Dividend growth: the second number every income investor needs.

1. The CAGR formula

Compound annual growth rate from period 0 to period n:

CAGR = (D_n / D₀)^1/n − 1

Where D₀ is the dividend per share at the beginning of the period and D_n is the dividend per share at the end. Expressed as a decimal or percentage. The CAGR is the constant annual rate that, applied for n years, would take you from the starting dividend to the ending dividend. It is preferable to an arithmetic-mean growth rate because it accounts for compounding.

2. Worked example: 5-year and 10-year dividend CAGR

A US consumer-staples company paid a dividend of $1.20 in 2015, $1.36 in 2017, $1.55 in 2019, $1.78 in 2021, $2.04 in 2023, and $2.32 in 2025.

10-year CAGR (2015→2025): (2.32 / 1.20)^1/10 − 1 = 1.93^0.1 − 1 = 6.81 %.

5-year CAGR (2020→2025): assuming 2020 dividend was $1.66, (2.32 / 1.66)^1/5 − 1 = 6.92 %.

Both periods give a similar growth rate, suggesting a reasonably stable compounder. A widening gap between 5-year and 10-year CAGRs would suggest either acceleration (recent growth higher) or deceleration (recent growth lower) and is worth investigating.

3. The endpoint problem

CAGR is sensitive to the choice of start and end period. A 5-year window starting in a recession trough and ending in a peak overstates trend growth; the reverse understates it. To mitigate the endpoint distortion, two practices help: use longer windows (10 years or more, where available), and compare the trailing 5-year CAGR to the trailing 10-year CAGR. A divergence is a flag.

4. The years-to-double calculation

Given a dividend growth rate g, the years required to double dividend income is:

N = ln 2 / ln(1 + g)

For small g, this is well-approximated by the rule of 72: N ≈ 72 / g_in_percent. At 6 % growth, dividend income doubles in approximately 12 years. At 8 %, in 9 years. At 10 %, in 7.3 years. The compounding effect over a 30-year holding period at 6 % growth is a 5.7× multiplier on initial dividend income — the engine that drives long-horizon income-equity returns.

5. The growth-rate plausibility check

A trailing CAGR of 12 % looks attractive but cannot persist indefinitely — it would require either a perpetually rising payout ratio or perpetually rising earnings, neither of which holds in the limit. The plausibility check is the relationship between dividend growth, earnings growth, and payout ratio:

If the payout ratio is stable, dividend growth equals earnings growth. If dividend growth exceeds earnings growth for an extended period, the payout ratio is rising and will eventually hit a ceiling. A 12 % trailing dividend growth rate combined with 4 % earnings growth is unsustainable; the dividend growth will revert toward the earnings growth rate, with a lag. For projection purposes, use the earnings growth rate (or a discount to it) as the long-run dividend-growth assumption, not the trailing dividend CAGR.

6. The yield-plus-growth heuristic

Total return on a dividend-paying stock can be approximated as yield + growth + multiple expansion. Over long horizons, the multiple-expansion term tends to wash out, leaving yield + growth as the dominant component. A 3.5 % yield growing 7 % per year delivers approximately 10.5 % annualised total return; a 6.5 % yield growing 1 % per year delivers approximately 7.5 %. The high-yield-low-growth profile and the low-yield-high-growth profile can produce comparable total returns but very different income trajectories — the former pays you more cash now, the latter pays you more cash later.

7. Two-stage and multi-stage growth models

Real dividend trajectories are rarely a single rate. A young, fast-growing dividend-payer may grow dividends at 15 % per year for the first decade and then slow to 6 % as the business matures. The two-stage Gordon model handles this case by valuing the high-growth phase explicitly and the steady-state phase via the constant-growth formula. For most retail income-equity analysis, a single trailing-CAGR projection is adequate; the two-stage model becomes necessary when the projected trajectory is materially different from the trailing.

8. The dividend-growth investor’s typical mistake

The most common mistake among long-horizon dividend-growth investors is overpaying for the growth profile by accepting a starting yield below the rate of inflation. A 1.5 % yield is below most countries’ central-bank inflation targets; the dividend has to grow at well above inflation just to maintain real income. The combination of low starting yield and a multi-decade horizon also produces substantial sequence risk: if the growth slows in the first decade, the compounded income outcome is materially worse than the projection. A discipline of accepting a starting yield no lower than (current 10-year sovereign yield + 1 %) provides a useful floor.