The Strait Is the Price: How Hormuz Risk Moved Through Oil Markets

The market was not pricing one simple barrel shortage. It was pricing a sequence in which military risk altered commercial decisions before the full consequences appeared in production statistics or the benchmark.

My Onyx submission argued that a barrel still in the ground—or even loaded at a Gulf terminal—was not equivalent to supply delivered safely and on time to the refinery that needed it.

This article tests that thesis against evidence available through 1 August 2026. It is written as a market case study: what happened to the delivery system, what the tape reflected, where the strongest signals appeared and what would have invalidated the view.

The separate technical note, From Geopolitical Event to Testable Market Thesis, explains the reusable research process.

1. The investment question

Onyx Capital Group asked entrants to evaluate the short- and long-term consequences of the US–Iran conflict for the oil market. The entry was limited to a 1,000-word paper or trade idea and a video of no more than two minutes [1].

I submitted both artefacts by 13 July. The related Future of Energy Trading event took place on 23 July [2].

Outcome: Submitted. No public ranking was issued.

The competition was the origin of the work, but the market question was narrower:

Was the shock primarily a loss of production, or was it also an impairment in the commercial delivery chain?

Flat crude combined realised supply loss, expectations of future disruption, precautionary inventory demand, freight, insurance, refinery outages, emergency-stock policy, non-Gulf supply, global demand, currencies and positioning.

Oil-market research has long distinguished physical supply shocks from demand and expectations shocks [13], [14]. A benchmark move was therefore evidence about the market’s aggregate belief, not proof of one mechanism.

2. From military risk to delivered supply

The core claim was simple:

Oil that exists but cannot be insured, transported and processed is not equivalent to delivered supply.

A Gulf barrel had to be produced, stored, loaded, insured, carried through Hormuz, delivered to the correct destination, processed by a compatible refinery and distributed as usable fuel. A failure between any two of those states could tighten the market without an immediate upstream production loss.

Insurance was part of supply

Reuters reported that indicative hull war-risk pricing reached about 3% of vessel value in early March, compared with roughly 0.25% before the conflict [8].

That was the illustrative hull premium on a $250 million tanker before cargo cover, financing, delay, operating risk and freight.

The economically important variable was not the headline quote alone. It was whether cover remained available in sufficient capacity and on terms compatible with the owner’s financing, charter obligations, crew policy and internal risk limits.

A route could therefore remain legally open while becoming commercially impaired.

Shipowners and charterers transmitted the shock

Governments changed the risk environment, but private firms translated it into physical behaviour. Underwriters chose capacity and exclusions; owners decided whether compensation justified vessel and crew risk; charterers chose whether to wait, reroute, replace the cargo or pay more for tonnage; financiers and counterparties determined whether the voyage remained permissible.

A naval escort could reduce attack probability. It could not compel ordinary underwriting, willing crews or commercially acceptable freight.

The relevant market state was not simply “open” or “closed”. It was a spectrum between ordinary transit and exceptional passage.

3. The physical chokepoint

Hormuz was structurally important before the conflict.

EIA estimated that 20.9 million barrels per day of oil passed through the strait in the first half of 2025—about one fifth of global petroleum-liquids consumption and one quarter of maritime-traded oil. The same route carried 11.4 Bcf/d of LNG, more than one fifth of global LNG trade [3].

The exposure was concentrated. Saudi, Iraqi, Kuwaiti, Qatari, Bahraini and Emirati exports relied on Gulf loading infrastructure; Asian refiners received most Hormuz crude; Qatar relied on the route for LNG exports; only part of normal oil flow could bypass the strait; and refinery compatibility and delivery timing limited substitution.

Bypass capacity was a buffer, not a replacement

EIA estimated that Saudi Arabia’s East–West pipeline and the UAE’s Habshan–Fujairah route could provide about 4.7 million b/d of combined bypass capacity [3].

versus

Even that comparison overstated immediately available relief.

Nameplate capacity was not spare capacity. Existing utilisation, terminal capacity, crude compatibility, storage, destination and pipeline operations determined what could actually be redirected.

The bypasses reduced the tail risk. They did not eliminate the chokepoint.

Physical risk remained observable

The International Maritime Organization listed 62 confirmed regional incidents as at 27 July 2026 [7].

That count did not measure barrels lost, but it supported the commercial logic behind elevated insurance and owner caution. The probability distribution facing a shipowner was not theoretical.

4. What the physical data and the tape said

The post-submission record did not provide a clean natural experiment. It did provide enough cross-market evidence to test the delivery thesis.

Crude flows recovered, but normality did not

The IEA reported that global oil supply rebounded by 4.1 million b/d in June to 98.8 million b/d, while remaining 9.4 million b/d below the pre-war level. Gulf exports, including bypass flows, reached 16.1 million b/d—well above the earlier disruption but still below the roughly 24 million b/d pre-war average [5].

That evidence rejected a permanent-total-closure thesis. It did not establish ordinary delivery.

A market could move from catastrophic scarcity to constrained transit without returning to its pre-war state.

