# The Three-Seas Crisis, the Refining Bottleneck, and France's Ledger: Notes on MacroVoices #554 Source: Realpha Blog (blog.getrealpha.com) Original article and charts: https://blog.getrealpha.com/en/blog/macrovoices-2026-10-22-macrovoices-554-dr-anas-alhajji-louis-vincent-gave/ > Educational listening notes on MacroVoices episode 554, covering the Middle East energy bottleneck and France's fiscal stress. It discusses frameworks and where the risks sit; it contains no stock recommendations, price targets, or buy/sell advice. Published: 2026-10-23 Locale: en Tags: energy, diesel, macro, geopolitics, european-debt, podcast-notes ![At dusk, a pipeline runs beside an empty highway toward a distant refinery, with a vacant tanker-truck loading bay in the foreground](/covers/macrovoices-2026-10-22-macrovoices-554-dr-anas-alhajji-louis-vincent-gave-cover.png) > The cauldron's leg is broken; the duke's meal is spilled. There is disgrace. Misfortune. —— *Book of Changes*, Ding hexagram, fourth line (pre-Qin China; my own translation) In the MacroVoices episode produced on 22 October 2026 (#554), energy analyst Dr. Anas Alhajji relayed what US Energy Secretary Chris Wright said at Riyadh Energy Week: crude moving through the Strait of Hormuz is back to 80–90% of pre-conflict levels, around 13 to 14 million barrels a day, while refined products are running at only 25–35% of pre-crisis volumes. Later in the same episode, Gavekal founder Louis-Vincent Gave noted that French tax receipts already run about 44% of GDP against roughly 24–25% in the US, so closing the deficit with tax hikes would collect less money, and that most French government debt now sits with foreign holders. Put the two halves together and the message I took away is that this round of energy stress is stuck in refining and shipping — a reading that holds as long as refineries keep being treated as military targets and governments keep cushioning consumers with subsidies. ## The one number I used to watch was Brent Let me start with what I had wrong. For the past year, whenever an attack came out of the Middle East, my first move was to pull up Brent. The circuit in my head was short: a producer gets hit, supply drops, crude spikes, energy names rally. Plenty of people were calling for $200 oil earlier this year, and although I never bought into the trade, I was running their framework — I assumed this crisis would settle itself in the crude price. That framework carried two assumptions I never audited. One was that the Middle East story stayed in the Middle East; French streets burning and European budget fights went into a separate folder. The other was that demand would hold, because the US numbers I happened to look at hadn't broken. Anas had said all along that $200 wasn't coming, and prices proved him right. I filed that as a good price call. Listening to this episode, I realised he had been reading a different table from mine the whole time. ![Three stacked horizontal bars show source, processing and delivery; the source bar is nearly full while the processing bar holds less than a third, so the gap appears in the middle layer.](/figures/three-layers-crude-vs-products-en.svg) ## The 25-to-35 number took my circuit apart Erik Townsend said the line that woke me up: there is no consumer of crude oil other than refineries. We put gasoline in cars, diesel in trucks, jet fuel in planes. So when crude flows recover to 80–90% while products sit at 25–35%, the missing piece can't be drilled back — it is stuck in the stage between the well and the pump. And a new refinery takes roughly ten years to build from scratch, which the world economy cannot sit through. Worse, that middle layer is being thinned from three directions at once. First the refineries themselves: California closed a batch, Germany closed a batch on climate policy, then Ruwais in the UAE was hit, Russian refineries were hit, Saudi Arabia's Jazan was hit. Anas was blunt about the people cheering strikes on Russian refining capacity — the bill lands on farmers in Africa and South America, and on households in Egypt and Kenya. Second, logistics: the Black Sea cost the world 800,000 barrels a day of diesel, and he says the press refused to touch it for months, until Trump raised it publicly. Third, biofuels, which is the link I had never traced: El Niño wrecked India's sugarcane crop, sugar prices rose, so cane flows to sugar instead of ethanol; in Indonesia, falling river levels stranded the barges that move palm oil, so biodiesel went with them. Neither of those has anything to do with refineries, and both land on the same hole. ![Three arrows point from the