Historically Erratic R&D Spending and Contemporaneous R&D Management

• In Event: Earnings Management and Market Reaction

Abstract

Earnings management is effective to the extent it is not detected. We argue that the detection of R&D manipulation is particularly difficult for firms with volatile R&D spending in the past. This is because investors are likely to factor the historically erratic spending into expectation and less likely to attribute current R&D surprises to earnings management. We show that this difficulty of detection increases firms’ likelihood of achieving earnings objectives through R&D cuts. We further show that the difficulty of detection is a particularly strong predictor of R&D manipulation within firms with high non-transient institutional holdings, and for firms with limited accruals flexibility.

Authors

• Crystal Xu, Rutgers, The State University of New Jersey, Newark

• Meng Yan, Fordham University