Properties of time-series estimates of degree of leverage measures

Empirical studies suggest that time-series regression estimates of the degrees of operating and financial leverage have a tendency to produce measures less than one. According to ex ante theory, these measures should be greater than one for firms operating above the breakeven point. There have also been suggestions that the biases in these estimates may be attributable to an underlying increase in unit sales. This work presents evidence that these counter-intuitive measures are produced by changes in the firm's operating parameters (unit price, variable cost, fixed cost and interest payments). It further suggests that attempts to control for the underlying change in unit sales substantially increase the volatility of predicted estimates.