Uddrag fra JP Morgan, først dansk resume, bagefter originial tekst:
As JPMorgan’s commodity team write in its latest note unlike a typical disruption where spare production capacity can be mobilized quickly, the location of the shock and the scale of current supply losses mean the immediate adjustment has to come from barrels already in storage. Put simply, inventories are acting as the shock absorber of the global oil system, according to JPM’s Natasha Kaneva.
Some history: the crisis began with stocks in relatively strong shape. Global inventories surged in 2020 and 2021 during the COVID demand collapse, then drew sharply through 2022 and 2023 as economies reopened, Russia invaded Ukraine, and markets tightened. That tightening phase reversed in 2024 and 2025, when supply once again outpaced demand, allowing stocks to rebuild. As a result, the world began 2026 with a healthy 8.4 billion barrels in storage.
Of that 8.4 billion barrels, roughly 6.6 billion are held onshore, with another 1.8 billion afloat. Some of the floating barrels are simply in transit from producers to consumers, others — such as sanctioned Russian and Iranian crude — effectively function as floating storage. By type, about 5.2 billion barrels are crude, while 3.2 billion are refined products. Visibility varies widely. OECD inventories are among the most transparent because member countries maintain strategic reserves, collect standardized data, and publish timely statistics—giving OECD stocks an outsized role in price formation. Much of the rest of the world is far less observable, particularly across the developing world. China is the notable exception, where inventory estimates began the year near 1.3 billion barrels.
Finally, it is important to distinguish between Strategic Petroleum Reserves—state-controlled emergency barrels—and commercial inventories, which are privately held stocks used in the normal course of trade and refining.
Even so, not every barrel can be drawn. Out of the 8.4 billion barrels in global inventories, JPMorgan estimates only 0.8 billion barrels are realistically available without pushing the system into operational stress. As we wrote previously, as of April 23rd, roughly 280 million barrels have already been consumed to cushion the impact of the conflict.
While on paper, that still suggests comfortable buffers, in practice the picture is more complicated. Floating storage can be tapped quickly, but only a slice of onshore inventories—around 580 million barrels—is readily accessible. The rest is effectively locked up in pipeline fills, minimum tank levels, and other operational constraints.
This is why inventory floors matter, as we explained at the start of the month. A market can still hold hundreds of millions of barrels, and yet become fragile once working stocks fall too low. Like blood pressure in the human body, the issue is circulation. Pipelines lose pressure flexibility, terminals cannot load efficiently, refiners struggle to secure the right grades on time, and traders bid aggressively for nearby supply. The system does not fail because oil disappears, it fails because the circulation network no longer has enough working volume. The same principle applies to refined products. Product inventories are somewhat more flexible than crude, but a meaningful portion must still be maintained as a strategic and operational buffer, particularly to support critical sectors such as transportation and aviation (this is the core thesis behind Trump’s blockade of Iran: the assumption that if enough shut ins take place, the country’s entire oil infrastructure will collapse as half of it is low pressure and thus extremely vulnerable to permanent failure).
OECD inventories offer a clear example of this operational floor. Historically, OECD product inventories, including both commercial and strategic reserves, have rarely fallen below ~35 days of forward demand, roughly 1.6 billion barrels, suggesting a practical lower bound.
In a prolonged disruption scenario, demand is therefore rationed well before inventories approach critically low levels. In theory, stocks could last much longer—but only at the cost of reduced consumption, lower refinery runs, and a broader economic slowdown. As a result, a full drawdown of global inventories is neither feasible nor likely, simply because prices will be so high it will become uneconomic to purchase oil.
Like an onion, inventory draws happen in layers. The sequence is determined by speed of access, economic cost, political willingness, and logistical ease, not by who has the most barrels. Easy barrels are peeled first, strategic barrels later, critical barrels rarely. The deeper the draw, the higher the price and the economic cost of each additional barrel withdrawn.
- Oil-on-water and floating commercial stocks (first layer). Cargoes already on tankers, storage vessels, or held offshore are the easiest to mobilize. They can be redirected quickly and typically require no policy decisions. Inventories on water began the year at 1.8 billion barrels and have fallen by 140 million barrels over the past two months, drawing at an average rate of 2.7 mbd. With the last cargo from Hormuz having arrived on April 20, the headline draw rate should now start to moderate.
- Commercial onshore stocks (second layer). Refinery tanks, terminal inventories, and merchant storage in hubs such as Cushing, ARA, or Singapore are next. These barrels are commercially owned and can respond to price signals relatively quickly. OECD commercial stocks have only recently begun to decline, drawing at about 2.2 mbd in April after a more modest 0.45 mbd draw in March. In absolute terms, OECD commercial inventories fell from 2.8 billion barrels in February to about 2.72 billion barrels today, a decline of 74 million barrels.
- Strategic Petroleum Reserves (third layer). Government-controlled emergency stocks such as the IEA coordinated systems or the US Strategic Petroleum Reserve are typically tapped only when disruptions are severe or politically sensitive. Current SPR releases are running at around 2.5 mbd across the US, Japan, and South Korea. Since March 26, when Japan released its first strategic barrels, OECD strategic reserves have dropped by 61 million barrels.
- Demand destruction replaces inventory draws (the pivot). Well before the system is emptied, high prices begin to ration demand. Consumers drive less, industry cuts runs, airlines trim schedules, and refiners reduce throughput. In effect, the market is shifting from a “managed” adjustment—driven by SPR releases and floating storage—to a “forced” adjustment, where price becomes the primary balancing tool. Observed global oil demand fell by an average of 2.8 mbd in March and is tracking a larger 4.3 mbd decline so far in April. Demand losses are expected to deepen to around 5.5 mbd in May, as demand destruction must scale to prevent the system from hitting operational floors. Given the magnitude of the demand pullback, JPM’s balance now suggests OECD commercial inventories are on track to approach operational stress levels by early June.
- Operational minimum stocks are typically the last barrels drawn, and, in practice, are deliberately avoided. Pipeline fill, tank bottoms, linepack equivalents, minimum terminal inventories, and the product stocks required for day-to-day continuity are rarely accessed, because drawing them materially increases the risk of operational disruption and broader system instability.
According to JPM calculations, OECD commercial stocks could fall to these operational floors by September if the Strait of Hormuz remains closed, assuming demand destruction stabilizes at 5.5 mbd. In fact, it is possible that demand destruction reverses as many governments are subsidizing oil prices to avoid social unrest, thus neutering the role of demand destruction as a natural inventory reduction buffer, which in turn accelerates inventory drainage and could hit “operational bottoms” even faster.
What this means, practically, is that all else equal, if the Strait of Hormuz is still blocked by September (and realistically by the operational stress level threshold expected, to be hit in June), the world will have no choice but to force a resolution and unblock the waterway as the alternative would be going into uncharted territory, and risking unprecedented energy collapse, and a global depression.