left, top and right at a central block for refining and fuel supply capacity, labelled plant closures and strikes, shipping, and biofuels.](/figures/three-forces-squeezing-refining-en.svg) That led Erik to a scenario I have been chewing on since. Crude caps out near $100 while crack spreads blow out to $200, diesel reaches twenty or thirty dollars a gallon, the world slides into a depression, and the depression then suppresses crude demand — so the barrel never moons, the processing margin does. Anas's answer was that they had worked through that last time, and the live question now is whether the crack spread has a ceiling and whether demand has a floor. ![Two stacked bars: the crude segment stays the same height while the processing spread above it grows from a thin sliver to taller than the crude itself.](/figures/crude-flat-crack-spread-up-en.svg) On demand he brought numbers. Last December the IEA expected US oil demand to grow 160,000 barrels a day in 2026 and OPEC expected 110,000; with monthly data now in for the first seven months, the year-over-year change is close to zero. So why no recession yet? Two pillars, he says: hundreds of billions of dollars of AI and data-centre capex, plus the extra fuel subsidies the EU pushed out in the second quarter. One detail I enjoyed — the only profitable segment left in global aviation is cargo, because all that AI equipment has to be flown somewhere, while everything passenger-related has declined. Pull either pillar and the recession arrives. ![Three vertical bars compared by height: the two forecast bars stand well above zero while the third actual bar sits almost flat on the zero line.](/figures/demand-forecast-vs-actual-en.svg) China reframed things for me too. I used to read "China bans product exports" as a master plan. Anas's version is that China was manoeuvred into the position: it holds domestic product prices below world prices, so its refiners lose money every day and can only earn on exports, which the state then rations by quota — in October, no exports at all. European leaders closing plants and California closing plants concentrated the world's remaining spare refining capacity in Chinese hands. Then he read out a news item: China and Russia are quietly advancing Arctic maritime cooperation, linking the maritime Silk Road to Russia's trans-Arctic corridor. That is an alternative to Hormuz, Malacca, Bab el-Mandeb and the Panama Canal — and it was at that point I finally understood why anyone wants Greenland. They want the route. ![On the left a slab is held level by two pillars; on the right one pillar is missing and the slab tilts down.](/figures/two-pillars-holding-the-slab-en.svg) His own remark at the International Energy Forum is the part I wrote down. Everyone is talking resilience and diversification, he said, but under the veneer, resilience means infrastructure redundancy, which costs money, and diversification brings more volatility than before, because optionality means buying imports when they are cheap this month and switching back to domestic supply when they aren't. None of that volatility, he says, is embedded in any energy market outlook. The French half picked up the thread. Gave's line is that economic activity is energy transformed — the price of food on the table has diesel inside it — so past a certain diesel price, anger reaches the street. He uses geographer Christophe Guilluy's three zones. Zone A is the centre of Paris, Lyon and Bordeaux: government, finance, knowledge industries, flying the EU flag. Zone B is the outer suburbs, largely migrants who work as nannies, Uber drivers and restaurant cooks for zone A, and whose flag in these demonstrations is the Palestinian one. Zone C is the rest of France, where shops are boarded up and hospitals and schools are closing, flying the tricolour. He even noted the three play different sports — tennis, soccer, rugby. Zone A held power for years by keeping B and C divided, and now A is running out of money to pay for B; B turns violent, A gets frightened, and A starts voting for the protection of C. That is the mechanism behind Le Pen polling in the mid-40s. He is equally blunt on the options: push the retirement age to 70, or devalue by 50%. The first is brutally hard, the second is impossible inside the euro, and with a presidential election six months out, not one candidate will say either word. ## Three gauges I watch now, and two things I can't call I rebuilt my dashboard after this. Brent stays, with two gauges added beside it: the diesel crack spread, and the share of refined products getting through the key straits. That 25-to-35 figure