What boring looks like on a thirty-year chart: the case for Canadian infrastructure compounding

The most boring chart in Canadian capital markets is the thirty-year total return on the senior Canadian infrastructure operators. The line goes up. The line goes up most years. The line continues going up through commodity stress, financial crises, and pandemics. The shape is unremarkable on a quarter chart. On a thirty-year chart it is the entire point.

A 6% earnings compounder, a 4% dividend yield, no glamour: that is the arithmetic of Canadian infrastructure. Compounded for thirty years, the math produces a 5.7x total return at constant valuation. Most of the desk's compounding work is done by names whose deck commentary will never appear at a conference panel and whose CEOs have not been the subject of an industry profile in five years. That is, in the strict sense, the design.

The cohort

The Canadian infrastructure cohort is small enough to name on one line: Fortis (FTS), Enbridge (ENB), TC Energy (TRP), Canadian National (CNR), Pembina, the Brookfield-controlled subsidiaries that trade as separate listings (BIP, BEP), and a handful of regulated utilities of more limited diversification.

The cohort sits at the boring end of the desk. The valuation multiples are unremarkable, the analyst coverage is dense, the deck commentary is interchangeable across the names. None of the operators is going to deliver an outsized single-year return. All of them have been delivering an unremarkable double-digit total return for three decades, in aggregate, with rolling dividends raised in essentially every year.

Fortis has raised the base dividend for fifty-two consecutive years. The streak is the longest in Canada. The shape of the streak — through 1981 inflation, the 1987 crash, the 1990 recession, the 1998 Asian crisis, the dot-com bust, the global financial crisis, the 2015 commodity reset, and the 2020 pandemic — is the part most readers find dull. It is dull because every year's raise looks like the previous year's. It is also, considered in aggregate, the cleanest single capital-allocation document in the Canadian listed universe.

The arithmetic of dull compounding

Take a Canadian infrastructure operator with the following profile, which is roughly the average of the cohort. Rate base grows at 6% per year. Authorized ROE on the rate base sits at 9.5%. Earnings per share grow at 6% per year. The dividend is set at 75% of earnings and grows in line. The dividend yield at the prevailing share price is 4%.

Total return at constant valuation: 6% earnings growth plus 4% dividend yield equals 10% annualized. Over thirty years, 10% compounds to 17.4x. Over twenty years, 6.7x. Over ten years, 2.6x. None of those numbers is dramatic on any single-year window. All of them are the consequence of the same dull arithmetic.

The valuation does not stay constant. Multiples expand and contract on a five- to seven-year cycle, and the buy-and-hold investor lives through the contractions as paper drawdowns. The 2015 pipeline contraction, the 2018 utility rate-cycle adjustment, the 2022 rate-rising-environment compression: each produced a 20–30% mark-to-market drawdown on the cohort. None changed the underlying compounding. The investor who held through the drawdowns received the math. The investor who sold did not.

What makes the cohort durable

Three structural features explain why the Canadian infrastructure cohort has produced this record where other ostensibly similar cohorts (US merchant power, certain European utilities, telecom) have not.

First, regulated rate base. Roughly 80–99% of cohort EBITDA is regulated or contracted under cost-of-service or take-or-pay structures. The earnings do not depend on the commodity tape, the freight tape, or the consumer tape. They depend on the rate base growing and the authorized return on the rate base holding. Both are functions of the regulatory regime and the operator's discipline in spending rate-base capital that earns its authorized return. The Canadian regulators — the CER, provincial commissions, the US FERC for cross-border assets — have produced a regulatory environment that, with some friction, allows the rate-base model to work.

Second, asset life. The pipelines, transmission lines, utility distribution systems, and rail track that the cohort owns are multi-decade assets. The Enbridge Mainline has been moving Canadian crude for sixty years. The Fortis distribution systems have been delivering electricity for longer. The CN main lines that connect Halifax to Prince Rupert have been in service for over a century. The asset life means the capital spent today is amortized over a long enough period that the unit cost stays low; it also means the regulatory accommodation is durable, because the regulator cannot reasonably ask the operator to stop providing a service whose alternative is more expensive.

Third, capital culture. The cohort is, in aggregate, conservatively financed, capital-disciplined, and shareholder-aligned in ways most other Canadian sectors are not. The 2015 and 2020 dividend records are public. Almost every cohort name raised the dividend in both windows. Almost no cohort name issued common equity at distressed prices. The culture is not accidental; the operators recruit and promote internally for cultural fit, and the boards on this cohort have been unusually stable.

