Utility Dividend Policy And The Partial Adjustment Model

The dividend policy of public utilities and the adjustment process in moving to a desired dividend payout ratio are examined. The Bowers, Moran, and Forbes methodology (1987) is applied to earnings and dividend data from public utilities to test if the partial-adjustment model is an accurate characterization of utility dividend policy formulation. It is shown that 2 potential problems exist with the use of the partial-adjustment model. One problem is that the usual form of the regression analysis does not generate the appropriate t-statistics. The other problem is that autocorrelated earnings could result in an infinite number of pairs of speeds of adjustment and target payout rates giving identical dividend payout streams. Empirical tests with representative utilities show that, in many cases, the partial-adjustment model gives results that are not unique. It is probable that, for the utilities studied, the partial-adjustment model is inappropriate for the measurement of dividend policy of the individual utility.