Which U.S. Stocks Grew Dividends Fastest Over the Last Five Years?

Old Dominion Freight Line (ODFL) grew its regular dividend per share at 30.00% a year from 2020 through 2025 without a single annual reduction, the fastest in AlgovestIQ's study of regular dividend-per-share growth among U.S.-listed dividend payers.

But the more useful finding is about the ranking itself.

Many of the apparent leaders by raw arithmetic are not dividend growth stories at all. They are companies whose 2020 was abnormal — suspended payments during the pandemic, or a token dividend replaced by a real one — measured from a floor they had just fallen to.

Quick answer: Among currently listed U.S. corporations that never reduced their regular dividend between 2020 and 2025, Old Dominion Freight Line grew its dividend fastest at 30.00% a year, followed by EMCOR Group at 25.59% and Monolithic Power Systems at 25.55%. Special dividends are excluded, and companies whose 2020 was unrepresentative — whether from a pandemic suspension, a dividend initiation or a cut made before the window — are not ranked.

Key findings

• ·Old Dominion led at 30.00% a year, raising its regular dividend in all five years from $0.302 to $1.120 per share.
• ·Companies whose 2020 was unrepresentative were removed — a small share of those that otherwise qualified, but a substantial share of the raw top-ranked names, because the distortion concentrates exactly where a reader looks.
• ·Technology had the highest median dividend growth of any sector at 9.99% a year, against 3.64% for Communication Services.

Why a no-cut filter is not enough

The intuitive screen is: rank every company by five-year dividend growth, then remove the ones that cut.

That screen produces Cigna at the top, with a 172.77% annual growth rate.

Cigna's regular dividend went from $0.040 in 2020 to $6.040 in 2025. It never cut inside the window. Every year rose. It passes a no-cut filter cleanly.

What actually happened is that Cigna paid a token $0.01 quarterly dividend for years, then initiated a real dividend of $1.00 a quarter in 2021. That is a dividend initiation, not five years of dividend growth.

The same pattern, from the opposite direction, affects companies that suspended payments in 2020:

A no-cut rule cannot catch any of these, because the cut happened in the starting year. From 2020 onward the series only rises.

The signal that does catch them is the number of payments in the starting year compared with the company's own established schedule. Texas Roadhouse paid four times in 2019 and once in 2020. Coca-Cola paid four times every year. No share price is needed to see the difference.

How the study was constructed

The universe is U.S.-listed common stocks with dividend history, read from AlgovestIQ's warehouse. ETFs and funds are excluded — they are covered by a separate study of fund distribution growth, which measures a different thing.

To qualify, a company must have a complete five-year regular dividend history with a payment in both the first and last year of the window, no reduction in its regular dividend in any calendar year of that window, a representative starting year, and enough payment classification to separate regular dividends from special ones. A separate, higher bar applies to the income-relevant table: a current yield of at least 0.5%.

Four decisions shape that screen.

Special dividends are excluded. Costco's $10 special in 2020 and $15 special in 2023 are not dividend growth. Where a payment carries no classification on any record and is far larger than the company's typical regular payment, it is marked unresolved and excluded pending review rather than counted as ordinary.

Dividends are restated for splits. Neither the raw payment nor the vendor's pre-adjusted figure can be summed across a split. Every amount is recomputed from the raw payment and the exchange split register.

Ranking happens inside an entity type. REITs, business development companies and master limited partnerships are ranked separately from ordinary corporations. Comparing Realty Income's distribution policy with Microsoft's dividend policy is not informative, and BDC distributions are largely investment income rather than corporate profit.

The yield floor is a presentation rule, not a research rule. Companies below 0.5% current yield stay in the research cohort and in the full ranking; they are held out of the headline table only, so a token dividend cannot top a list about dividend income. Yield moves with price and is no part of the historical calculation.

The ranking: fastest five-year regular dividend growth, no cut

Growth is measured on split-adjusted regular dividend per share across completed calendar years. The two endpoint columns are what the growth rate is computed from, so any reader can check the arithmetic.

Research rank is computed across all research-eligible companies. Display eligibility does not renumber ranks — a company held out of the hub's income table for a low yield keeps its place here.

