemporaneous characteristics. At the core of our approach is the tensor factor model --- a generalization of APT which explains returns on contemporaneously-sorted characteristics portfolios across many investment horizons. Equivalently, the model provides a parsimonious representation of contemporaneous returns sorted on multiple lags of characteristics. The model fits the data well and generates robust multi-horizon predictions and MVE portfolio weights for long-horizon investment.