On 11 July 2008, WTI crude printed an all-time high of $147.27. By mid-September, as Lehman filed for bankruptcy and Merrill Lynch was sold under duress, oil had already given back more than a third of its gains; by year-end it would trade below $40. Demand had not simply softened—it had evaporated. In response, the Fed slashed the funds rate to the zero lower bound and, within months, opened the QE spigots, purchasing trillions in mortgages and Treasuries to stabilise a financial system that had bet everything on perpetual appreciation. The trajectory was almost too clean: energy shock → tightening → demand destruction → recession → monetary pivot.
I have long argued that in 2026 the same machinery is running, though the staging has changed. The trigger this time is the Middle East war and the resulting supply shock through Hormuz. Brent is projected to average $86, with tail scenarios at $115. Diesel has now surged to record levels. The Fed, seemingly following a rehearsal script, hiked 25 basis points on 16 September—its first increase since 2023—with Chair Warsh declaring inflation “still too high.”
The Fed, however, is not “responding” to inflation. It is performing the tightening phase of a cycle whose endpoint, I argue, is already decided. The tightening is less a policy choice than the necessary precondition for the next loosening. Simply put, the system cannot justify another round of emergency credit creation without first precipitating the emergency. It has spent years preparing to manage exactly this kind of crisis. In this respect, the 2008 pattern was more a template than an accident.
I have traced this sequence many times before: the 2019 repo crisis, the BlackRock white paper calling for “going direct,” the pre-COVID pandemic simulations, the lockdowns that simultaneously suppressed inflation from monetary flooding and opened the door for unprecedented fiscal-monetary coordination. The patient was already on the operating table in September 2019, and COVID supplied the anaesthetic.
Today, the patient is again in pre-op. The AI bubble is the new subprime: a debt-financed structure built on fictitious capital, with hyperscalers burning cash they don’t have, off-balance-sheet commitments at $3.1 trillion, and $266 billion in AI-related bond issuance this year alone. The difference is that this time, the collapse is being rehearsed in public.
In my latest piece, I highlighted how Dario Amodei’s midnight warning about a rogue AI seizing the internet, Sam Altman’s sudden reluctance to IPO, and the bipartisan U.S. AI safety bill miraculously revived were not responses to a real threat. They are the ideological scaffolding for the federal backstop that OpenAI and its peers have sought from the beginning.
The sequence is familiar: private credit finances the buildout; mounting fragility demands a public guarantee; the guarantee requires an emergency vocabulary. In 2008, the vocabulary was “systemic risk” and “too big to fail.” In 2020, it was “pandemic.” In 2026, it is, above all, “AI safety.” While the form changes, the function stays the same. And it is here that a diesel export ban goes from being a policy detail to the visible rehearsal of the mechanism itself.
The debate is now public, and unresolved. Trump confirmed on 22 September that he has pushed advisers to consider a ban: “let’s not send out the diesel.” Treasury Secretary Scott Bessent confirmed that the administration is examining whether a full or partial ban would work, adding that a decision would come “fast, one way or the other.” Energy Secretary Chris Wright has warned that a ban could raise fuel and gasoline prices on the coasts and prompt refiners to cut production. Interior Secretary Doug Burgum cautioned against retaliatory measures by other energy exporters. Dozens of industry groups, from the API to the Chamber of Commerce, have lobbied hard against the idea. Some Republican lawmakers and midterm candidates have called for a ban ahead of midterms; others, including senior senators, have dismissed itas ineffective or a “gimmick.”
But there’s no doubt that the muzzle is loaded. The U.S. exports roughly 1.5-1.7 million barrels per day of diesel, and Europe’s reliance on those barrels has become acute. In August 2026, U.S. diesel accounted for more than half of total European imports, with the UK and Netherlands sourcing 62 to 72% of their diesel from American refiners. Analysts now describe refining capacity—not crude—as the main choke point in global energy markets. A ban would strand product on the Gulf Coast, force refiners to cut crude runs, and remove a significant chunk of global seaborne diesel trade—a policy-made choke comparable to Hormuz itself. European diesel futures already spiked on the rumour alone, while U.S. retail diesel has hit a record above $6.50 a gallon.
Whether the ban happens in full, partial, or surgical form is secondary. What matters instead is that the system is actively searching for the trigger. The public debate should be seen as part of the apparatus: it normalises the idea that supply can be weaponised—sending shockwaves from global stock markets down to localised agriculture and modern food delivery networks—that markets can be shocked by decree, and that the resulting instability is something the state must then manage.
This is emergency capitalism in what we could call its mature form. The system is addicted to shock and rescue because emergencies unlock the stimmy. Where there is addiction, there is a dealer who does not necessarily cause the crisis but knows when the crisis is coming and positions accordingly.
The 2008 template is being run again, with different actors and a different trigger. But the destination is the same: a Fed forced to loosen, a federal backstop for private losses, and another round of debt-financed survival for a system that can only reproduce itself by devaluing its fiat currencies and “managing” the inevitable immiseration of most of us.
