An eighty-year record
Here is a fact that sounds like exaggeration and isn't: no Big Six Canadian bank has cut its common dividend since the 1940s.
Not in 2008, when American and European banks were slashing payouts, taking bailouts, and in some cases disappearing. Not in March 2020, when the world locked down and every bank on earth braced for a credit apocalypse. Not through any Canadian recession, oil crash, or housing scare in between. Royal Bank, TD, Scotiabank, BMO, CIBC and National Bank have paid, and mostly raised, through eight decades of everything the world could produce.
For an income investor, that's about as strong a track record as exists in any market. It's also, as we've just finished arguing about streaks, the kind of credential that invites exactly the wrong conclusion — "safe forever, stop checking." So the useful question isn't whether the record is impressive. It's what produces it, and what would actually have to break.
Why the record is structural, not lucky
Eighty years without a cut across six independent institutions isn't a run of good fortune. It's the output of a system built, deliberately, to produce that result.
The oligopoly is real and protected. Six banks hold the overwhelming majority of Canadian banking assets, in a market where foreign entry is heavily restricted and bank mergers are effectively blocked. That structure is unfashionable to defend and enormously profitable to own: competition is limited, pricing power is durable, and the resulting earnings are steadier than almost anything in a genuinely competitive industry. Boring, protected profits are the raw material of a reliable dividend.
The regulator is unusually strict, and it worked. Canada's banking supervisor holds capital and liquidity requirements above international minimums and has for decades — the reason Canadian banks entered 2008 overcapitalized while their peers were leveraged into the ground. Mortgage underwriting rules are tighter, most Canadian mortgages are full-recourse, and the riskiest slice of housing lending carries government-backed insurance. Regulation that looks like a drag in good years is precisely what keeps payouts intact in bad ones.
Dividend policy is institutionally conservative. The Big Six generally target payout ratios in the 40–50% range of earnings — roughly half of profits retained. That's not an accident of arithmetic; it's a stated discipline, and it means earnings can fall substantially before the dividend is even mathematically threatened. A bank earning half again as much as it pays out has room a 90%-payout REIT simply does not have.
Put those together and the eighty-year record stops looking like luck. It looks like an oligopoly with regulated capital buffers and a deliberately conservative payout, which is a machine designed to keep paying.
Where the real risks actually live
Structural strength is not immunity, and the honest risks aren't the ones that trend on social media.
The regulator can freeze raises — and did, recently. In March 2020, Canada's supervisor instructed banks to halt dividend increases and buybacks, preserving capital against a feared credit wave. That restriction stayed in place until November 2021. Note what happened and what didn't: no bank cut, and when the freeze lifted, several announced large catch-up increases. But for roughly eighteen months, holders of the country's most reliable dividend growers got no growth at all — by order. That's the realistic modern downside case: not a cut, a pause. Worth knowing if your plan assumes uninterrupted annual raises.
Credit is the actual cyclical lever. Bank earnings swing on loan losses, and provisions can jump hard and fast in a downturn. A severe housing correction with sustained high unemployment is the scenario that genuinely pressures earnings — and Canadian household debt levels are high by international standards, which is a real vulnerability rather than a talking point. The payout ratio is the thing to watch here: at 45% there's enormous slack, but a severe enough earnings drop pushes that ratio up quickly, and a ratio running persistently above the 50s is a bank operating without its usual cushion.
Rate cycles cut both ways. Higher rates widen lending margins, which is good, and simultaneously stress borrowers, which is not. The net effect depends on where in the cycle you are and how fast the move happened. It's rarely the thing that threatens a payout on its own, but it shapes the earnings the payout comes out of.
And the structure itself could change. Everything above rests on a protected oligopoly. Serious competitive opening, a regulatory regime shift, or a genuine technological displacement of retail banking would rewrite the premise — none of which appear imminent, all of which are the kind of thing eighty-year records fail to anticipate, because that's what long records are structurally blind to.
Reading a bank's payout properly
Here's where the standard toolkit breaks again, and it's worth stating plainly because it's the third sector in a row where it does.
Free cash flow — the workhorse of ordinary dividend analysis — is close to meaningless for a bank. For a normal company, operating cash minus capex approximates money available to shareholders. For a lender, "operating cash flow" swings wildly with changes in loans and deposits, which are the business itself rather than leftovers from it: a bank that writes a lot of profitable mortgages can show sharply negative cash flow doing so. Run a bank through a generic dividend-safety screen and you'll get numbers with decimal points that mean nothing at all — the same disease we described in REIT payout ratios and utility capital spending, third verse.
What matters for a bank instead:
- Payout ratio against earnings — the direct measure, with the 40–50% norm as the reference and the trend mattering more than the level.
- Capital adequacy, published as a CET1 ratio — the regulatory buffer standing between loan losses and the dividend. This is the single most bank-specific safety number, and it's disclosed every quarter.
- Return on equity — how profitably the bank uses shareholder capital, the earnings-quality measure that replaces free-cash-flow thinking here.
- Credit quality trends — provisions for credit losses and impaired loans, which lead earnings pressure rather than following it.
- Valuation via price-to-book, not the debt ratios you'd apply to a manufacturer, since a bank's balance sheet is its inventory.
That list is, more or less, the shape of our own scoring: DivSight grades banks on a separate branch, substituting return on equity where free-cash-flow coverage sits for other companies and price-to-book where debt-to-equity would go — versioned like every methodology decision we make. Without that branch, the country's most dependable dividend payers would score like distressed companies, which is the same wrong-lens failure this series keeps documenting: a confident number computed about the wrong question.
So — safe?
As safe as dividend income realistically gets in a public market, with two honest asterisks.
The asterisks: growth can be suspended by regulators in a crisis even when the payout survives, as 2020 demonstrated; and "safe" here means the dividend, not the share price, which can fall hard while every cheque still arrives on time. Bank shares dropped sharply in both 2008 and 2020. The dividends didn't miss.
What that record buys you isn't permission to stop looking — it's a high prior, and a genuinely well-founded one. Check the payout ratio and its direction, glance at capital ratios, watch credit provisions when the economy turns. The numbers are published quarterly and the checks take minutes, which is how we run them continuously across all six. Eighty years is the strongest evidence available that the machine works. It's still evidence about the machine, not a promise about next year — and the difference between those two is the entire discipline.