The Gap: Oil Shorts, Refinery Damage, and a Canadian Economy Trapped Between Both Outcomes

What the Paper Market Is Pricing

By late June, managed money in ICE Brent crude futures had turned net short by approximately 52,000 contracts, according to CFTC and ICE positioning data. [1] On the NYMEX WTI side, managed money remained net long at roughly 83,000 contracts, but that position was shrinking: net length fell by more than 13,000 contracts in the week ending June 23 alone. [1]

The directional shift matters, but the liquidity picture may matter more. Reuters reported that Brent crude open interest had fallen approximately 17% by mid-June, the steepest decline since at least 2009, with front-month open interest at its lowest in over a year. [2] Capital is not just rotating from long to short. It is leaving. The market that prices the global crude benchmark, and by extension Canada’s most important export, has less capacity to absorb shocks than at any point since the financial crisis.

What is this positioning consistent with? The International Energy Agency’s June Oil Market Report forecast that global oil demand would decline by 1.1 million barrels per day year over year in 2026, driven by higher fuel prices and product disruptions. [14] The U.S. Energy Information Administration made an equivalent revision: demand falling 1.1 million barrels per day, compared to its February expectation of 1.2 million barrels per day of growth. [15] OPEC+ has added approximately 800,000 barrels per day of supply since April. [16] Shipping through the Strait of Hormuz has partially resumed following the June ceasefire, though volumes remain below pre-war levels. [17] Saudi Arabia has cut official selling prices. The Brent curve has weakened into contango, with prompt supply swamping the market. [2]

The positioning is consistent with a market pricing demand destruction and supply normalization. What it creates, in the interim, is a structural condition: a thinner, less liquid market with diminished capacity to absorb a physical supply shock if one arrives. That is not a claim about what traders intend. It is a description of what the positioning enables.

What the Physical Market Is Showing

The physical market has not resolved. It has been buffered.

In Russia, Ukrainian drone strikes have systematically degraded refining capacity over the course of 2026. Energy Intelligence estimates that Russian crude processing fell to approximately 3.95 million barrels per day in June, down roughly 25% from year-earlier levels, the lowest throughput in more than two decades. [3] Gasoline output fell to approximately 850,000 barrels per day, down 17% from 1.03 million. [3] Reuters reported that the Moscow oil refinery, struck on June 16, is unlikely to resume production this year after damage to a primary processing unit accounting for 53% of the plant’s capacity. [18] Fuel rationing has been imposed in more than half of Russia’s 83 federal entities, and Russia has requested 50,000 metric tonnes of gasoline from Kazakhstan. [19]

On July 6, Ukraine reported striking the Omsk Oil Refinery, a Gazprom Neft facility processing approximately 460,000 barrels per day, the largest refinery in the country. [4] Independent damage assessment was not available at publication, but Ukrainian forces described it as the last of Russia’s eleven biggest gasoline-producing refineries to be hit. The strike occurred roughly 3,000 kilometres from the front line.

In the United States, the Strategic Petroleum Reserve has been used to buffer the displacement. SPR stocks fell to 319.5 million barrels as of the week ending July 3, the lowest level since May 1983. [5] The Department of Energy released a record 9.9 million barrels in the week ending May 15 alone, as part of a 172-million-barrel program to offset global shortages. [5] Before the 2022 emergency release cycle, SPR stocks sat above 560 million barrels. The reserve has been drawn down by roughly 240 million barrels in four years.

Refined product margins tell a parallel story. The Gulf Coast 3-2-1 crack spread stood at $57.84 per barrel on July 2 per EIA data, and market indicators suggest it has widened past $60 since. [6] The structural driver is downstream capacity. U.S. Gulf Coast refineries were running at 93 to 95% utilization before the Hormuz disruption, with almost no spare capacity to absorb additional demand. [28] No meaningful new refining capacity is coming online domestically, and several European refineries have closed permanently. The bottleneck has moved from crude supply to refining capacity, and that constraint does not resolve quickly. Earlier in 2026, cracks surged further still: jet fuel margins peaked near $74 per barrel in March, and gasoline margins reached approximately $35 per barrel. [20] The fact that spreads are widening even as crude prices fall is itself the signal: the physical shortage is downstream, not upstream.

At Cushing, Oklahoma, the WTI delivery point, stocks sat at 21.6 million barrels as of mid-June, a level Reuters described as “nearing operational minimums.” [21]

In February, this publication’s 2007 Benchmark analysis noted that energy and materials markets were showing late-cycle price action independent of broader economic slowing, echoing the pattern that preceded the 2008 commodity peak. Four months later, the physical damage has deepened even as the paper market has shifted toward pricing resolution. Crude processing is at a two-decade low. The SPR is at a four-decade low. But the Brent curve has moved into contango and OPEC+ spare capacity is being deployed. Both conditions are true simultaneously, and that is what makes the gap unusual.

