In late February 2026, the Strait of Hormuz, the narrow waterway through which roughly 20% of globally traded oil and 20% of LNG passes, effectively closed. What followed was described by the International Energy Agency as the largest supply disruption in the history of the global oil market.
Brent crude surged from around $80 to $114 per barrel within weeks. Diesel prices in Europe, already under pressure from earlier disruptions, spiked sharply. The EU estimated that gas prices rose 70% and oil by 50%, adding an additional €13 billion to European fossil fuel import costs. Analysts at Brookings warned of scarcity pricing in diesel specifically, the fuel that powers road haulage, logistics, and a significant share of European fuel retail volume.
For fuel retail operators across DACH and Italy, the weeks that followed were a test of something most had not explicitly designed their pricing systems for: operating under conditions outside the historical norm.
When rules-based pricing meets a novel situation
Most fuel retail pricing systems, whether manual, rules-based, or competitor-matching, are calibrated against historical market conditions. The rules reflect what worked before: price within a band of the market, respond to competitor moves by a set increment, protect margin by holding a floor.
Those rules perform well when the market behaves as it has historically. They perform poorly when the market moves in ways that fall outside the conditions they were calibrated on.
The Hormuz closure was precisely that kind of event. Crude prices did not drift — they jumped. Supply constraints did not emerge gradually — they arrived at scale within days. Competitor behaviour became less predictable as operators across networks made independent decisions about whether to pass on cost increases immediately or absorb them temporarily to protect volume.
Rules calibrated on historical conditions cannot tell you what the right price is when conditions change fast enough to invalidate the history.
In this environment, a rules engine faces a structural problem. Its recommendations are, in effect, a function of past market behaviour. When past behaviour stops being a reliable guide to current dynamics, the engine continues to produce outputs — but those outputs are extrapolations from a world that no longer applies.
What causal models do differently under pressure
The distinction between a rules-based and a causal pricing system is not primarily one of speed or data volume. It is one of what the system actually measures.
A causal pricing model — one built using methods like Double Machine Learning — measures the relationship between price and demand directly, from each station's own data, with non-price influences removed. That measurement is specific to the station, specific to the fuel type, and specific to the customer mix the station actually serves.
When crude prices spike and the market moves into a regime that historical rules did not anticipate, the causal model has something the rules engine does not: a validated estimate of how your own customers respond to price, independent of what the market around you is doing. A fleet-heavy site with price-insensitive customers can absorb a cost increase differently from a high-street petrol forecourt where retail customers have three alternatives within five minutes of driving. The rules engine treats both with the same logic. The causal model does not.
Pricing Intelligence re-prices daily against live competitor price feeds. During the Hormuz period, as competitor pricing became volatile and unpredictable, that daily recalibration, combined with per-station price sensitivity, meant recommendations could respond to actual market conditions rather than to historical patterns that no longer held.
The Kiel Institute finding
Research from the Kiel Institute for the World Economy, published in March 2026, modelled the economic consequences of a full Hormuz closure. One of its key findings is directly relevant to how pricing systems should be designed: in the short run, firms cannot easily switch suppliers, reroute shipping, or renegotiate contracts. The short-run scenario produced sharply larger price spikes precisely because rigidity in the system — refineries configured for specific crude grades, LNG terminals needing months to secure alternative cargoes — amplified the shock rather than absorbing it.
Pricing systems face an analogous rigidity. A system built around historical rules and competitor-matching is configured for the market conditions it was trained on. When those conditions change sharply, the rigidity shows up as recommendations that are slow to adjust, or that adjust in the wrong direction, or that treat all stations within a network as if they face the same market dynamics.
The operators best placed to manage a supply shock are not the ones with the most data. They are the ones who know how their own customers actually respond to price.
What this means for how you think about pricing resilience
The Hormuz crisis is, in one sense, a tail event. Most years, pricing operates in more stable conditions. But the lesson it offers is not specific to extreme disruptions.
Market conditions change. Competitors enter and exit. Regulatory changes shift the pricing environment — the 12-Uhr-Regel debate in Germany is a recent example of policy intervention reshaping how operators can set prices. Consumer behaviour evolves as fuel mix changes and EV penetration grows.
A pricing system that knows your own customers' price sensitivity — measured causally, validated statistically, refreshed from your own ongoing data — does not need to be redesigned each time the external environment shifts. The core measurement adapts because it is grounded in your actual market, not in a historical approximation of it.
The Hormuz closure was an unusually sharp stress test. Most pricing environments apply a slower, quieter version of the same pressure. The question it forces is not whether your pricing tool survived March 2026. It is whether your pricing system is built to know the difference between what the market is doing and what that means for you.
All market data in this piece is sourced from public reporting as noted below. Backwell Tech has no commercial relationship with any of the cited institutions.
About Backwell Tech
Backwell Tech is a Berlin-based high-tech company specializing in predictive AI solutions. The platform offers companies scalable AI models for profit maximization by utilizing historical and real-time data and ensuring data integrity. Since its founding in 2019, Backwell Tech has combined cutting-edge research with practical innovation in explainable algorithms. The company focuses on ethical AI development and delivers reliable, interpretable forecasts that enable informed business decisions. More information at www.backwelltechcorp.com.
Backwell Tech Corp contact:
Maximilian Gismondi
Sources
- Market disruption scale: International Energy Agency (IEA), quoted in World Economic Forum. 'Beyond oil: 9 commodities impacted by the Strait of Hormuz crisis.' April 2026.
- Crude price data: Bloomberg. 'From chokepoint to crisis: The Strait of Hormuz and global oil markets.' March 2026. Brent crude $80 to $114 per barrel.
- EU cost estimate: Euronews. 'Strait of Hormuz shutdown: What implications for Europe?' March 31, 2026. Gas +70%, oil +50%, €13bn additional import cost.
- Diesel scarcity risk: Brookings Institution. 'From chokepoint to crisis: The Strait of Hormuz and global oil markets.' May 2026.
- Short-run rigidity finding: Hinz, J., Mahlkow, H., and Wanner, J. (2026). 'The Cost of Closing the Strait of Hormuz.' Kiel Policy Brief 206. Kiel Institute for the World Economy.
- Oil volumes through strait: U.S. Congressional Research Service. 'Iran Conflict and the Strait of Hormuz.' 2026. ~20 million barrels per day, ~27% of global maritime oil trade.
- Backwell methodology: Chernozhukov, V., Chetverikov, D., Demirer, M., et al. (2018). Double/debiased machine learning for treatment and structural parameters. The Econometrics Journal 21(1), C1-C68.
- Pilot data: Backwell Tech internal pilot, 4-station network, 110-week backtest. Results indicative of opportunity, not a guarantee of future performance.