Dividend Growth Stocks: What Is Inside the Dividend Growth Theme

The theme applies one test: can this company keep raising its dividend for years, funded by the business rather than by borrowing? In practice that means a multi-year record of consistent increases, durable conversion of earnings into free cash flow, and a balance sheet strong enough to defend the payout through a cyclical trough.

The word doing the work is growth, and it is what separates this theme from every high-yield list. Those two screens do not merely emphasise different things, they frequently select opposite companies. Yield is a ratio whose denominator is the share price, so it rises mechanically when the shares fall. A stock appears at the top of a yield screen either because the payout is unusually large or because the market has already marked the equity down, and in practice the second explanation is common enough that the highest-yielding names are often the ones whose dividend is least safe. Screening for the direction of the payout rather than its current size selects for the opposite profile: a modest yield with a lot of room underneath it.

The second structural choice is that the theme spans funding mechanisms rather than sectors. A dividend is only as durable as whatever pays for it, so grouping by industry hides the thing that actually matters. Contracted industrial revenue, repeat consumer demand, surplus technology cash flow, and regulator-approved bank capital fail at different points in a cycle and for different reasons, and a roster spanning all four is not making one bet on one kind of durability. For the general idea, see thematic investing.

If a dividend growth screen is not looking at yield, what is it looking at? The first answer is the payout ratio: the share of earnings a company hands out as dividends. It is the closest thing the category has to a safety measure, because it says how much has to go wrong before the payout is at risk. A dividend consuming a small slice of earnings survives a bad year with no decision required. A dividend consuming nearly all of them is one disappointing quarter from a board discussion, and that is true no matter how long the increase streak has run.

The better version of the same test is free cash flow coverage: the dividend measured against operating cash flow after capital spending, rather than against reported earnings. Dividends are paid in cash, and earnings are an accounting figure that can diverge from cash for years, particularly at companies building factories, plants, or data centres. Free cash flow coverage is the harsher and more honest test, because a heavy investment programme drains it immediately while leaving the earnings-based payout ratio looking comfortable.

Neither number has a universal safe threshold. The right band depends on how predictable the underlying cash flow is, which is why a contracted industrial gas supplier can sustainably run a higher ratio than a steel producer, and why comparing a company to its own sector is more informative than comparing it to a rule of thumb. This is also why the roster below is organised by funding mechanism: the ratio only means something once you know what is behind it.

The most boring source of dividend safety is a customer who comes back on a schedule. Consumer businesses with pricing power and non-cyclical demand generate cash in most economic conditions, which is what lets management commit to raising a payout years in advance rather than reviewing it every quarter. This layer is where the theme's steadiest funding sits, and it is deliberately not the highest-yielding part of the roster, because a business with this quality of cash flow rarely trades at a price that produces a big yield.

How this layer relates to the rest. This layer is the theme's stability anchor. Its cash flow is the least sensitive to interest rates, commodity prices, and capital cycles, so it is the part of the roster least likely to face a payout decision in a bad year. Everything else in the theme carries some cyclicality that this layer does not.

The second way to make a dividend predictable is to sell under a contract that runs for years. Industrial gas is the clearest example in listed markets: a supplier builds a plant next to a customer's facility and signs a long-term agreement with minimum-volume terms, so revenue arrives whether or not the customer runs at full rate. That structure converts an industrial business into something closer to an infrastructure annuity, and an annuity is the easiest thing in the world to pay a rising dividend out of.

How this layer relates to the rest. This layer supplies the visibility the cyclical names below cannot. Its revenue is contracted years ahead rather than set by this year's demand, so it is the part of the roster where the next several dividend increases are the most foreseeable. The trade-off is that its growth is tied to a heavy capital programme, which the consumer layer does not carry.

This layer is the most instructive in the theme, because the businesses in it are genuinely cyclical and the dividends have grown anyway. That is only possible for one reason: the payout was set low enough relative to mid-cycle earnings that a trough year still covers it. It is the payout ratio doing the work, not the stability of the industry, and it is the single best argument for why the ratio matters more than the yield when you are judging whether a dividend will survive.

How this layer relates to the rest. This layer is where the theme's central safety test is visible in the open. The consumer and contracted layers make a dividend look safe by making the revenue steady; these two make it safe by keeping the claim on earnings small. Both routes work, and holding both means the roster is not implicitly betting that steady industries are the only safe ones.

A company starts paying a dividend when it can no longer reinvest all of its cash at an attractive rate, which is why a technology dividend used to be treated as an admission of slowing growth. That framing is now out of date. Several large technology businesses generate more free cash flow than their own investment plans can absorb, and they pay some of it out while still growing. These names typically carry the lowest yields in the theme and the fastest increase rates, which is the dividend growth trade in its purest form.

