The dividend that survived: 26 consecutive raises at Canadian Natural

Canadian Natural Resources has raised its base dividend in every year for twenty-six consecutive years, including 2015 and 2020. The raise announced in March 2020 — a six per cent increase — landed the same week the WTI front-month went briefly negative. The arithmetic of that raise is the cleanest single piece of evidence about Canadian energy capital allocation in the public record.

The record is not a dividend story. The dividend is a downstream effect. The record is a capital-allocation document about how a Canadian oil and gas business is built to survive the bottom of its own cycle.

The twenty-six raises, in three windows

The dividend has been raised in three distinct macro windows that tested whether the prior raises could continue.

The first window: 2009–2010, the global financial crisis. CNQ raised through the trough. The asset base in 2010 was meaningfully smaller, the share count meaningfully larger relative to today's; the dividend per share was a fraction of the current rate. The raise was modest in absolute terms. The discipline was set.

The second window: 2015–2016, the oil cost-curve reset. WTI averaged the high US$40s for two years; AECO traded persistently below US$2/GJ; Canadian heavy-light differentials widened to historic extremes. CNQ raised the dividend in both years and made no equity issuance. Capital expenditure was cut meaningfully. The 2016 sustaining capex per barrel printed in the low US$10s, which was the structural answer to the price-tape question. The cohort behaviour in the same window included multiple dividend cuts, several equity issuances at distressed prices, and the early stages of the restructurings that closed two years later.

The third window: 2020, the COVID demand collapse. The March 2020 raise was the signature event. The operator's commentary at the time was deliberate: the dividend was covered at the trough oil price, and the raise was the most public way to communicate that fact to a shareholder base that had been told by every other operator's announcement that the dividend was at risk. The arithmetic supported the commentary. The 2020 free cash flow at the low WTI tape was still positive on a maintenance basis. No equity was issued. Capital expenditure was cut hard.

The arithmetic that made the raises possible

The dividend record is downstream of three structural choices the operator has made over thirty years.

The asset mix. CNQ's production is predominantly long-life and low-decline. The Horizon and Albian mining and upgrading complex, the Pelican Lake and Primrose thermal projects, and the legacy heavy-oil and natural-gas assets together produce a base that declines at a low single-digit rate without sustaining capex. Compare this to a Permian-shale operator whose base production declines at 40% per year; the sustaining capex required to hold flat is a fundamentally different number. CNQ pays for low decline up front in the asset acquisition; the dividend in 2020 is the receipt.

The cost discipline. Sustaining capex per barrel printed in the low US$10s in 2015 and 2020. The 2025 number is slightly higher but remains first-quartile in the Canadian cohort. The discipline is operational: the operator's mining and SAGD sites have been in continuous improvement for two decades, and the unit cost numbers reflect that.

The capital allocation culture. Murray Edwards is the architect of the culture. He has been on the board for thirty years and a meaningful shareholder throughout. Tim McKay has been the senior operating presence for two decades. The continuity of the capital allocator is the part that does not show up in any operating metric but explains every decision that produced the record.

What 2020 actually communicated

The March 2020 raise was a signal, and the signal was about more than the dividend. The operator was saying, in public, that it did not need to participate in the cycle's worst quarter the way the rest of the cohort would. The hold positions through the trough were the part the operator could see in the asset base; the equity-issuance restraint was the part the operator was choosing; the dividend raise was the visible artifact of both.

The follow-on actions through 2020 and 2021 confirmed the signal. CNQ continued to buy back shares through the trough at prices well below the prior peak. The buybacks were modest in absolute size, by design — the operator's posture has been "do not embarrass the dividend by promising more than the buyback can deliver" — but the cumulative effect over the 2020–2025 period has been meaningful per-share reserve and production growth.

Suncor (SU) had a different 2020. The dividend was cut by 55%. The cut was rational; the conditions that forced it were the result of five years of incremental cost-discipline erosion at Fort Hills, the upgrader, and the safety record. The CEO change in April 2023 was the direct response. The post-Kruger reset is the recovery; the pre-Kruger drift is the cautionary tale that the same management team can produce one record in 2015 and a meaningfully worse one in 2020.

