Aggregate Cost Stickiness in GAAP Financial Statements and Future Unemployment Rate

Abstract

ABSTRACT We examine whether aggregate cost stickiness predicts future macro-level unemployment rate. We incorporate aggregate cost stickiness into three different classes of forecasting models studied in prior literature, and demonstrate an improvement in forecasting performance for all three models. For example, when adding cost stickiness to an OLS regression that includes a battery of macroeconomic indicators and control variables, we find that a one-standard-deviation-higher cost stickiness in recent quarters is followed by a 0.23 to 0.26 percentage point lower unemployment rate in the current and following quarter. In out-of-sample tests, we find significant reductions in the root mean squared errors upon incorporation of cost stickiness for all three models. Additional tests suggest that professional macro forecasters, particularly those employed in nonfinancial industries, do not fully incorporate the information contained in cost stickiness. Finally, we find a stronger predictive power of cost stickiness toward the end of recessionary periods; we also assess cross-sectional variation of this predictive ability. JEL Classifications: M41; E24; J60.