How to read a Canadian oil & gas operator in seven numbers

Most sell-side decks open with three things: a logo, a price target, and a chart of production growth. None of those tells you whether the business is good. A Canadian oil and gas operator is read in seven numbers, and the order matters more than the magnitudes. We go through them the same way every quarter, on every name in the coverage universe.

The order is deliberate. If a name fails on numbers one through three, we do not bother with the rest. If a name passes on those and stalls on four through seven, the read is "interesting, not yet." If a name passes on all seven, we have already spent four years reading it, and we already own it.

1. Sustaining capex per barrel of oil equivalent

This is the number that tells you whether the business is real. Take the operator's average annual sustaining capital expenditure across a five-year window. Divide it by the operator's average annual production in BOE. The result is the dollar cost, per barrel, of keeping production flat — before any growth capex, before any debt service, before any return to shareholders.

A first-quartile Canadian operator pays under US$15 per barrel of sustaining. The cohort average sits in the high teens. The structurally disadvantaged operators sit at US$25 or higher; those are the ones who quietly issue equity in every downturn because the business cannot fund its own depletion.

The number matters in two directions. Low sustaining capex is the structural advantage every other number on this list builds on. And a sudden _rise_ in sustaining capex, particularly one that does not show up in management's commentary, is the leading indicator of an operator that is moving up the cost curve as the best acreage gets depleted. Read Canadian Natural (CNQ) for the upper-end of the Canadian sustaining-capex discipline; the low-decline, long-life asset mix is the structural reason the number stays low.

2. Free cash flow break-even WTI

This is the number that tells you whether the business survives the next trough. Calculate the WTI price at which the operator generates exactly zero free cash flow after sustaining capex, after the base dividend, and after maintenance interest. The lower the number, the more cushion the operator has when the price tape turns.

Most Canadian operators print a break-even between US$40 and US$55 WTI at current strip differentials. The first-quartile names sit in the high US$30s. The marginal names sit at US$60 or higher; the marginal names with high decline rates sit at US$70 or higher. The arithmetic is unforgiving. An operator with a US$65 break-even and a US$58 average WTI does not "wait it out." It issues equity, cuts the dividend, or both.

We track the break-even at flat differentials each quarter for every operator on the desk. The number that moves slowly is the structural read; the number that moves a lot quarter-to-quarter is telling you about apportionment, transport, or hedging changes, none of which are the business.

3. Net debt over cash flow at strip

The third number tells you whether the operator can take a punch. We compute net debt divided by trailing-twelve-month operating cash flow at the current strip price for WTI and AECO. Anything under 1.0× is conservative; the 1.0–2.0× range is normal for a Canadian E&P with a credible discipline record; above 2.0× is a problem the next downturn will surface.

The two extremes matter for opposite reasons. A name like Tourmaline (TOU) operates at effectively zero net debt and a meaningful cash balance; that is a soft signal about Mike Rose's view of his own cycle. A name with 2.5× net debt to cash flow heading into a trough is a name whose dividend is going to test the patience of its shareholder base.

Capital allocation under stress is the only honest test. The 2015 and 2020 records of every operator on the desk are public; the 2026 forward decisions are not.

4. Per-share production growth, ten-year compounded

The fourth number separates operators from speculators. We take per-share production in 2015 and 2025 and compute the compounded annual growth rate. Absolute production growth is easy; issue shares, buy assets, watch the headline rise. Per-share production growth is the part the operator earns by keeping the share count disciplined.

A first-quartile Canadian operator delivers 3–5% per-share production CAGR over a decade with no share-count growth. The cohort average is closer to 1–2%. The marginal names print zero or negative per-share growth despite issuing equity every other downturn; those are the operators whose investor decks always feature the absolute number and never the per-share one. ARC Resources (ARX) is one of the cleaner per-share growth examples among the gas-weighted Montney operators since the 2021 Seven Generations merger.

5. Insider open-market buys, last four quarters

The fifth number is the only one that costs the operator something to fake, which is why we read it last among the operator-facing metrics. Options grants are loyalty to a quarter; RSUs vest on schedule; restricted stock prints on the proxy. Open-market purchases by named executives with their own money are loyalty to a decade.

We track the absolute dollar value of insider buys, the number of distinct buyers, and the share-price level at which each buy was executed. A consistent four-quarter buy record from multiple executives at multiple share-price levels is a signal we trust. A single CEO buying a small lot near the prior peak and selling on the way up is a different signal. Both are public.