Enbridge (ENB) had a different 2020 again. The dividend was raised through both windows, with no equity issued at distressed levels, and the contractual structure of the pipeline business absorbed the macro shock without operational consequence. The CNQ record on the E&P side and the ENB record on the infrastructure side are the two cleanest dividend-and-discipline records in Canadian energy across both stress windows.

The dividend held through 2015 is a signal. The dividend held through 2020 is a signal. The dividend held through both is a different signal, and the operators who delivered it are a short list.

The compare to Berkshire's record

The Canadian Natural record draws a comparison most readers in Canadian capital markets do not make often enough. The cleanest parallel to Murray Edwards's capital allocation record at CNQ is, in our reading, Warren Buffett's record at Berkshire Hathaway in the operating period from 1985 to 2010. The structural similarities are not coincidental.

Both built capital cultures around a small number of repeatable rules. Buy assets that produce cash without requiring growth capital to do so. Hold assets through cycles. Treat the dividend, or in Berkshire's case the share-count, as the visible discipline. Do not issue equity at distressed prices. Buy back stock when the market gives you the opportunity to. Promote internally and prefer continuity over freshness in operational leadership. The two operators apply the rules in different sectors, but the rules themselves rhyme more than the conference-panel commentary tends to acknowledge.

The differences are also informative. CNQ pays a dividend; Berkshire does not, on the grounds that retained earnings compound faster inside Berkshire than they would in shareholder hands. The CNQ dividend is a function of the operator's cash structure: the assets generate so much free cash that, after sustaining capex and growth capex, there is more capital available than the operator can productively deploy. The dividend is the residual; the buyback is the calibrated lever. The Berkshire model is different because the holding-company structure allows for higher-return reinvestment opportunities than CNQ's asset base typically presents.

The most useful single sentence on the Edwards record is one Murray Edwards himself approximates in CNQ's letters: the dividend is the operator's promise, the buyback is the operator's option, and the equity issuance is the operator's failure. The 2015 and 2020 records demonstrate the first sentence in two different cycles; the post-2015 and post-2020 buyback records demonstrate the second sentence in two different drawdowns; the absence of any equity issuance at distressed prices over thirty years demonstrates the third. Three sentences. Thirty years of receipts.

What we read going forward

The twenty-six-year record is one of the cleanest capital-allocation documents in North American energy. It is also a record of a particular generation of capital allocator. Murray Edwards is in his sixties. Tim McKay is the operating successor, but the cultural transmission past the next decade is a real question that the operator's commentary touches on but does not fully resolve.

The succession question is the open Pillar II question on the desk. We read every proxy filing for the board composition, the executive bench, and the compensation plan tilt. The post-2022 compensation reset moved the plan further toward per-share metrics, which is the right direction. The next decade will tell whether the culture is operator-specific or institutional.

What survives the founder

The most important Pillar II question on the CNQ desk is also the most uncomfortable: what survives Murray Edwards. The Edwards record is one capital allocator's record. The institutional question is whether the discipline that produced the twenty-six raises is encoded deeply enough in the board, the management bench, and the compensation structure to outlast the original capital allocator's tenure.

The honest read is partial. Tim McKay's promotion to CEO in 2020 was a continuity choice; McKay had been the operational architect for two decades and produced the unit-cost discipline that the dividend record sits on. The post-McKay generation of executives is visible at the management ranks but not yet tested at the apex. The board composition has been stable, the compensation plan has been per-share-aligned since the early 2010s, and the corporate culture is famous in Calgary for its quiet seriousness. All of these are encouraging.

The cultural transmission test is the next decade, not the next quarter. We track the proxy filings for board renewals, the compensation plan for any drift toward absolute-production metrics, and the executive bench for the third generation of capital allocators. If the cultural transmission holds, the dividend record continues into a fourth decade. If it does not, the record peaks somewhere in the late 2020s and the operator becomes a normal Canadian heavy-oil name. The base-case probability, in our framework, is the first; the tail probability is the second; both deserve to be named explicitly because the entire CNQ thesis sits on the answer.

In the meantime, the 2026 capital plan reads as a continuation of the prior twenty-six years. Sustaining capex remains first-quartile; the dividend is raised; the buyback is modest and consistent; no equity is issued. The same playbook produces the same record. The reason most readers find it dull is exactly the reason we read it: dull is what compounding looks like when you write it down every year. The chart that records it does not need a price target, a panel discussion, or a thesis change. It needs a steady hand and a willingness to wait.