NorthHelm Advisory | Report 01
Reserve Growth and Equity Returns
More reserves. What reaches the shareholder?
Read the report ↗Report 01 of 08 · September 14, 2026
More producing reserves per share. A stronger stock return?
We did not establish a reliable forecasting relationship in this study.
The practical question is what the growth costs, how it is funded and what each existing share retains.
What we tested
Growth in proved developed producing (PDP) reserves per year-end basic share against following June-to-June stock returns: 103 company-years, 32 Canadian oil and gas companies, four annual periods. PDP reserves are assigned to producing wells.
What the result means
This study does not give investors a reliable basis for choosing stocks on PDP growth per share alone. That does not prove there is no relationship, or that reserve growth lacks economic value.
The study window: reserve year-end to following-year returns
We pair a year-end reserve measure with a later full year of stock returns. Public disclosure sits between the reserve year-end and the study entry, subject to the historical-verification limits below. The measured return runs from the following June close to the next June close. It does not begin on the announcement date or measure the immediate news reaction.
- Reserve year-end
Measure the company’s reserves and shares. - Public disclosure
Information becomes public; historical timing is only partly verified. - Following June close
The measured stock-return period begins. - Next June close
The full-year return period ends.
Same companies. Same years. Two ways to measure returns.
Unadjusted stock returns: the stock’s return with the effects of oil prices, gas prices, currency and the wider market still included. Dividends are included, and stock splits are accounted for so they do not create artificial gains or losses.
Factor-adjusted returns: the return remaining after removing the model’s estimated contribution from oil prices, gas prices, currency and the wider market. The model uses each stock’s sensitivity to those influences, estimated before the return period. It cannot remove every outside influence or establish causation.
Both use split- and dividend-adjusted prices as a total-return proxy. The factor adjustment is a model estimate, not the result of an actual hedge; neither specification is proved right.
How an investor should read this
How we describe relationship strength
Consistent descriptive scale: using the absolute, unrounded correlation: below 0.10 = near-zero; 0.10 to below 0.30 = weak; 0.30 to below 0.50 = moderate; 0.50 to below 0.70 = strong; 0.70 to 1.00 = very strong. Positive or negative states the direction separately. These are editorial descriptions used throughout this note, not universal statistical thresholds or tests of reliability. Uncertainty intervals still govern what can be concluded.
Unadjusted stock returns: +0.113 indicates a weak positive relationship. On average across the four years, companies with higher PDP growth per share tended to rank slightly higher on subsequent stock returns. The uncertainty interval, −0.090 to +0.304, includes no relationship and relationships in either direction.
Factor-adjusted returns: −0.023 indicates a near-zero negative estimated relationship. After the model accounts for oil, gas, currency and market influences, the average relationship between PDP-growth-per-share rankings and return rankings is near-zero and negative. Its interval, −0.195 to +0.165, also spans negative, zero and positive relationships.
Both numbers are relationship scores, not percentage gains or losses. Each averages four annual rank correlations on a scale from −1 to +1. The dot shows the estimate; the bar shows its uncertainty. Their difference does not prove that market factors caused the unadjusted relationship.
Analytical cutoff: September 8, 2026. FY2021–2024 reserve measures; returns begin at a delayed June entry. This is not a test of immediate announcement reactions. Historical coverage and verification against original filings remain incomplete.
01 · What does each measure tell you?
Growth is only one part of the shareholder story.
PDP is defined above. Future development costs (FDC) are reported costs to develop reserves; the funding measure below uses proved-reserve FDC.
Five measures tested. No reliable return-ranking signal established here.
What each result means, and what the measure asks
These descriptions summarize the estimates, not proven forecasting strength. “Near-zero” and “weak” describe estimate size; the intervals determine how much confidence to place in it.
These five tests compare company rankings, not the size of a stock-price move. A positive number means companies ranking higher on the measure tended to rank higher on later factor-adjusted returns; a negative number means they tended to rank lower. For measures labeled “lower,” the ranking is reversed so that less share growth or a lighter funding burden ranks higher. Even a higher-ranked return can be a loss. The bar shows uncertainty about the relationship, not a range of stock returns. [S1]
PDP growth per share: -0.023 indicates a near-zero negative estimated relationship. The measure asks: did the PDP reserve base grow faster than the basic share count?
Total PDP growth: -0.004 indicates a near-zero negative estimated relationship. The measure asks: did the company grow that base, regardless of ownership changes?
