« Back to Results

High-Dimensional Methods with Applications to Forecasting and Policy Evaluation

Paper Session

Saturday, Jan. 4, 2025 2:30 PM - 4:30 PM (PST)

Hilton San Francisco Union Square, Union Square 11
Hosted By: Econometric Society
  • Chair: Jann Lorenz Spiess, Stanford University

Machine Learning Who to Nudge: Causal vs Predictive Targeting in a Field Experiment on Student Financial Aid Renewal

Susan Athey
,
Stanford University
Niall Keleher
,
Innovations for Poverty Action
Jann Lorenz Spiess
,
Stanford University

Abstract

In many settings, interventions may be more effective for some individuals than others, so that targeting interventions may be beneficial. We analyze the value of targeting in the context of a large-scale field experiment with over 53,000 college students, where the goal was to use "nudges"" to encourage students to renew their financial-aid applications before a non-binding deadline. We begin with baseline approaches to targeting. First

Can Machines Learn Weak Signals?

Zhouyu Shen
,
University of Chicago
Dacheng Xiu
,
University of Chicago

Abstract

In high-dimensional regression scenarios with low signal-to-noise ratios, we assess the predictive performance of several prevalent machine learning algorithms. Theoretical insights show Ridge regression’s superiority in exploiting weak signals, surpassing a zero benchmark. In contrast, Lasso fails to exceed this baseline, indicating its learning limitations. Simulations reveal that Random Forest generally outperforms Gradient Boosted Regression Trees when signals are weak. Moreover, Neural Networks with l2-regularization excel in capturing nonlinear functions of weak signals. Our empirical analysis across six economic datasets suggests that the weakness of signals, not necessarily the absence of sparsity, may be Lasso’s major limitation in economic predictions.

Forecasting GDP Growth Rate: A Large Panel Micro-Level Data Approach

Yongmiao Hong
,
University of Chinese Academy of Sciences
Naijing Huang
,
Central University of Finance and Economics
Yicheng Wang
,
Peking University
Zixuan Zhao
,
Central University of Finance and Economics

Abstract

Economists and econometricians typically use aggregate macroeconomic and financial data for inflation prediction. However, aggregation often results in a loss of valuable information, diminishing key features like heterogeneity, interactions, nonlinearity, and structural breaks. We propose a novel microeconometric approach to inflation forecasting, making use of a large panel of individual stock prices. By employing machine learning algorithms, we can effectively exploit this micro-level information to achieve substantially more accurate inflation forecasts. Our findings highlight the advantages and potential of utilizing micro-level data for macro prediction, diverging from conventional macro-forecasting approaches that rely on aggregate data to forecast macro variables.

Inference for CP Tensor Factor Model

Bin Chen
,
University of Rochester
Yuefeng Han
,
University of Notre Dame
Qiyang Yu
,
University of Rochester

Abstract

High-dimensional tensor-valued data have recently gained attention from researchers in economics and finance. We consider the estimation and inference of high-dimensional tensor factor models, where each dimension of the tensor diverges. Specifically, our focus lies on the factor model that admits CP-type tensor decomposition, allowing for loading vectors that may not be orthogonal. Based on the contemporary covariance matrix, we propose an iterative simultaneous projection estimation method. Our estimator exhibits robustness to weak dependence among factors and weak correlation across different dimensions in the idiosyncratic shocks. We establish an inferential theory, demonstrating consistency and asymptotic normality under relaxed assumptions. Within a unified framework, we consider two tests for the number of factors in a tensor factor model and justify their consistency. Through a simulation study and two empirical applications featuring sorted portfolios and international trade flows, we illustrate the advantages of our proposed estimator over existing methodologies in the literature.
JEL Classifications
  • C45 - Neural Networks and Related Topics
  • C53 - Forecasting and Prediction Methods; Simulation Methods