How the Gap Resolves

There are two transmission mechanisms through which this dislocation can reach Canada. The first is direct. The second runs through the financial system.

The direct channel works through futures pricing. When managed money covers short positions en masse, it creates buying pressure in futures contracts. If that buying is concentrated in prompt-month contracts, the front of the curve can shift from contango, where future prices exceed spot prices, to backwardation, where spot prices command a premium. That shift changes physical trader behaviour: holding inventory becomes less attractive when prompt barrels are worth more than stored ones. Traders with optional barrels sell into the spot market. Refiners and consumers accelerate purchases. Physical prices can rise through spot differentials and prompt cargo premia. This mechanism amplifies or reveals existing scarcity. It does not create it. But the physical constraints documented above are real, and the open-interest decline means fewer participants remain to cushion the move if covering begins.

A second, less direct channel runs through financial plumbing. In March, this publication’s Late-Cycle analysis documented that roughly $16 trillion in government bond-backed repurchase agreements operate globally, with half at overnight maturity and approximately 70% of the non-centrally cleared segment carrying zero collateral cushion, according to BIS and ISDA data. [27] Oil-market volatility can tighten financial conditions through risk reduction, margin calls, and bond-market spillovers. If those spillovers reach the repo market, government bond yields can spike, which in turn affects Canadian fixed mortgage rate benchmarks. Over two million Canadian mortgages are set to renew at rates locked below the Bank of Canada’s current 2.25%. [7] This channel is less direct than the gasoline, inflation, and fiscal mechanisms documented below, and there is no established precedent for a commodity dislocation triggering this specific chain. But the BoC’s own Financial Stability Report warned that vulnerabilities in parts of the financial system have increased, and the channel illustrates how oil-market stress could transmit to Canadian households through pathways the direct price mechanism does not capture. [22]

No exact recent parallel lines up all four conditions: bearish crude positioning, declining open interest, physical refinery destruction at scale, and strategic reserve depletion. The 2020 COVID crash featured extreme positioning and demand destruction, but supply was not being simultaneously destroyed. The 2022 Russia-Ukraine invasion carried physical supply risk, but speculative positioning was not comparably bearish. The 2008 financial crisis is the closest match: commodity prices spiked through mid-2008 while the broader economy had already entered recession in late 2007. But even in 2008, refining capacity was not being physically disabled at this scale. Each prior episode contains part of the current pattern, but not the full structure. That does not make an outcome predictable. It makes the standard playbook less reliable.

If Oil Breaks Higher

In this scenario, physical reality forces repricing. Shorts cover. Brent moves toward $90 or above.

Canadian gasoline would be the first casualty. The national average reached approximately 190¢ per litre in early May even with the federal fuel excise tax suspension absorbing 10¢/L. [13] [12] That suspension expires September 7. Without it, today’s pump price of 160.9¢/L would stand closer to 171¢/L before any change in underlying crude or refining costs. [13] A return to spring supply conditions without the tax buffer could push the national average above $2.00/L.

Gasoline already accounts for the largest single contributor to headline inflation: 33.2% year over year in the May CPI print, pushing headline CPI to 3.2%. [8] Excluding gasoline, CPI ran at approximately 2.2%. Higher oil would widen that gap further, and the Bank of Canada has stated explicitly that it is focused on ensuring higher energy prices do not become persistent inflation. [7] The rate stays at 2.25%, or potentially moves higher. Household debt at 179.6% of disposable income means the cost of that hold is borne disproportionately by the most leveraged borrowers. [9]

On the production side, Canada’s ability to capture crude upside has improved. The Trans Mountain pipeline is running at its full expanded capacity of 890,000 barrels per day, and the WCS-WTI discount narrowed to $11.65 per barrel in June, the tightest spread since November. [11] [10] This is genuine structural progress from the pre-expansion period. But it has limits. TMX is already fully utilized, meaning there is no additional export buffer if demand exceeds current capacity. And the channel through which higher crude prices used to benefit the broader economy, a rising Canadian dollar, has weakened. FCC Economics, Scotiabank, and National Bank of Canada have all documented the decline: FCC reported that the loonie’s responsiveness to commodity prices has diminished since 2022, while NBC found the rolling correlation between daily Canadian dollar moves and WTI had turned negative in recent months. [23] [24]

This publication’s analysis of the Canadian dollar’s decoupling from oil documented the structural mechanisms: foreign ownership channels roughly $58.4 billion in annual profits out of the Canadian oil sector, and the Bank of Canada-Federal Reserve rate differential of 125 to 150 basis points creates a persistent drag on the currency. Those mechanisms have not changed. Higher oil prices still generate profits that flow disproportionately to foreign shareholders rather than recirculating domestically. The natural hedge that once made oil price spikes a mixed blessing for Canada, where a stronger loonie offset higher import costs, does not activate under current conditions.