Products carried the cleaner scarcity signal

Crude recovered faster than refined products and LPG. The IEA reported that product exports remained below half their pre-war level while crude recovered to nearly three quarters, with product cracks and refinery margins reaching multi-year highs [5].

Reuters reported the European low-sulphur gasoil crack at $74.66/bbl on 30 July, even with crude trading well below its wartime peak [10].

Consumers did not use crude directly. They used diesel, jet fuel, gasoline, LPG and petrochemical feedstocks. A crude cargo at sea did not repair a damaged refinery, restore the correct product slate or refill inland distribution inventories.

For an investor, the product market was therefore not a secondary detail. It was evidence about where physical scarcity remained most acute.

Inventory composition mattered more than the headline total

The IEA reported that global observed stocks rose by 21 million barrels in June, the first increase in four months. But oil on water increased by 117 million barrels while onshore stocks continued to draw by roughly 96 million barrels. OECD inventories fell by another 62 million barrels, including barrels supplied through government releases [5].

More oil at sea was not the same as replenished tanks at consuming centres.

The market was moving cargoes again while the buffers nearest to end use remained depleted.

Emergency stocks bought time

IEA member countries made 400 million barrels of emergency stocks available in March, the agency’s largest collective action [6].

Those stocks could replace part of a temporary shortfall, reduce panic buying, protect essential demand and buy time for rerouting, repair or diplomacy.

They could not restore insurance capacity, create tanker availability, repair refineries, replace every crude grade or support an indefinite disruption.

A stock release was therefore bearish for the prompt shortage without proving that the route mechanism had disappeared.

Financial repricing and physical disruption were related but distinct

A geopolitical event could affect markets through a broad financial channel and a physical oil-contract channel at the same time. The first worked through risk appetite and growth expectations; the second worked through insurance, freight and delivered supply.

That distinction prevented a broad risk-off move from being treated automatically as evidence of a new physical barrel loss.

The benchmark traded beliefs faster than the system repaired

Brent settled at $84.09 on 28 July after several days without new strikes [9].

By 31 July it settled at $90.12 as flow concerns returned [12]. Two laden VLCCs exited Hormuz that day, but overall traffic remained thin [11].

The sequence did not prove that one headline caused one exact price change. It showed a timing mismatch.

The benchmark could reprice de-escalation before the physical system normalised. It could also rally on renewed risk before a new barrel was visibly lost.

5. What the market was actually pricing

Flat Brent was the net result of several premia moving in opposite directions.

The cleanest interpretation was not that one market was “right”. Different markets priced different parts of the chain.

Insurance priced concentrated voyage loss. Freight priced usable tonnage and delay. Spreads and differentials priced prompt location-specific scarcity. Cracks priced refinery and product constraints. Brent priced the aggregate balance plus macroeconomic feedback.

That was why a falling benchmark could coexist with continuing physical stress.

6. Positioning implications at the 1 August cutoff

This section is a research read-through, not a live recommendation. A real position would still require current curves, options surfaces, freight, liquidity, carry, margin and execution costs.

My regime call at the 1 August cutoff was constrained transit. Gulf exports had recovered to 16.1 million b/d but remained materially below the roughly 24 million b/d pre-war average; product exports were still below half of normal levels; war-risk cover remained elevated; and the European gasoil crack stood at $74.66/bbl. That combination was inconsistent with full de-escalation because the commercial and product systems had not normalised. It was also weaker than a sustained-disruption call because crude flows had recovered substantially and some laden transit was occurring. The table below defines what would have moved the view away from constrained transit.

What I would not infer

I would not infer commercial normality from:

• one escorted voyage;
• one tanker exit;
• a ceasefire headline;
• a lower Brent close;
• an announced stock release.

Nor would I treat one composite score as a substitute for conflicting evidence.

If freight, insurance and products remained tight while Brent fell, the disagreement was analytically useful. It suggested that the benchmark was absorbing a demand or policy offset rather than disproving the delivery constraint.

Portfolio read-through

For a diversified asset manager, the shock had several channels beyond crude beta.

Product scarcity could improve margins for unaffected refiners while raising feedstock and operating risk for exposed plants. Tanker economics could benefit from scarce tonnage and higher freight even as vessel, crew and insurance risks rose. Importing countries with weaker inventories, currencies or fiscal buffers faced faster transmission into inflation and external balances. Energy-intensive industries could continue paying high delivered fuel costs even when flat crude fell. Headline-driven repricing against slower physical repair also created gap risk that linear positions handled poorly.

Those read-throughs required security-level and regional analysis. They were not automatic buy or sell signals.

7. What survived, what changed and what failed

What survived

The post-event evidence supported the central mechanism. Delivery reliability mattered alongside production; insurance and freight transmitted geopolitical risk; crude and products recovered at different rates; onshore inventories revealed more than aggregate stock headlines; relative and delivery-sensitive markets often carried cleaner information than flat Brent; and bypasses and emergency stocks reduced the shock without repairing the route.

What changed

The original phrase route risk was too compressed.