A regulated utility is the boring half of the desk. The boring half is the part that pays for the wait.

Where the cohort breaks down

The cohort is not uniform. Three failure modes are worth naming.

Regulatory error. A regulatory environment that turns against the operator can structurally impair the rate-base model. The post-Wexit-rhetoric Alberta of 2019–2020 was a brief stress test on the Canadian pipeline cohort; the regulator did not change but the political backdrop did, and the cohort traded down. The deeper failure mode is a regulator that genuinely reduces authorized ROE meaningfully, which has happened in some US state utility cases. The Canadian cohort has not seen this at scale.

Capital indiscipline. TC Energy's pre-2024 capital program, anchored by Coastal GasLink, is the cleanest recent example of a Canadian infrastructure operator overspending on a signature project. The over-run was real, the consequences were a net debt elevation that took years to walk down, and the South Bow spinout in late 2024 was part of the cleanup. The operator is still in the cohort, but the post-2018 record is meaningfully weaker than the pre-2018 record.

Stranded-asset risk. A long-life regulated asset can become uneconomic if the underlying demand structurally disappears. The most-discussed risk on this side is Enbridge's liquids Mainline in an aggressive Canadian-oil-demand-decline scenario. The probability of the worst case is, in our read, low. The probability of a meaningful demand erosion over thirty years is non-trivial. The cohort's diversification into US gas distribution (Enbridge's 2023–2024 Dominion utilities acquisition) and renewable rate base is a partial hedge. It is not the same as the original toll-road economics.

Why the cohort is not bigger

A reasonable question, looking at the small cohort, is why the Canadian infrastructure compounders are not a larger group. The Canadian listed universe contains hundreds of names with adjacent characteristics — regulated utilities of various scale, REITs in the multi-residential and industrial categories, telecoms with substantial regulated assets, agricultural cooperatives with quasi-regulated returns. Why does the desk's compounding work concentrate on roughly half a dozen names?

The answer is in three criteria, each of which eliminates a meaningful share of the listed universe. The first criterion is asset durability. A regulated asset whose underlying technology can be displaced — a coal-fired power utility, a copper-wire telecom, a fossil-fuel pipeline whose feedstock declines over thirty years — does not produce the same compounding profile as an asset whose technology is structurally durable. The Canadian infrastructure cohort holds assets whose displacement risk over thirty years is meaningfully lower than the median of the listed regulated-asset universe. That filter eliminates many of the names a casual reader would expect to qualify.

The second criterion is regulatory durability. The regulator's willingness to grant a fair return over a multi-decade window is not constant across Canadian and US jurisdictions. Certain US state utility commissions have produced authorized ROEs that have ratcheted down meaningfully over the past two decades. Certain Canadian provincial commissions have produced regulatory friction that has impaired the operator's ability to spend rate-base capital cleanly. The cohort on the desk operates in the jurisdictions where the regulatory relationship has been most consistent — primarily the federal level (CER, FERC), the more institutionally stable provincial commissions (Ontario, British Columbia), and the US state commissions that have a track record of fair-return rulings.

The third criterion is capital culture. A regulated asset run by a poorly disciplined capital allocator is, in the long run, no better than a non-regulated asset run by a great one. The cohort on the desk meets a high bar on capital culture — measured by the 2015 and 2020 dividend records, the equity-issuance restraint, the buyback discipline, and the per-share growth profile. The bar excludes a number of names whose underlying assets are excellent but whose capital cultures have, at points, eroded the underlying advantage.

The compounded effect of the three filters is a cohort of roughly half a dozen names. That is not, in our view, a defect of the framework; it is the framework's expression. A research desk that produces a list of seventy-five compounders is a desk that has not done the work. The desk that produces six is the one that has applied the criteria. Six is enough.

Why the boring half pays for the wait

A research desk that earns its keep over decades does so by allowing the exciting parts of the portfolio to compound at high single digits while the boring parts compound at high single digits without the volatility. The shape of the total-return distribution matters. A 12% compounder with 30% annualized volatility is, in many practical investor situations, a worse outcome than a 9% compounder with 10% volatility, even though the long-run point estimate of the first is higher.

The Canadian infrastructure cohort is the low-volatility part of the desk. We do not own it for the headline return; we own it because the lower variance allows the higher-variance parts of the desk to be held through their drawdowns without forced selling. The compounding that the boring half produces over thirty years is the dividend that makes the impatient half tolerable.

That is, in five words, the case for Canadian infrastructure compounding. The math is unspectacular. The chart is unspectacular. The thirty-year aggregate is exactly the point.