tells me more about the state of the middle layer than the barrel price does. The third gauge is government subsidies and AI capex, because those two pillars are absorbing the pain for everybody, and the first two gauges only start to bite once they loosen. ![On the left is a single gauge; on the right, three gauges sit side by side with their needles resting at low, high, and loosening positions.](/figures/dashboard-one-to-three-en.svg) Here is a question I suspect some readers share with me: crude went up, so why didn't my energy position follow? This episode gave a concrete answer — the margin that widened is collected in the refining layer, and what you own may sit upstream or in services. So now I make myself write one sentence before sizing anything: which layer is this money earned in — source, processing, or delivery? If I can't write it, I bought a direction rather than a cash flow. This is the same discipline as valuation: before deciding what a business is worth, identify which square of the chain it occupies and whether that square is loose or tight right now. The harder question is whether to cut positions when the headlines read like this. Gave, at 52, says he cannot remember a stretch in his career with this many large tail risks stacked at once — France, energy, and whether the AI capex boom ends. His response isn't to bet on collapse. He points out that shorting French debt against German debt costs 150 basis points a year in carry, and that if the ECB turns around and monetises French debt, the position gets crushed. So he prefers hedges with positive carry: selling the euro against the Norwegian krone, the Icelandic krona and similar, collecting 150 to 500 basis points, with the side bet that the krone offers some oil protection. What I took from it is that the cost of a hedge decides how long you can hold it, and insurance that bleeds every year tends to get closed out the month before you need it. ![Above and below a zero line stand two bars, the lower one marked as paying away 150 basis points a year and the upper one as collecting 150 to 500 basis points a year.](/figures/hedge-carry-direction-en.svg) Now the parts I can't call. Recession gets confirmed only after two quarters of data, so until then no number on my screen can tell me which phase we are in. And I don't know whether China keeps this semi-closed export regime for the long run. Anas's reasoning — that a congested world serves China while the Arctic alternatives are being built — sounds plausible to me, but it gets tested by construction progress, not by next year's prints. The detail I keep coming back to is from the end of his interview. Anyone glued to X watching explosions and fires in Saudi Arabia, Anas said, would be shocked on arrival in Riyadh: terminal three had just been hit, and the streets were congested, malls full, cafés packed with people watching soccer, in a city of ten million. He took a moment to praise Riyadh as a capital of restaurants. A man back from a war zone spending time on where to eat — that is when I noticed how much of the picture in my head had been fed to me by a timeline. ## Worth a look - MacroVoices #554, produced 22 October 2026, hosted by Erik Townsend and Patrick Ceresna, with Dr. Anas Alhajji and Louis-Vincent Gave; available on the show's site - Energy Outlook Advisors: Alhajji's institutional newsletter and Daily Energy Report - Gavekal's research subscription; the two pieces Gave wrote on France were linked in that week's research roundup email - IEA and OPEC monthly reports, useful for comparing demand forecasts against what actually printed - EIA diesel price and crack spread data, as a read on how tight the middle layer is ## One Thing to Take Away One idea: in anything you depend on, the middle layer is the easiest to overlook and the hardest to rebuild. A broken source makes the news, a broken delivery you feel the same day, but when the processing layer fails, both ends look fine for a while and the damage shows up months later — and it happens to be the slowest layer to build back. Something I tried that you could do today: pick one thing you can't do without, and it doesn't have to involve money — your commute, a friendship, a tool you work with. On a sheet of paper, break it into three layers: source, processing, delivery. Circle the link whose failure would waste the other two. Then ask yourself one question: last month, did I spend any time, any money, or one real conversation on that layer? The first time I did this, the link I circled was the one fixed meal I eat each week with family, and the time I'd scheduled for it last month was zero.