† carries a comparability flag: a single-year step large enough to suggest a policy reset rather than steady compounding. ‡ below 0.5% current yield, so it is ranked here but held out of the income-relevant table on the dividend hub.

Ares Management and Goldman Sachs share rank 8, as do NVIDIA and Zoetis at rank 14, because their growth rates are identical as displayed. Forcing one above the other would invent a distinction the data does not carry.

The absolute-change column is there for scale, not for ranking. A company going from $0.02 to $0.12 and one going from $2.00 to $4.00 both show high growth rates and are not the same proposition. Note also that absolute change does not solve the base-effect problem on its own: Cigna's $6.00 increase is the largest in the whole universe, and its growth rate is the most distorted.

Four of the top twenty pay less than 0.5% today. NVIDIA's dividend went from $0.016 to $0.040 per share — a real 20% compound rate on a dividend that is economically almost irrelevant. That is why the hub's income table applies a yield floor and this table does not: the growth rate is a fact worth ranking, and whether it represents income is a separate question the reader should be allowed to see.

The base can be depressed from before the window, too

The cadence test asks whether the starting year looks like the years around it. That is not sufficient on its own.

GE Aerospace's dividend went from $0.199 in 2020 to $1.440 in 2025, a rate of 48.51% — the highest in the universe. It never cut inside the window, it paid its full schedule in 2020, and 2020 matched 2019 almost exactly. Every test above passes.

The cut that set that floor happened in 2018, before the window opens. GE went from roughly $4.19 a share in 2017 to $0.05 a quarter by the end of 2018, held there through 2019 and 2020, then rebuilt. Measuring from 2020 captures the rebuilding, not five years of dividend policy.

So the study adds one more test: the starting year is compared with the median of the three years before the window. A company starting below half that level is not ranked.

The median matters rather than the peak. Those pre-window years contain large one-off payments that the vendor labels as ordinary — Keurig Dr Pepper's $103.75 merger payment in 2018 and Thomson Reuters' $4.64 return of capital carry routine frequency labels — and a peak-based comparison would read an entirely normal base as a 99% cut.

GE Aerospace is the clearest case, starting the window at roughly a tenth of its pre-window level. The rule leaves companies whose low base was simply their settled level — Cameco cut in 2018 but had paid a stable $0.06 for two years before 2020 and grew genuinely from there, so it stays ranked.

The threshold was set from the measured distribution rather than chosen to remove any particular company: qualifying companies cluster well above it, and the few that fall below sit far below, so a wide range of cutoffs selects the same names.

Growth and income are not the same thing

The highest-growth dividends are mostly not the highest-yielding ones. Old Dominion, Monolithic Power and Universal Display all pay well under 1.5% today. A reader screening for income and a reader screening for dividend growth are looking for different companies.

That is also why current yield, payout ratio and free-cash-flow coverage are shown on each company's own page as live values rather than frozen into this study. They move with price and with each quarterly filing; the growth rate here covers closed calendar years and does not change until 2026 completes.

Median dividend growth by sector

Across the qualifying cohort:

Technology's lead is partly a base effect at sector scale: technology companies started paying dividends more recently and from lower payouts. Energy and Communication Services have few enough qualifying dividend payers that their medians should be read as descriptive of those companies rather than of the sector.

REITs, ranked separately

REITs are ranked in their own cohort because their distributions are governed by a payout requirement rather than by discretionary dividend policy.

What this study does not claim

It is a screen of companies listed today, not a history of the market. Every company here survived to 2026. A company that paid a growing dividend in 2020 and was acquired in 2023 is absent. That is correct for a reader asking "which currently available stocks grew dividends fastest" and wrong for anyone asking what the 2020 dividend universe did. This study answers the first question only.

It ranks past dividend policy, not dividend safety. Nothing here forecasts whether a dividend continues. Current payout ratio and free-cash-flow coverage, shown on each company's page, speak to affordability today.

It is a frozen annual measurement, watched daily. The 2020–2025 window does not change until 2026 completes. But a company that reduces its dividend after publication has the "no cut" description withdrawn from its page within the week, rather than standing until the next annual rerun. FedEx, which appears in the full ranking, is currently under exactly that review.

Readers can examine any company's dividend record on its own page — ODFL, MPWR, GS — or start from the dividend research hub.