Lower growth in basic shares: +0.025 indicates a near-zero positive estimated relationship. The measure asks: did existing owners face less expansion in the share denominator? Consolidation adjustments remove the mechanical effect of recorded share consolidations. The measure tracks net changes in basic shares outstanding, including issuance and share reductions; it does not include potential dilution from warrants or convertibles. [S1]
Net acquisition share: -0.063 indicates a near-zero negative estimated relationship. The measure asks: what proportion of the reserve additions included in this measure came from acquisitions after dispositions? The total combines development additions, revisions including economic factors, and net acquisitions; negative category totals count as zero. The acquisition example shows the calculation.
Lower FDC/funds flow: +0.137 indicates a weak positive estimated relationship. The measure asks: how small were reported proved-reserve FDC relative to positive annual funds flow? Neither net acquisition share nor lower FDC/funds flow captures the return earned on invested capital. [S1]
Why a valuable asset can still be an unattractive investment
Economic interpretation, not a tested causal explanation: reserve volume does not capture margin, decline, development spending, tax or claims ahead of common equity. Financing can change the owner’s share of the asset, and the entry price can already reflect an expected improvement. [S1]
02 · A single average conceals very different years
A positive year is not a repeatable signal.
The latest year looks encouraging. The earlier years do not tell the same story. From June 2025 to June 2026, companies with higher PDP growth per share tended to rank higher on factor-adjusted stock returns. In the three earlier years, the relationship was negative or almost absent. An investor therefore cannot treat the latest year as evidence of a dependable pattern. The question is whether it can repeat, and under what conditions. [S1]
The same metric has not ranked returns consistently across these years
Why might the latest year look different?
The study does not establish why, so these are hypotheses to investigate, not findings. Reserve growth may have coincided with stronger cash generation, better financing or other company news that investors valued. The companies represented also differ across years, and a few companies’ rankings or ordinary sampling variation could influence an annual result. The factor model may leave some differences in commodity or market exposure unaccounted for.
A stronger oil price alone would not settle the explanation: this chart already uses factor-adjusted returns. To distinguish the possibilities, we would compare the same companies across years, test sensitivity to removing individual issuers, and reconcile their reserve changes with dated cash-flow, financing and news records. Until then, +0.308 is a reason to investigate, not a demonstrated explanation or a forecast.
The definition of excess return also matters
Three alternative factor adjustments on the same 103 company-years
| Return specification | PDP/share rank correlation |
|---|---|
| Oil and gas only | +0.05 |
| Oil, gas and market | −0.01 |
| Energy-equity beta | +0.10 |
S1 Exhibit 4, p4, at its published precision. Raw and primary estimates: the opening exhibit. Other window, sample and price-return checks: S1 pp4–5. Separate check that reduces extreme estimated market sensitivities: S1 p13.
The primary model estimates issuer sensitivities before entry; the energy-equity model uses sensitivity to the iShares S&P/TSX Capped Energy Index ETF (XEG), a Canadian energy-equity benchmark. Subtracting a common index return from every stock cannot change their within-year ranking. Neither approach guarantees commodity neutrality. [S1]
Canadian gas and heavy-oil benchmark price differences (AECO and Western Canadian Select, or WCS), exposure to Brent oil prices, local contracts, hedges and changes in debt or assets are not fully captured. This is model attribution, not causal management performance or a realized hedge. Trading costs, borrow and market impact are not modeled. [S1]
Boards and investors should challenge what each model removes and retains. Four annual paths establish neither the correct return model nor performance across future regimes.
03 · 13.3% and 54.1% can both be correct
Same company. Two correct percentages.
Before comparing acquisition-led growth, ask what was bought and what was sold. This worked example uses Surge Energy’s FY2021 reconciliation: acquisition share is 13.3% after subtracting dispositions and 54.1% using gross acquisitions. The cards and table show how both the acquisition amount and total additions change. These percentages describe proved-reserve additions, not stock returns; the headline forecasting test uses PDP reserves per share. [S1]
Surge Energy FY2021: reconcile the convention before comparing the share
Surge Energy FY2021: reconcile the convention before comparing the share
| Annual convention | Acquisitions included | Total additions included, using net or gross acquisitions as labeled | Share |
|---|---|---|---|
| Acquisitions after subtracting dispositions | 1,656 Mboe | 12,481 Mboe | 13.3% |
| Gross acquisitions | 12,767 Mboe | 23,592 Mboe | 54.1% |
Surge Energy proved-reserve reconciliation, S1 p8. Mboe = thousand barrels of oil equivalent. [S1]
What makes up the total reserve additions included?
