Publication Type
Working Paper
Version
publishedVersion
Publication Date
12-2015
Abstract
We use the adaptive LASSO from the statistical learning literature to identify economically connected industries in a general framework that accommodates complex industry interdependencies. Our results show that lagged returns of interdependent industries are significant predictors of individual industry returns, consistent with gradual information diffusion across industries. Using network analysis, we find that industries with the most extensive predictive power are key central nodes in the production network of the U.S. economy. Further linking cross-return predictability to the real economy, lagged employment growth for the interdependent industries predicts individual industry employment growth. We also compute out-of-sample industry return forecasts based on the lagged returns of interdependent industries and show that cross-industry return predictability is economically valuable: an industry-rotation portfolio that goes long (short) industries with the highest (lowest) forecasted returns exhibits limited exposures to common equity risk factors, delivers a substantial alpha of over 11% per annum, and performs very well during business-cycle recessions, especially the recent Great Recession.
Keywords
Complex industry interdependencies, Predictive regression, Adaptive LASSO, Central node; Industry-rotation portfolio, Business cycle, Multifactor model, Principal components, Target-relevant factors
Discipline
Business | Finance and Financial Management
Research Areas
Finance
Areas of Excellence
Finance and Financial Markets
Citation
RAPACH, David E.; STRAUSS, Jack; Tu, Jun; and ZHOU, Guofu.
Industry Interdependencies and Cross-Industry Return Predictability. (2015).
Available at: https://ink.library.smu.edu.sg/lkcsb_research/4515
Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-No Derivative Works 4.0 International License.
Additional URL
https://ssrn.com/abstract=2566541