Two streaks, two endings
Put the two stories side by side, because together they teach what neither teaches alone.
Fortis has raised its dividend every year for more than half a century — a streak that survived oil shocks, double-digit rates, 2008, and a pandemic, and one we recently argued deserves better measurement tools than it gets. Allied Properties, until this past December, carried its own admired record: over a decade of consecutive annual distribution increases, enough to sit among the Canadian dividend aristocrats, the kind of history income screeners are built to find. Then Allied cut its distribution by 60%, and the streak went from credential to trivia overnight.
Same credential. Opposite endings. Which raises the question this post is about: what does a raising streak actually prove? Because the honest answer is "something real, and less than you think" — and the gap between those two is where investors get hurt.
What a streak genuinely proves
Start with the real information content, because streaks are not noise.
A dividend raise is a public, hard-to-reverse promise, and cutting one is expensive in every currency a board cares about — the price drops, income holders leave, and the headline writes itself. Which means a long raising streak is what economists call a costly signal: cheap to claim, expensive to fake. A company doesn't stumble into forty consecutive raises. It gets there by organizing itself around not disappointing you — building payout policy into capital planning, choosing projects with the raise in mind, treating the streak as identity. Boards protect streaks the way airlines protect safety records, and for the same reason: the record is the brand.
So a streak is genuine evidence about two things: management's demonstrated priorities, and the historical stability of the business underneath — you cannot raise through four recessions on top of a fragile enterprise. That's real. It's why the streak screen finds Fortis, and why the aristocrats fund built an entire index on the idea.
Notice, though, what both pieces of evidence share: they're about the past. The streak proves the machine worked, and that management wanted it to keep working, under every condition the past happened to contain. It says nothing about whether the machine works now — and the conditions the past happened to contain is a quietly load-bearing clause.
What the streak can't see
Allied's streak was earned honestly. Urban office was a genuinely strong business for the entire era the streak was built in; the raises were covered when they were declared. The streak wasn't a lie about the past. It was a lagging indicator wearing a leading indicator's costume — and when the ground shifted, everything the credential couldn't see was exactly what mattered.
What it couldn't see: a payout ratio that had climbed to roughly all of available cash. Occupancy sliding as office demand failed to recover. Debt at a multiple of the REIT's market value with refinancing on the clock. Every one of these was public, current, and pointed the other way from the streak — the exact stack of warnings on the yield-trap checklist. An investor reading the coverage saw trouble years out. An investor reading the streak saw an aristocrat, right up until the press release.
The rule that falls out of the contrast is short enough to keep: a streak measures the past; coverage measures the present; when they disagree, coverage wins, and it isn't close. Fortis's streak deserves trust not because it's long, but because the machine currently underneath it — regulated returns, sanctioned capital plans, a payout sitting comfortably inside earnings — still works, checkably, today. The streak is the receipt. The coverage is the inventory. You can be robbed holding a very good receipt.
When the streak becomes the problem
There's a darker wrinkle worth naming: past a certain length, a streak stops being only evidence and starts being an incentive — and incentives cut both ways.
The mild version is the token raise: a company nudging its payout up a fraction of a percent, not because the business earned it but because the streak demands an entry every year. The raise is real, the signal is hollow, and screeners count it identically to a real one. The serious version is holding on too long: a board maintaining a payout the cash no longer covers, precisely because the streak has become the company's identity and breaking it feels existential. Cutting a year earlier is almost always cheaper for shareholders than cutting a year late — but the streak votes against it every time, right up until it can't. It's hard to look at Allied's final stretch — reassurances in August, a 60% cut by December — without seeing some of that gravity at work.
And one mechanical trap on top, flagged in our comparison of the dividend ETFs: the Canadian aristocrats index weights its members by yield. When a streak-holder gets into trouble, its price collapses, its yield spikes — and the index responds by weighting it more heavily, mechanically maximizing your exposure at the exact moment the streak is likeliest to break. The credential concentrates you into its own failures. Investors who held the aristocrats fund through certain famous Canadian cuts have lived this arithmetic.
How to actually use one
None of this argues for ignoring streaks. It argues for demoting them — from verdict to tiebreaker.
Used properly, a streak is the last check, not the first: run the coverage, the balance sheet, the sector-appropriate metrics — the right lens for the kind of business — and if the payout passes on today's numbers, a long streak is a genuine reason to prefer it over an equally-covered payer without one. It tells you management will fight for the thing you're buying. That's worth something, sometimes worth a premium. What it's never worth is overriding what the current numbers say — the mistake of reading the receipt as the inventory.
That demotion is how our own engine treats it, for what it's worth: streaks earn points as supporting evidence, and a confirmed cut truncates one on the spot rather than letting the ghost of old raises keep scoring — a lesson we learned in public, on Allied. The scoring weight lives with coverage, where the present tense is.
Fifty years of raises is a beautiful thing. Allied had more than a decade of them, and the next entry was a 60% cut announced four months after the payout was called comfortable. The streak was true the whole time. It just wasn't the part of the story that was about you — the current numbers were, and they always are.