A paper by Henry Lahr, Research Associate at Centre for Business Research (CBR), has been published in the European Accounting Review.
The paper, “An improved test for earnings management using kernel density estimation”, describes improvements on methods developed by Burgstahler and Dichev (1997) and Bollen and Pool (2009) to test for earnings management by identifying discontinuities in distributions of scaled earnings or earnings forecast errors.
The main advantage offered by the bootstrap procedure over prior methods is that it endogenises the bandwidth selection step for kernel density estimation. Results for scaled earnings and earnings forecast errors in US firms over the period 1976-2010 show that significance levels found in earlier studies are greatly reduced by the new test procedure, often to insignificant values. Discontinuities cannot be detected in analysts’ forecast errors, while such findings of discontinuities in earlier research can be explained by a simple rounding mechanism.