Finally, the timing of asset sales by fiscal quarter is examined. Prior research (see,
e.g.. Hand et al. 1990, 78) has shown that real earnings-smoothing activities (through insubstance defeasances) cluster in the fourth fiscal quarter. The evidence in table 3 on
asset sales (measured in dollar amounts) by fiscal quarter reveals a similar pattern. The
percentage of assets sold by the sample firms in each fiscal quarter increased monotonically
during the fiscal year. In addition, the difference between the median (mean)
percentage of asset sales for the fourth quarter and the median (mean) for any of the
other three fiscal quarters is statistically significant at the 1 percent level for one-tailed
tests. Furthermore, the sale of long-lived assets in the fourth fiscal quarter exceeds that
in the three other fiscal quarters combined.