When should I use a three-stage DDM instead of a two-stage and how is the middle stage modeled?

Question

AcadiFi · Accepted Answer

The three-stage DDM addresses the unrealistic jump from high growth straight to stable growth in the two-stage model. The middle ('transition' or 'declining') stage linearly tapers the growth rate from g_high to g_stable over T periods, better reflecting how firms mature as markets saturate, margins compress, and reinvestment opportunities diminish. Structure: - Stage 1: years 1 to N₁, dividends grow at constant g_high - Stage 2: years N₁+1 to N₂, growth declines linearly each year - Stage 3: years N₂+1 onward, Gordon with g_stable The linear decline each period: g_t = g_high − (t − N₁)/(N₂ − N₁) × (g_high − g_stable). So if g_high=20%, g_stable=4%, and stage 2 runs 6 years, growth falls ~2.67% per year through stage 2. ```mermaid flowchart LR A[Stage 1
High Growth
20% for 5yr] --> B[Stage 2
Declining
20 to 4% over 6yr] B --> C[Stage 3
Stable
4% forever] ``` When to use: (1) clearly profitable growth firms with identifiable deceleration runway (SaaS companies hitting market saturation, medical device makers with patent cliffs), (2) cyclical industries mid-cycle, (3) emerging market champions expected to converge to developed-market growth rates. Empirically, the three-stage model reduces the terminal-value's dominance and produces more defensible fair values — but the cost is more assumptions (N₁, N₂, slope). Sensitivity-test each parameter; a reviewer will ask why stage 2 is 6 years and not 4 or 10. Practical tip: the H-model is a closed-form approximation: V₀ ≈ [D₀ × (1+g_L) + D₀ × H × (g_S − g_L)] / (r − g_L), where H = half the duration of the declining period. It gives a quick sanity check against the full discounting.

When should I use a three-stage DDM instead of a two-stage and how is the middle stage modeled?

Master Level II with our CFA Course

Related Questions

Practice Questions