The best argument against everything on this blog
Every post on this site operates on a shared assumption: that picking the right dividend-paying stock or fund is a skill worth developing, and that our job is to help you develop it. There's a body of finance theory that doesn't quarrel with which dividend payers you pick — it questions whether the whole category deserves the attention. Not "you bought the wrong yield," but "the yield was never the thing that mattered."
That argument deserves better than a footnote, and dismissing it would make everything else on this site harder to trust. So here it is, stated the way its proponents would state it — and then the honest accounting of where it's right, and where it isn't the whole story.
The homemade dividend
Start with the mechanical fact underneath the whole argument, because it isn't in dispute — you can watch it happen on any trading screen. When a company pays a dividend, its share price drops by very close to the amount of the payment on the ex-dividend date. A $50 stock paying a $1 dividend opens the next day around $49, all else equal. The company didn't create value by mailing you a cheque. It moved a dollar from the share-price column to the cash-in-your-account column, and the market repriced the shares to reflect that the company now holds a dollar less.
Run the arithmetic on both paths and they land in exactly the same place. Own 100 shares of that $50 stock — $5,000 total. Take the dividend: $100 cash arrives, and your 100 shares are now worth roughly $4,900. Total: $5,000. Instead, sell 2 shares at $50 the day before: $100 cash arrives, and your remaining 98 shares are worth $4,900. Total: $5,000. Identical outcome, to the penny, whether the company decided to pay you or you decided to pay yourself.
That's the core insight economists formalized more than sixty years ago, usually called dividend irrelevance and credited to Modigliani and Miller: in a frictionless market — no taxes, no trading costs, no one knowing more than anyone else — a company's choice to pay dividends instead of retaining the cash doesn't change how much wealth its shareholders hold. Any dividend you don't want, you can reinvest by buying more shares. Any income you want from a stock that pays nothing, you can manufacture by selling a few shares yourself — a "homemade dividend," sized to the exact amount you need, on the exact day you need it, rather than whatever a board happens to declare. If dividends genuinely created value from nothing, someone would have found the machine that prints free money. Nobody has.
Where the critique is completely right
Three parts of this hold up under real scrutiny, and they're worth conceding before making any case for the other side.
The yield doesn't buy you extra return. Sort stocks by dividend yield and you're not selecting for better businesses — you're mostly selecting for maturity. Companies that pay large, steady dividends tend to be older, slower-growing, and more capital-intensive, and those characteristics correlate with expected return through the market's well-documented value and quality dimensions, not through the dividend itself. Two companies with identical underlying economics, one paying a dividend and one retaining everything, should be expected to deliver the same total return before any yield-chasing enters the picture. The dividend is a symptom of the kind of company you bought. It isn't the cause of anything.
The concentration cost is real. Canadian dividend investing runs into this constantly: screen for high yield in this market and you land overwhelmingly in banks, energy, telecoms, and pipelines — a handful of sectors that dominate the TSX's dividend-paying universe and crowd out nearly everything else. A "diversified" Canadian dividend portfolio is frequently a concentrated bet on four industries wearing a diversification costume, and the critique is dead right that this is a cost you're paying, not a feature you're getting, unless you've chosen it with your eyes open.
The mental accounting is exactly what it looks like. Treating a dividend cheque as the safe money you're allowed to spend, while treating share-price appreciation as paper gains you shouldn't touch, has no basis in what your money actually is. A dollar of dividend and a dollar of unrealized gain spend identically at the grocery store. The distinction lives entirely in your head, and a genuinely rational, frictionless investor wouldn't draw it.
Where the critique misses something real
Four things, and together they explain why careful, non-foolish people keep choosing this approach despite a theory that says they shouldn't.
The stickiness is information, not an illusion. A company's dividend record is a costly signal — boards raise dividends only when they're confident the increase can be sustained, and they resist cutting because the market punishes it severely. That reluctance means a dividend stream tends to move more slowly and more deliberately than the share price, which reacts to every headline, rate decision, and mood swing within minutes. In principle this shouldn't matter — the share price already reflects everything, dividend included. In practice, a management team's willingness to keep paying is a genuinely different, and often calmer, read on the underlying business than whatever the market is doing to the stock this week.
Sequence-of-returns risk is a real cost, not a fully avoidable one. Someone drawing income by selling shares in a falling market sells more shares for the same dollar amount precisely when the portfolio can least afford it — a permanent impairment that a portfolio living on actual cash distributions doesn't automatically suffer, because resilient payers tend to keep paying through exactly the downturns that would force a total-return investor to sell low. The honest caveat: a sufficiently disciplined total-return investor holding a cash buffer sidesteps this by simply not selling in the bad years. But "sufficiently disciplined" is carrying a lot of weight in that sentence, and dividend income removes the need for the discipline instead of assuming it exists.
Which is really the same point as this one: the self-control value of "don't touch the principal" is real, even though it isn't rational in the strict sense above. A heuristic doesn't need to reflect the true fungibility of money to change behaviour for the better — the same way an envelope budget works on people who know, intellectually, that it's all one bank account. Investors who treat dividends as spendable and price appreciation as untouchable tend to panic-sell less than investors managing one undifferentiated pool. For most actual outcomes, that behavioural edge is worth more than the small efficiency lost to holding the "wrong" mental model.
And in Canada specifically, the tax argument doesn't transfer as cleanly as it does elsewhere. Eligible dividends from Canadian corporations get a gross-up and dividend tax credit that meaningfully softens the "dividends are tax-inefficient" critique for domestic payers held in a non-registered account — a wrinkle the withholding-tax post touched from the other direction, and one that deserves its own full treatment. The short version: a dollar of eligible Canadian dividend income and a dollar of capital gain aren't taxed nearly as differently as the generic version of this argument assumes.
So which is actually better?
Neither, in the sense the question usually intends. This isn't a returns competition with a declared winner. It's a choice between two ways of drawing cash from a portfolio, and the honest answer depends on what you're actually optimizing for.
If pure long-run wealth is the only goal, and you trust yourself to sell shares calmly into a 30% drawdown without flinching, the total-return case is close to airtight: broader diversification, full control over the timing of every taxable event, and no built-in bias toward whichever handful of sectors happen to pay the biggest dividends this decade. If predictable cash flow, the behavioural guardrail of an untouched principal, and a payment stream that carries real information about business health matter to you — which for a great many actual retirees, actual humans, they legitimately do — dividend-focused investing is a deliberate, defensible tool. Not a mistake. A different tool for a different job.
Here's the sting in the tail, and it's why this site exists rather than an argument against it: choosing the dividend path doesn't make the theory's critiques disappear. It just changes what you owe in response to them. If the yield itself buys you nothing extra, chasing a bigger one buys you nothing, faster. If concentration is a real cost, paying it by accident is worse than paying it on purpose. And if the entire case for this approach rests on the dividend actually being the calm, information-rich signal the theory says it shouldn't need to be — then whether that specific dividend is earned or manufactured, covered or merely scheduled, stops being a nice-to-know and becomes the load-bearing assumption underneath your whole strategy.
Pick the tool for honest reasons. Then use it properly — which is the rest of what we do here, on every payer we cover.