It also had to distinguish legal reopening, exceptional passage, partial physical recovery and ordinary commercial normality. Those were different states.

What failed

An earlier draft used a weighted route-risk score:

where $R$ represented route conditions, $S$ shipping, $G$ the supply gap, $P$ product tightness and $M$ price impulse.

I removed it before submission.

The weights were not calibrated. The update speeds were unspecified. The result did not map defensibly to a position.

Most importantly, the score averaged away the disagreement that mattered. A benchmark decline alongside worse traffic and stubborn product cracks was not a data error. It was evidence that different parts of the market were pricing different shocks.

8. What would invalidate the thesis

The persistent-delivery-premium thesis should be reduced or abandoned when several observations occur together:

• ordinary underwriting capacity returns;
• war-risk terms fall materially;
• multiple laden vessels complete safe voyages without exceptional arrangements;
• freight and delays normalise;
• production and refinery runs recover;
• onshore inventories rebuild;
• product cracks normalise;
• the markets expected to confirm the mechanism fail to respond.

No single observation was sufficient in every case.

A political agreement mattered most when insurers, owners, producers, refiners and inventory data behaved as though it was credible.

9. Limits

The evidence had important limits.

AIS visibility was incomplete because vessels could disable transponders or lose coverage. Public vessel counts were therefore lower bounds.

Insurance figures were indicative quotes, not a public panel of executed transactions. Terms varied by vessel, owner, cargo, charter, route, flag and underwriter.

Freight, crude-grade, refinery and product data were uneven and revised on different schedules.

The late-July Brent moves were event-window descriptions. They did not identify a unique causal shock because demand expectations, inventories, non-Gulf supply and positioning were changing at the same time.

The map shows structural routes and approximate pipeline paths, not live vessel positions or navigational detail.

The correct response was not false precision. It was to keep the claim narrow:

A persistent Hormuz premium required evidence that commercial and physical delivery remained impaired after the first political headline changed.

Conclusion

The conflict exposed a delivery network, not one barrel balance.

Oil could exist at the wellhead while remaining unavailable to the final buyer because the voyage was uninsured, the tanker was absent, the route was exceptional, the refinery was constrained or the required product was missing.

That was why the strongest evidence did not sit in Brent alone. It appeared across insurance, vessel behaviour, freight, Gulf export volumes, inventories, refinery operations and product cracks.

The submission’s central thesis survived in a narrower and more useful form:

The persistent premium was the price of moving energy through a concentrated network whose commercial and physical links recovered at different speeds.

At the reconstruction cutoff, that evidence described constrained transit. The premium should be considered resolved only when the commercial and physical links normalise together.

References

[1] Onyx Capital Group, “Research Competition,” 2026. [Online]. [Accessed: Aug. 2, 2026].

[2] Onyx Capital Group, “The Future of Energy Trading,” Jul. 23, 2026. [Online]. [Accessed: Aug. 2, 2026].

[3] U.S. Energy Information Administration, “World Oil Transit Chokepoints,” Mar. 3, 2026. [Online]. [Accessed: Aug. 2, 2026].

[4] U.S. Energy Information Administration, “Global Energy Security Data,” 2026. [Online]. [Accessed: Aug. 2, 2026].

[5] International Energy Agency, “Oil Market Report—July 2026,” Jul. 2026. [Online]. [Accessed: Aug. 2, 2026].

[6] International Energy Agency, “IEA member countries to carry out largest-ever oil stock release amid market disruptions from Middle East conflict,” Mar. 11, 2026. [Online]. [Accessed: Aug. 2, 2026].

[7] International Maritime Organization, “Middle East—Highlighted (Confirmed) Incidents,” Jul. 27, 2026. [Online]. [Accessed: Aug. 2, 2026].

[8] Reuters, “Maritime insurance premiums surge as Iran conflict widens,” Mar. 6, 2026. [Online]. [Accessed: Aug. 2, 2026].

[9] Reuters, “Oil prices fall 5% to two-week low after several days without US-Iran strikes,” Jul. 28, 2026. [Online]. [Accessed: Aug. 2, 2026].

[10] Reuters, “Refinery attacks keep European fuel prices around record highs,” Jul. 30, 2026. [Online]. [Accessed: Aug. 2, 2026].

[11] Reuters, “Two VLCCs exit Strait of Hormuz, where traffic remains thin,” Jul. 31, 2026. [Online]. [Accessed: Aug. 2, 2026].

[12] Reuters, “Oil prices settle more than 1% higher, log strongest month since March,” Jul. 31, 2026. [Online]. [Accessed: Aug. 2, 2026].

[13] L. Kilian, “Not All Oil Price Shocks Are Alike: Disentangling Demand and Supply Shocks in the Crude Oil Market,” American Economic Review, vol. 99, no. 3, pp. 1053–1069, 2009.

[14] D. R. Känzig, “The Macroeconomic Effects of Oil Supply News: Evidence from OPEC Announcements,” American Economic Review, vol. 111, no. 4, pp. 1092–1125, 2021.