6. The 2015 and 2020 dividend behaviour

The sixth number is a backward look. Most operators on the desk have lived through two near-existential commodity stress tests in the last decade: 2015–2016 and 2020. The dividend behaviour in each is the cleanest single record of how the operator's capital culture performs under pressure. We compile the table for every name.

The cohort splits roughly into four buckets. Operators who held the dividend in both stress windows: CNQ, ENB, TOU, FTS. Operators who held in one and cut in the other: SU (cut 2020, held 2015), several Canadian midstream names. Operators who cut in both: most of the Canadian heavy-oil pure-plays. Operators who had no dividend to cut because they were too distressed to pay one: a long list of names that are no longer on the desk.

The 2015 cohort is informative. The 2020 cohort is informative. The operators who passed both are the ones whose 2026 decisions we trust without re-doing the work.

7. Ten-year reserve replacement at flat prices

The seventh number is the one most operators do not want to print. Reserve replacement at flat prices means the operator's drilling and step-out activity produces enough new proved reserves to offset depletion without the help of a rising commodity price. A 100% replacement at flat prices means the business is, in the long run, durable. Below 100% means the operator is in a slow drawdown of the asset base.

The trick the industry uses is reserve replacement at strip prices — a generous price deck makes uneconomic reserves "economic" and inflates the number. The flat-price replacement is the honest one. Most disclosed Canadian operators report it; the ones that do not should be asked why.

What you do not need to read

We do not weight headline production growth as a Pillar I metric. We do not weight EBITDA multiple expansion as a Pillar III metric. We do not read management's price-deck slide, ever. We do not read the analyst-day strategic-priorities deck; we read the proxy compensation plan instead, because the proxy is the part the operator has signed.

Most of what gets written about Canadian energy is decoration on top of these seven numbers. If the seven are good, the writeup is interesting. If they are not, the writeup is a story.

What the seven numbers do not see

The framework has limits. Three of them are worth naming, in case the reader mistakes the seven for a complete picture.

The first limit is jurisdictional and political. A Canadian heavy-oil operator with a clean read on all seven numbers still owns assets that sit downstream of two specific political variables: the Canadian federal carbon and emissions framework, and the US administration's posture on Keystone-class pipeline approvals. The seven numbers will tell you the operator's break-even at flat differentials; they will not tell you how the differentials themselves move under a regulatory regime change. The 2018–2019 widening of WCS-WTI to over US$50 was a regulatory and pipeline-capacity event, not an operator event. The clean operators came through it; the operators with thin sustaining-cost margins did not.

The second limit is reservoir quality at the per-asset level. Sustaining capex per BOE is an aggregate number across the operator's full asset base. A name with a healthy aggregate sustaining number can be drilling its best acreage today and its second-tier acreage in three years, with a sustaining number that drifts higher even though the operator's framework appears unchanged. We track the per-asset disclosure where it exists; many operators provide enough field-level data to read it. The exceptions, where field-level data is sparse, are themselves a soft Pillar I signal.

The third limit is technology. A meaningful unit-cost reduction from a new completion technique, a new water-handling design, or a new well-pad layout can change the break-even by US$5–10 per barrel within a two-year window. The 2018–2024 step-change in Montney well productivity was the cleanest recent example; the operators positioned in the right acreage compounded a per-share advantage that the prior seven-number read did not anticipate. The framework is a structural read, not a forecast.

These limits do not invalidate the framework. They are the reason we do not run the seven numbers in isolation. Each operator's Halvren Checklist scorecard sits alongside the seven; the checklist captures the cyclical and people questions that the numbers cannot.

What the framework does well

Where the framework earns its keep is the rejection step. Most of the work of any research desk is sorting names into "worth reading carefully" and "not worth the next hour." The seven numbers, applied in order, produce a clean reject decision in under fifteen minutes per name. An operator with sustaining capex of US$25 per BOE and net debt of three times cash flow does not need a sixty-page diligence pack to be classified; the seven numbers have already told the reader what they need to know.

The accept decisions take longer. An operator who passes the first three numbers cleanly is worth the rest of the framework, and the rest of the framework typically takes weeks of reading. The 2015 and 2020 records are public, but the texture of the decisions inside those windows is in the operator's letters, the proxy filings, the conference-call transcripts, and the principal interviews from the period. The seven numbers tell you whether to do the work. The work itself is still the work.

The seven numbers are not novel. They are not proprietary. They are the same seven numbers a careful operator would use to read its own business. The reason we publish them is that the sell-side framework keeps treating production growth as the operating metric and the dividend as the capital metric, when both are downstream of these seven. Read the seven first. The rest follows.