Oil balance outlooks need scenario discipline

Observed facts: the near-term balance is assumption-heavy

The International Energy Agency's July Oil Market Report describes a market recovering from a severe disruption but still operating below pre-conflict supply. It forecasts global oil demand declining by 1 million barrels a day in 2026 before rebounding by 2 million barrels a day in 2027. On supply, it expects a 3.7 million barrel-a-day annual decline in 2026, followed by a 7.5 million barrel-a-day increase in 2027 if transit volumes improve. The IEA says the balance could move back into surplus, but makes that result contingent on de-escalation, recovering tanker traffic and the restart of production and refining capacity [1].

The US Energy Information Administration's publicly available July Short-Term Energy Outlook uses its own recovery path. It assumes that most crude production returns to near pre-conflict averages by the end of 2026 and that most remaining shut-in output returns in the first quarter of 2027. The EIA also reports that trade-flow adjustment and demand reduction had exceeded its earlier expectations [2]. Those are not minor modelling details: the timing of reopened routes, restarted fields and recovered consumption changes whether inventories draw, stabilise or rebuild in any quarter.

Observed facts: producer policy is part of supply, not background noise

OPEC+ policy adds another conditional variable. On 2 August, seven participating countries announced a 188,000 barrel-a-day production adjustment for September. They also reaffirmed compensation for earlier overproduction and said they would continue monthly reviews of market conditions [3]. This is an observed policy decision, not evidence that every announced barrel will reach the market on schedule. Conformity, compensation, infrastructure availability and the capacity to reverse or pause adjustments all affect realised supply.

The distinction between horizons matters as well. OPEC's World Oil Outlook 2026 projects global oil demand reaching 124 million barrels a day by 2050 and estimates substantial long-term investment requirements [4]. The IEA and EIA materials cited here concentrate on the short-term shock, recovery and inventory cycle [1][2]. A long-range demand pathway and a twelve-to-eighteen-month balance forecast answer different questions. Treating their headline numbers as if they were direct substitutes creates a false disagreement before the underlying assumptions have even been aligned.

Oakhampton inference: compare the bridges, not just the endpoints

Oakhampton's inference is that commodity diligence should translate each outlook into a small set of explicit bridges between today's observable market and the forecast endpoint. The first bridge is physical restoration: shipping access, field restarts, refinery throughput and product availability. The second is demand response: how quickly consumption returns after disruption, and whether price, economic conditions or substitution leave a lasting effect. The third is inventory behaviour: whether stock releases and draws are masking a continuing deficit or providing time for supply to recover. The fourth is producer conduct: announced adjustments, actual output, compensation and the option to change course.

A practical comparison can therefore hold the headline forecast lightly and test the same commercial exposure under several coherent cases. A faster-restoration case combines improving transit, quicker production recovery and inventory rebuilding. A slower-restoration case retains constrained flows, delayed refinery recovery and continuing product tightness. A demand-weakness case allows consumption to recover more slowly even if crude supply improves. These cases are not predictions and should not be assigned false precision. They are a way to expose which assumption drives a contract, acquisition, working-capital need or operating plan.

• Record each source's forecast date, horizon and units before comparing headline numbers.
• Separate crude availability from refinery throughput and refined-product availability.
• Track announced producer policy against realised output and compensation obligations.
• Reconcile inventory direction with stock releases, oil on water and onshore commercial stocks.
• Stress timing and liquidity effects even where the annual average appears balanced.

Remaining uncertainty: public outlooks cannot settle asset-level exposure

The largest uncertainty is path dependence. The cited outlooks were prepared at different dates and embed different assumptions about geopolitical developments, shipping access, demand elasticity and production recovery. Their figures will be revised as those conditions change. Public materials also cannot establish a particular asset's crude slate, storage access, refinery configuration, hedge book, contractual pass-through rights, shipment timing or financing headroom. Those details determine whether a market-level swing becomes an operating loss, a cash requirement or a manageable timing issue.

The defensible conclusion is therefore narrower than choosing a winning forecaster. Current official sources show a market in which demand, restored supply, inventories and OPEC+ policy can move at different speeds [1][2][3]. Long-term outlooks add a separate question about capacity and investment [4]. Decision-grade analysis should preserve those distinctions, state the assumed recovery path and identify the observation that would invalidate it. Scenario discipline does not eliminate uncertainty; it makes the commercial consequences of being wrong visible early enough to manage.