The dividend discount model has been criticised because of its simplistic assumption about a constant rate of growth. As such, it is unable to cope with situations of no growth, delayed growth and variable growth. How would you respond to this criticism?
Most valuation models have a sound theoretical base. However, as valuation models require estimates of future events these are subject to the problems of forecasting. Hence, necessarily there is subjectivity in the practical implementation of valuation models. That is, the problem is not the model itself, but rather obtaining the necessary values of the input parameters to implement the model. The present value model requires estimates of future cash flows and assumptions can be made to simplify the forecasting task. However, the assumptions themselves usually require further subjective input such as a growth rate in the case of the dividend discount model. Moreover, any model which is based on the present value concept requires a discount rate which incorporates the time value of money and relative risk. There are many different approaches to obtaining values for the input parameters, especially the growth rate of cost of capital. Most approaches seek to utilise available data to provide some objectivity to the task. Examples include trend lines based on past data, time-series models such as martingales, expert forecasts, formal models (e.g. the CAPM), inverting of valuation models where the price is known and the use of other financial data (e.g. the plowback technique). No one approach will always provide the most accurate estimate and hence each technique results in estimation, and ultimately, valuation errors. The extent of the error depends upon individual circumstances prevailing at the time. However, it is not uncommon for small input errors to result in large valuation errors, especially when the values of the cost of capital and growth rate are close to each other.