Dividend Growth
Dividend growth investing selects companies that keep increasing their payout over time, rather than those paying the most right now. A lower yield growing steadily overtakes a higher static one, and the year in which it does so depends only on the two rates involved.
The ordinary version of income investing sorts by yield. This one sorts by the rate the payout is rising, which is a different question with a different answer — and the arithmetic that separates them is simple enough to do in one line.
How it works
You buy companies raising their dividend rather than those paying the most today. The starting yield is usually modest and the expectation is that it compounds.
The growth has to be funded by earnings. A company raising its payout faster than it raises profit is closing the gap between what it earns and what it distributes, and that has an end.
Which is why the payout ratio matters more here than the yield does. The ratio says how much room the increases have left.
A worked example
Take 100,000 in a holding yielding 2% and raising the payout 8% a year, against 100,000 in one yielding a static 5%.
Year one pays 2,000 against 5,000.
Year ten pays 3,998 against 5,000 — still behind.
Year thirteen pays 5,036 against 5,000, and from there the gap widens permanently.
Thirteen years is a long time to be behind, which is the honest cost of the approach and the reason it suits accumulation rather than someone who needs income now.
Yield on cost
Yield on cost divides today’s payout by what you originally paid. On the example above it reaches 5.04% in year thirteen while the current yield may still be near 2%.
It rises automatically and it is not a return. Any holding whose payout grows produces a rising yield on cost, including one that has performed badly, because the denominator never updates.
The decision-relevant number is the current yield — what the same money would buy today. Yield on cost describes a purchase made years ago and answers no question about whether to keep holding.
The record is not a promise
Long increase records are the usual screen — twenty-five or more consecutive years is a common threshold — and they are genuinely informative about management priorities.
They are also survivor-biased. The list contains the companies that managed it; the ones that started a streak and broke it are not on it, and there is no way to see how many there were.
And a streak creates its own pressure. A company that has raised its payout for decades has a strong incentive to keep doing so through a year when it should not, which is exactly when the payout ratio starts telling you more than the record does.
Concentration
Any income screen concentrates. Companies able to raise payouts for decades cluster into a few industries — consumer staples, industrials, utilities, healthcare — so a portfolio of thirty of them is usually a bet on a handful of sectors.
Which is the same warning the sector funds page makes, arriving through a different door.
The practical response is to check the sector weights before the holding count. Thirty companies across four industries is a portfolio; thirty across two is a position, and the screen that produced it will not say which one you have.
Costs and tax
A cut usually means selling and replacing. Every trade pays the bid-ask spread, and in a taxable account the sale is a disposal.
The payout is taxable annually in a taxable account, growing as the dividend grows. A 3.5% yield taxed at 35% is a 1.22% annual drag, which is covered on the dividend tax page.
The original data
Of the 24,971 unique videos in the finance search study this site keeps, 19 have a title about dividend growth, at a median of 4,380 views across 15 channels — and 42% of those titles are instruction-shaped. Dividend investing appears in 137 videos at 5,503 and reinvestment in 3 at 65,554.
19 videos at 4,380 against 137 at 5,503 for yield-based investing. The version that selects on growth has a seventh of the coverage of the version that selects on level, despite the arithmetic favoring it on any horizon past thirteen years.
The question: Yield on cost is 7%. Hold regardless?
The answer is that yield on cost is not a reason to hold anything. It is a fact about a purchase made years ago. The question is whether you would buy this company at today’s price for today’s yield — and if the answer is no, the 7% figure is describing your history rather than the holding’s future.
When it fails
The failure is a company defending a streak it can no longer afford, and the screen cannot see it coming. The payout ratio climbs year after year as increases outpace earnings, the record stays intact because the increases keep happening, and the holding stays on every dividend-growth list right up to the announcement. What broke was the earnings; what the screen measured was the payout, and the two came apart years before anything was visible.
The second failure is treating yield on cost as performance. It rises mechanically.
A third is buying the streak rather than the coverage. The ratio is the forward-looking number.
A fourth is ignoring the concentration. Long-streak companies cluster in a few sectors.
A fifth is expecting income now. The crossover takes over a decade at typical rates.
And a sixth is holding it in a taxable account by default. The tax grows as the dividend does.
Related
Dividend tax is the annual cost this strategy is charged. REITs are the structure obliged to distribute rather than choosing to. And index funds already hold every one of these companies. Dividend Aristocrats is the S&P 500 index built on the twenty-five-year rule, with its tracking fund’s record against the S&P 500.
The figure that gets misused here is yield on cost. It rises every year simply because the denominator is frozen at what you paid a decade ago, and it tells you nothing about whether the holding is worth keeping today. The number that matters is the yield on what it is worth now, which is what you would be buying at.
— Michael Whitman
This page is educational, not financial advice. Test every idea on your own charts before risking money.