| Proved-reserve component (Mboe) | Net basis | Gross basis |
|---|---|---|
| Drilling and development additions | 4,003 | 4,003 |
| Revisions, including economic factors | 6,822 | 6,822 |
| Acquisitions, net or gross as labeled | 1,656 | 12,767 |
| Total reserve additions included | 12,481 | 23,592 |
The denominator adds these three categories. A negative combined category amount counts as zero. Production and the reconciliation line labeled "other" are excluded. The result is neither total reserves nor the net annual reserve change. [S1]
This study uses net acquisition share. Booked, Built or Bought uses gross acquisitions and aggregates several years. The cards show only the annual definition difference. S1, page 8 retains the exact formulas, missing-component treatment and zero-denominator rule.
Put the comparison to work
Boards: show acquisitions and dispositions separately before the net result, then link the reserve change to development commitments and financing. This lets directors distinguish the size of the transaction from what remains after asset sales.
Investors: align drilling, revision and acquisition definitions before comparing issuers. Then examine what each share paid for, and retains from, the additions. A low net acquisition share does not establish drilling-led or purely organic growth; revisions can contribute.
Acquisition intensity remains an exploratory research question, not an established trading rule or a measure of the cost borne by each share. [S1]
04 · Does the funding ratio add information, or repackage another exposure?
A lighter funding burden: useful question, unproven signal.
Funding ratio: reported proved-reserve FDC divided by reported positive annual funds flow. A lower ratio may indicate less development burden relative to cash generation. It can also describe a more developed reserve base, a favorable commodity year or different leverage. These are competing interpretations, not established causes of share performance. [S1, S2]
The test includes 91 observations from 29 issuers. Seven observations with nonpositive funds flow and five with missing funds flow are excluded, not assigned favorable ratios; one missing-funds-flow observation also lacks FDC and is not counted twice. Missing is not low risk. The ratio is not committed finance, a near-term shortfall or a schedule for completing development. [S2]
Does accounting for leverage change the funding signal?
Same 66 company-years from 20 issuers. Does a lighter development burden go with better subsequent factor-adjusted returns? These cards compare that relationship before and after also accounting for leverage.
Relationship between a lower FDC/funds-flow ratio and better return rankings, after accounting for PDP growth per share and momentum (the stock’s prior return).
The same comparison, now also accounting for reported net debt relative to funds flow.
See all six matched-sample comparisons
Add one explanation at a time, comparing the same rows
| Added control | Rows / issuers | Before → after | 95% interval after |
|---|---|---|---|
| None: PDP/share + momentum base | 91 / 29 | +0.135 → +0.135 | [−0.112, +0.359] |
| Reported net debt / funds flow | 66 / 20 | +0.189 → +0.066 | [−0.219, +0.349] |
| PDP / proved reserves | 91 / 29 | +0.135 → +0.168 | [−0.087, +0.413] |
| PDP reserve scale | 91 / 29 | +0.135 → +0.158 | [−0.078, +0.413] |
| Annual gas share of boe | 81 / 26 | +0.146 → +0.143 | [−0.161, +0.400] |
| FY-end cash flow / basic-share equity value | 54 / 18 | +0.221 → +0.229 | [−0.288, +0.616] |
S2 p1. Before = funding partial rank correlation controlling for PDP/share growth and momentum. After adds the named control only. Each row has its own matched sample. Intervals condition on four annual paths; these are not percentage returns.
The funding measure may overlap with the balance-sheet story. Its estimated relationship with returns becomes smaller after accounting for leverage, but the uncertainty is too wide to establish that leverage explains it. Both ratios use funds flow in the denominator, and companies report net debt differently. Read the funding ratio alongside the balance sheet rather than treating it as a separate, proven return signal. [S2]
The valuation row is especially limited: it uses fiscal-year-end, not June-entry, equity value in a restricted CAD sample. It does not establish independence from entry valuation. Its cash-flow measure does not deduct sustaining capital and is not free-cash-flow yield. Gas share is a volume control, not full economic or hedge-adjusted exposure. [S2]
High FDC alone does not establish that a company will need to raise external funding. FDC describes development spending associated with the reported reserves; it is not, by itself, a near-term cash shortfall or a financing plan. The funding question depends on when spending is planned or committed, the cash the business can generate, available liquidity and other obligations. Boards and investors should request that dated cash-and-spending bridge before inferring a need for new debt or equity. This is a practical interpretation of the measure, not a financing outcome established by the return test.