How this layer relates to the rest. This layer supplies the growth rate. The consumer and industrial layers offer durability at a modest pace of increase; these names start from a smaller yield and raise it faster, so they are the reason the theme is about the direction of the payout rather than its current size. They are also the most exposed to a capital-spending cycle that could absorb the cash that currently funds increases.

Banks are the one place in the theme where the dividend is not purely a management decision. Large US banks size their capital returns against annual supervisory stress tests, so the payout is gated by a regulator's view of how the balance sheet performs in a severe hypothetical downturn. That makes bank dividend growth a different animal: the constraint is capital adequacy rather than earnings, and it is the reason a bank can be highly profitable and still not raise the payout in a given year.

How this layer relates to the rest. This layer diversifies the failure mode. Everything above it fails on operating cash flow. A bank dividend fails on capital and credit, which arrives at a different point in the cycle, so including a bank means the roster is not exposed to a single mechanism by which a dividend gets cut.

Read across the roster, the theme is a set of different answers to one question: what pays for the dividend, and what would have to happen for that source to stop? The consumer layer answers with demand that repeats. The contracted industrial layer answers with agreements signed years in advance. The cyclical industrial layer answers with a payout deliberately kept small enough that a trough year still covers it. The technology layer answers with cash generation that outruns the company's own reinvestment plans. The financial layer answers with a capital position a supervisor has already stress-tested.

Those five sources do not fail together. A recession pressures consumer demand and bank credit while the industrial gas contracts keep paying. An industrial downturn hits Linde, Air Products, Nucor, and Deere while leaving software and payments cash flow largely intact. A capital-spending cycle in semiconductors or cloud infrastructure absorbs the free cash flow that funds technology increases without touching a steel producer's payout at all. Holding them together is what makes the theme something other than a concentrated bet on one kind of durability.

It also means the ten names will not move as a block. Nucor trades on steel prices, Deere on farm income, Bank of America on rates and credit, Microsoft and Broadcom on an AI capital cycle. What they share is not a shared driver but a shared discipline, and that is a much weaker form of correlation than most thematic rosters carry, which is arguably the point of an income-and-stability core rather than a tactical tilt.

The most common misunderstanding in this whole category deserves stating plainly. A dividend is not a bonus paid on top of your investment, it is a transfer of value from the share price into your pocket. On the ex-dividend date the shares typically fall by roughly the amount of the payout, because the company is worth exactly that much less cash than it was the day before. Nothing has been created. Money has moved from one place you owned it to another.

Two consequences follow. The first is that yield is not return. What you earn is total return, price change plus dividends together, and a portfolio can pay a generous income while losing value. The second is tax: in a taxable account dividends are generally taxable in the year you receive them, whether you spend the cash or reinvest it, so an income-heavy holding creates a yearly tax drag that a non-payer does not. That is why many people hold income-focused positions inside a Roth IRA or another tax-advantaged account. Tax rules depend on your situation, so confirm the current details or ask a tax professional.

This is precisely why the theme is built around the quality and durability of the underlying businesses rather than around harvesting income. If the payout were free money, the sensible move would be to maximise it, and the highest yields would win. Because it is a transfer, the only thing that makes a dividend worth having over decades is a business healthy enough to keep funding it while still growing, which is the same thing as saying the business is a good one. The dividend is evidence, not the product.

A membership test is only credible if it excludes things. These are the names and categories people most often expect to find in a dividend theme, and the specific reason each one does not qualify here.

• The highest-yielding names on any screen. Yield is a ratio, and its denominator is the share price, so a yield rises when the price falls. The names at the top of a yield screen are frequently there because the market has already marked the stock down in anticipation of a cut, which makes the highest yield an unusually poor filter for the thing this theme is trying to select for. A dividend growth screen looks for the opposite profile: a modest yield with plenty of room underneath it.
• Broad income payers such as the staples and healthcare dividend stalwarts. Companies like the long-streak consumer staples and large pharmaceutical payers are excellent dividend stocks and are held in the broader dividend theme instead, where current income across sectors is the point. They are excluded here because their increases tend to run at or near the rate of inflation, so they express income durability rather than payout growth.
• REITs. Real estate investment trusts are required to distribute most of their taxable income, so a high payout is a structural feature of the entity rather than evidence of a management choice about capital. They are also valued against bond yields and financed with debt, which makes them rate-sensitive in a way the rest of this roster is not, and they belong in the REIT theme.
• Regulated utilities. A utility earns an allowed return on the capital it invests in the network, which produces a steady and generous dividend but very little payout growth, and the growth it does have is funded by issuing debt and equity to expand the rate base. That is a different mechanism from earning surplus cash and choosing to pay some of it out, so utilities sit in their own theme.
• Monthly-dividend payers. Paying twelve times a year instead of four is a cash-flow convenience, not a signal about the business, and monthly payers are concentrated in real estate and credit vehicles where the distribution is mandated or leveraged. Cadence tells you nothing about whether the payout can rise for a decade, which is the only question this theme asks.

