H-model is an alternative to the two stage dividend discount model (DMM-2). Unlike the DMM-2 that assumes a constant initial high growth period followed by a sudden drop to the terminal growth rate, the H-model assumes that the dividend growth rate decays gradually over the next 2N years. This difference is illustrated in Figure 1. Figure 1: Dividend growth rates for two stage dividend discount model (DMM-2) [green line] and H-model dividend discount model (DMM-H) [red line]. While there is no exact formula for the H-model, on excellent approximation to for NPV is In Equation (1), D_{0} is the current dividend, r is the discount rate (the desired rate of return for the investment), g_{1} is the initial high growth rate, g_{2} is the terminal growth rate and N is the number of years where the growth has slowed to half way between g_{1} and g_{2} (see Figure 1). As an example DMM-H, we will analyze the MDT and INTC that were our test case for the DMM-2. The results are shown in Table 1 with the DMM-2 results included for comparison. As Table 1 shows, the DMM-H gives slightly lower valuation for MDT that has very high initial dividend growth rate g_{1}; however, the valuations for INTC come about equal. So which valuation model is the right one? DMM-2 assumes a constant growth with a sudden drop which seldom happens in practice but it is not unheard of either: A company, for example, might change its dividend policy with a stated goal of reaching a certain payout ratio in X number of years. In this case, DMM-2 is a realistic model. In general, DMM-H that assumes a slow decay in the growth rate is more close to actual behavior for dividend increases. Both models, however, should be used with skepticism as no-one can reliably predict dividend growth for more than a few years into the future. The safest approach is to make conservative estimates for both models and use to lower valuation. Table 1. H-model dividend
discount model (DMM-H) compared to the two stage dividend discount model (DMM2)
for MDT and INTZ.
You can modify this Google spreadsheet for your own analysis. |