Improved Inference in Regression with Overlapping Observations
40 Pages Posted: 23 Jun 2004 Last revised: 15 Aug 2014
Date Written: March 3, 2010
Abstract
We present an improved method for inference in linear regressions with overlapping observations. By aggregating the matrix of explanatory variables in a simple way, our method transforms the original regression into an equivalent representation in which the dependent variables are non-overlapping. This transformation removes that part of the autocorrelation in the error terms which is induced by the overlapping scheme. Our method can easily be applied within standard software packages since conventional inference procedures (OLS-, White-, Newey-West- standard errors) are asymptotically valid when applied to the transformed regression. Through Monte Carlo analysis we show that it performs better in finite samples than the methods applied to the original regression that are in common usage. We illustrate the significance of our method with three empirical applications.
Keywords: Long horizon, stock return predictability, induced autocorrelation
JEL Classification: C20, G12
Suggested Citation: Suggested Citation
Do you have negative results from your research you’d like to share?
Recommended Papers
-
Consumption, Aggregate Wealth and Expected Stock Returns
By Martin Lettau and Sydney C. Ludvigson
-
Risks for the Long Run: A Potential Resolution of Asset Pricing Puzzles
By Ravi Bansal and Amir Yaron
-
Dividend Yields and Expected Stock Returns: Alternative Procedures for Interference and Measurement
-
Resurrecting the (C)Capm: A Cross-Sectional Test When Risk Premia are Time-Varying
By Martin Lettau and Sydney C. Ludvigson
-
Stock Return Predictability: Is it There?
By Geert Bekaert and Andrew Ang
-
Stock Return Predictability: Is it There?
By Geert Bekaert and Andrew Ang
-
Resurrecting the (C)Capm: A Cross-Sectional Test When Risk Premia Wre Time-Varying
By Martin Lettau and Sydney C. Ludvigson