Simultaneous Estimation of the Implied Values of Franked (Tax-Free) Dividends, Required Rates of Return and Growth Rates Using a Modified Residual Income Valuation Model

Abstract

Do investors price tax rebates on dividends? Since President George W. Bush's proposal to eliminate taxation of dividends at both the corporate and personal tax levels, there is a revival of interest on how dividend, corporate and personal taxes affect the pricing of equities. This study explores a setting, Australia, where corporate and personal tax systems are fully integrated in the form of an imputation tax system to examine the pricing of franked (tax-free) dividends using a valuation approach. A procedure is developed to simultaneously estimate the value of franked dividends, the cost of equity, and growth rates implied by share prices, book values, and analysts' earnings forecasts. The analysis is done on an after corporate tax but before personal tax basis by incorporating tax credits as part of payoffs. The estimation procedure employs a system of two equations that are equivalent price expressions based on different periods of earnings forecasts. Via simple manipulation and rearrangement, the system of equations leads to two (regression) relations with four estimates (two constant terms and two slope coefficients). The four estimates are functions of the implied value of franked dividends, the implied value of required rate of return, and the two implied growth rates based on different periods of earnings forecasts for a portfolio of firms. The estimation procedure is performed on a portfolio of firms followed by I/B/E/S from January 1st 1993 to December 31st 1999. The results show that investors do price the tax rebates with the estimates of a dollar of franked dividends being statistically greater than $1. In the estimation procedure, optimistic bias inherent in earnings forecasts would drive the estimate of market value of franked dividends upwards. Another estimation procedure is performed in a setting without imputation tax credits (i.e., the US) to examine the extent to which investors adjust one-year-ahead earnings downwards, knowing that these forecasts are, on average, optimistically biased. To the degree that forecast biases are comparable across the two countries (i.e., Australia and the US), the optimistic bias in earnings forecasts alone cannot fully explain the estimate of the value of franked dividends being greater than $1.