If Oil Stays Low

In this scenario, the shorts are right. Demand destruction dominates. Brent settles in the mid-$60s or lower.

That outcome would confirm the IEA and EIA demand forecasts: global oil demand declining by 1.1 million barrels per day in 2026. [14] [15] It would mean the paper market was not mispositioning. It was pricing a global recession signal. Canada, with an economy that has stalled around zero growth for two consecutive quarters, would be entering that global demand downturn with elevated household leverage and limited fiscal room. [9]

The revenue side narrows. This publication’s Late-Cycle analysis documented Canada’s fiscal starting position in March: a deficit at 2.2% of GDP, debt-to-GDP at 41.2%, and PBO-estimated fiscal sustainability room of $46 billion. That room shrinks if the commodity side of federal and provincial revenue declines. The deficit spending strategy documented in The Squeeze has even less runway if oil stays low because the growth it was meant to produce has not materialized: GDP edged up just 0.1% in Q1 2026 after contracting 0.1% in Q4 2025. [9]

The Trans Mountain pipeline, which represents a genuine improvement in export infrastructure, becomes a different kind of asset at persistently low prices. The expanded 890,000-barrel-per-day system still needs to be serviced. The $35 billion investment documented in this publication’s Feedback Loop analysis carries a fixed cost structure that does not scale down with revenue. A pipeline at full capacity is an asset when prices are high. At low prices, the capacity generates diminishing returns against the same fixed obligations.

On the employment side, the May Labour Force Survey showed genuine improvement: 88,000 jobs gained, unemployment falling to 6.6%, a result that exceeded expectations and recovered most of the earlier losses. [25] That is a real resilience signal. But the Late-Cycle analysis identified employment as the last indicator to turn in every major downturn of the past fifty years, and a sustained commodity price decline would hit Alberta and Saskatchewan first, where full-time resource-sector employment is most exposed. Employment resilience does not disprove late-cycle vulnerability on its own.

The Bank of Canada, in this scenario, would have room to cut. Falling energy costs would ease inflation pressure. But the reason inflation eases in this scenario is demand destruction, not normalization. The BoC’s June statement framed policy as balancing “soft growth” against “increased inflation.” [7] If energy prices collapse, the framework changes but the problem does not disappear. It shifts from an inflation problem to an output problem, and rate cuts chasing a commodity-led contraction downward have limited effectiveness when the driver is external demand, not domestic credit conditions.

The Narrowing Buffer

Canada has real buffers. They should be stated plainly.

TMX is running at full capacity and the WCS discount has narrowed to its tightest level in months. Canadian producers are capturing more value per barrel than at any point since the pipeline expansion. [11] [10] Employment gained 88,000 jobs in May. [25] Gasoline has moderated from its May peak to 160.9¢/L. [13] The Bank of Canada’s Financial Stability Report stated that households and businesses “remain stable overall.” [22]

But the phrase “stable overall” describes the average, not the distribution. Consumer insolvency filings hit their highest first quarter since 2009. [26] Household debt stands at 179.6% of disposable income, a ratio that has risen even as rates have come down from their 2023 peak. [9] The federal fuel excise tax suspension expires September 7, which would add roughly 7% to the pump price overnight, absent any other change in market conditions. [12] The SPR has been drawn down by roughly 240 million barrels in four years and now stands at 319.5 million barrels, the lowest in four decades. [5] And Brent open interest has declined approximately 17% in 2026 alone, meaning the market pricing Canada’s primary export has notably less liquidity than it did at the start of the year. [2]

In February, this publication’s 2007 Benchmark noted that peak capital flows and commodity prices can coexist with late-cycle conditions, and that Canada’s starting position in 2025 was weaker than in 2007 on every measured dimension. In March, Late-Cycle Signals documented six late-cycle indicators running simultaneously and concluded that Canada’s starting position was weaker than either 2007 or 2019. In June, The Squeeze documented the domestic data confirming those warnings: consumer insolvencies, stalled GDP, gasoline-driven inflation, and federal projections that assume the average rather than the median.

The oil positioning data adds a new dimension to that cluster. The gap between paper and physical markets creates a condition where a correction, in either direction, arrives into a thinner market and is absorbed by an economy carrying elevated leverage and limited policy room. Canada does not need to predict which side of the gap is right. It needs both sides to resolve gently. The buffers exist, and some of them have genuinely improved since the pre-TMX era. But they are narrower than they were when these warning signals were first documented, and the margin for a gentle landing depends on conditions that neither the paper market nor the physical market fully controls.