05 · June entry asks a different investment question
The timing of the test matters.
The entry date defines the question. This study starts at the following June close, when reserve news may already be reflected in the share price. It tests later returns, not the immediate response to an announcement. An inconclusive result here does not establish that reserve announcements lack economic significance. [S1]
See the study-window graphic near the opening. The evidence limitations below explain what has and has not been verified.
What keeps the forecasting question open
An incomplete historical universe. The test includes 32 of the source collection’s 50 issuers. It is not a complete historical security census and does not comprehensively reconstruct failed, acquired or delisted issuers. The direction of the selection effect is not established. [S1]
The sample is concentrated in repeat observations. Twenty-one issuers supply all four years, accounting for 84 of 103 observations (81.6%); eighteen source-panel issuers contribute none. The analysis starts from 133 annual candidates, reaches 128 after source/comparability checks, and retains 103 usable return observations. These are coverage facts, not proof of completeness. [S1, pp5–6]
Availability is not value certification. A purposeful audit checked 12 company-years across four companies: current-year reserve and financial documents preceded entry by 108–127 days. The other 91 observations lack that timing verification. The original values at both ends of every paired growth measure have not been fully checked against the original filings; some financial values come from later comparative disclosures. [S1]
Four market paths, not future regimes. The intervals account for repeated observations from the same issuer but condition on four observed annual paths. They do not measure uncertainty across future regimes. The headline also depends on the model used to adjust returns for market and commodity moves. [S1]
A release-day study would ask another question. It needs the original timestamp, the market’s prior expectation and an event window that handles after-hours releases. Annual earnings and reserves are often released together, so even an immediate price move cannot automatically be attributed to reserve news alone. [S1, p3]
For a board, this separates the economic content of a disclosure from the market response around it. For an investor, it separates information that was available from information that was still an opportunity. That original-filing verification remains incomplete. [S1]
06 · Make the next growth claim answerable
Four questions to take into the next decision.
Connect added reserves to spending, cash, financing and the share count. These diligence questions support a board or investment discussion; they are not validated return predictors.
Evidence to request, then a decision to revisit
| Question | Boards: request and act | Investors: examine and act |
|---|---|---|
| What must be spent? | Tie additions to dated development commitments. Revisit timing when costs change. | Separate drilling, revisions, acquisitions and dispositions. Compare associated future development costs. |
| Who supplies the cash? | Match cash under weaker conditions to funding dates and terms. Identify unfunded decisions. | Check funds-flow definitions, recurrence and timing. Investigate missing or nonpositive denominators. |
| What does each share retain? | Show what existing owners contribute and retain. Reassess financing terms when ownership changes. | Reconcile year-end basic shares, consolidations and issuance. Check contingent dilution, excluded from basic shares. |
| What is already in the price? | State the capital-allocation assumptions. Revisit the decision when they change. | Use information available at entry and a consistent share basis. Examine what the price assumes. |
NorthHelm diligence framework, informed by this study [S1, S2].
Monitor: compare the bridge with the following year’s actual development spending, financing and share count. Boards revisit capital decisions; investors revisit entry assumptions. New facts do not supply an automatic buy or sell signal.
Before a stronger forecasting claim: complete the historical security universe and checks against original filings, then test a fixed rule on independent evidence. Reanalyzing this sample is not independent validation.
Now available: Promises and Delivery — Report 02 ↗. Published September 30, 2026, the second report examines share-price and trading-volume responses to guidance, revisions and annual results. This publication update leaves the present study’s analysis unchanged.
Evidence you can inspect
S1: Does reserve growth predict equity returns?
Reader Package pages 8–21. Estimates: S1 pp1–5; definitions: p8; timing: pp9–12; exploratory checks: pp13–14. Page ranges refer to the unchanged original Reader Package.
S2: Does funding burden add information beyond simpler explanations?
Reader Package pages 22–24. Funding controls, exclusions and limitations. Page ranges refer to the unchanged original Reader Package.
Resampling whole issuer histories recognizes repeated observations, but the intervals still condition on four annual market paths. They do not measure uncertainty across future regimes. More bootstrap draws cannot substitute for more independent periods. [S1]
Original published Reader Package
Start with the decision you are weighing
Which growth claim needs better evidence before your next board or investment decision? Discuss it with NorthHelm.