The risk premium decays before the risk does
Tanker traffic through the Strait of Hormuz collapsed in March and has not recovered. Brent went to $117 and then gave most of it back. Both of those things are still true, which is the problem.
There is a habit in risk work of using the oil price as a proxy for geopolitical risk in the Gulf. It is convenient, it updates continuously, and it is free. It is also, in a specific and dangerous way, measuring something else.
The last six months are an unusually clean demonstration, because the physical disruption and the price have been telling completely different stories and the divergence is documented on both sides.
What the traffic did
The Strait of Hormuz is the narrowest point in the seaborne crude system: bounded by Iran on the north side and Oman on the south, roughly twenty-one nautical miles across at its tightest, and the only maritime exit for the Gulf's producers. Tanker movements through it are observable from vessel transponder data, which means the physical fact does not depend on anyone's announcement.
Traffic ran between twenty-eight and forty-two tankers a day from September through February, drifting down over the winter and recovering in February. In March it went to under one. It has not been above five in any month since, and the most recent reading is 1.3.
Whatever one believes about the politics, that is not a scare or a threat or a headline. It is a sustained physical interruption of the single most important maritime route in the energy system, now in its sixth month.
What the price did
Brent averaged $70.9 in February and $103.1 in March, the month the traffic stopped. It peaked in April around $117, roughly 65% above the pre-disruption level.
Then it came down. By June it was $85.4, by July $83.8, and August has run at $90.8. From the April peak the price has retraced something like sixty per cent of the move, while the transit count has stayed at roughly a tenth of normal throughout.
Why this happens, and why it is a trap
Markets do not price levels of risk. They price changes in expectations about levels of risk, and then they get used to things.
Three mechanisms drive the retracement, and none of them require the underlying situation to improve:
Adaptation is real and it is fast. Pipelines that bypass the strait get run at capacity, storage gets drawn, tankers reroute, and cargoes that were destined for one refinery go to another. Physical systems have more slack than they appear to, and the market discovers that slack over weeks.
The tail gets repriced downward simply by not happening. In March the distribution included genuinely catastrophic outcomes: escalation to regional conflict, strikes on production rather than transit. Each month those do not occur, the market assigns them less weight. This is rational updating and it is also exactly how everyone gets caught, because the mechanism is indistinguishable from complacency until afterwards.
Positioning unwinds. Some of the spike was speculative length that has since been reduced. That flow reverses regardless of the fundamentals.
The result is a price that says the situation has substantially normalised, sitting on top of transit data that says nothing of the sort. An institution using the oil price as its Gulf risk gauge marked its exposure down through the summer while the physical position did not improve at all.
The price is not measuring the risk. It is measuring how surprised the market still is by the risk, and surprise has a short half-life.
Reading it properly
The discipline is the one that applies to every risk indicator that is also a traded price: separate the state from the news about the state, and use different instruments for each.
For the state, use the physical series: transits, storage, spare capacity, pipeline utilisation, freight and insurance rates on the specific route. These describe the situation and are largely immune to sentiment. War-risk insurance premia are particularly useful here, because unlike the flat price they are quoted on the specific hazard rather than on the global supply-demand balance.
For the news, the price is exactly right, and its moves are informative about what the market has just learned.
Conflating them produces the two classic errors in opposite directions. Reading a spike as evidence of a fundamental shift buys the top of a move that is mostly surprise. Reading a retracement as evidence of resolution, which is the live error today, writes down an exposure that has not actually improved.
What the price is for
The two series in this piece describe the same situation and disagree about it, and only one of them is what most risk processes actually read.
That disagreement is not a market failure. A price is supposed to reflect the change in expectations, and expectations genuinely have changed: adaptation happened, the catastrophic tail did not materialise, positioning unwound. Every one of those is a real update. The price is doing its job.
The difficulty is that the job it is doing is not the job it gets used for. A gauge that measures surprise, read as a gauge that measures danger, will always be most reassuring at the point where a persistent risk has been absorbed into the baseline, which is exactly the configuration on the page, and exactly when a hedge is cheapest and hardest to justify to anyone.
← Back to all insights