Three of those exclusions have a home elsewhere on Walnut, and the boundaries are worth being explicit about. The broad income payers sit in the dividend theme, which selects for established companies paying a well-covered dividend across sectors and is weighted toward current income rather than the rate of increase. Property companies sit in the REIT theme, where the distribution is a requirement of the structure and the shares trade against bond yields. Regulated electric, gas, and water companies sit in the utility theme, where the payout is funded by an allowed return on invested capital and expansion is financed rather than earned. All three are reasonable ways to own dividends. None of them is a growing payout funded by surplus cash, which is the only thing this theme is testing for.

The same ten names, grouped by what funds the dividend rather than ranked by yield, so the shape of the theme is visible in one view.

Five funding mechanisms across 10 companies, with no single sector holding more than three places. That spread is the theme's central design decision, not an accident of which large caps happen to pay a dividend.

The passive route is a dividend fund, and it answers a related but different question. The theme names two proxies. SCHD applies a quality screen that combines dividend history with balance sheet and profitability measures, which is conceptually the closest fund to what this theme does. VYM takes a much broader approach, selecting above-median-yield US companies without a quality filter, so it holds far more names and tilts toward current income rather than growth in the payout. Both apply their rule mechanically and hand you the average result at weights you do not choose.

A theme inverts the trade. You know exactly which ten companies you hold, which funding mechanism each one represents, and what weight each carries, and you accept that ten names is a much narrower roster than a fund holds. Neither is automatically better. The fund is the simpler instrument and diversifies a dividend cut across hundreds of positions, the theme is the more deliberate one and lets you concentrate in the specific compounders you find most durable. Plenty of people hold a broad dividend fund as a core with a focused roster beside it.

A list of ten names is an input, not a portfolio. What turns one into the other is structure: which funding mechanisms you want exposure to, what weight each name carries, and whether the income and the concentration you end up with were chosen or inherited.

• Decide the funding mix first, then the names. The split between contracted cash flow, consumer demand, and surplus technology cash changes the character of the position far more than swapping one industrial gas supplier for another.
• Be honest about what you want the dividend for. A roster weighted toward the technology layer produces very little income today and a faster-rising payout later. One weighted toward the industrials produces more now and grows more slowly. Both are defensible; drifting between them by accident is not.
• Set target weights that sum to 100. Equal weighting across ten names is a choice, and so is tilting toward the longest records. Not deciding is what leaves you concentrated in whichever name ran hardest.
• Mind the account. In a taxable account dividends are generally taxed in the year they are paid, reinvested or not, so where an income-tilted portfolio lives affects your after-tax result more than a fraction of a percent of yield does.
• Judge it on total return. Dividends plus price change is the scorecard. A rising income stream attached to a shrinking capital base is not a good outcome, and only the combined figure shows that.

This is what Walnut is built for. You describe the thesis, the AI assistant proposes constituents and weights you can edit, the portfolio tracks as one performance line against the S&P 500, and you place trades you approve yourself at your own broker. Walnut is informational and does not tell you which stocks to buy.

For the companion view of which dividend growers are most widely held and discussed, see best dividend-growth stocks. For the income-now side of the same decision, see best high dividend stocks.

The dividend growth theme is ten companies across five funding mechanisms, and the mechanism is the whole idea. Costco and TJX are paid for by demand that repeats. Linde and Air Products by contracts signed years ahead. Nucor and Deere by a payout kept small enough that a cyclical trough still covers it. Microsoft, Broadcom, and Texas Instruments by cash generation that outruns their own reinvestment. Bank of America by a capital position a regulator tests every year.

Read as a yield list, the roster looks unimpressive, because selecting for a growing payout usually means accepting a small one today. Read as five different answers to the question of what makes a dividend survive, it is a structure, and the structure is what you are deciding whether to own. A high yield is a warning at least as often as it is an opportunity, a dividend is a transfer rather than a gift, and the only durable version of this strategy is owning good businesses. Nothing here is a recommendation, and Walnut is not an investment adviser.