The fraudsters and impersonators have migrated from X to pretty much every social-media platform these days, including Substack, Telegram and even WhatsApp.

Please note: I will NEVER reach …

Read more

Executive Summary

I keep reading, often with a fair amount of irony attached, that oil analysts were wrong to expect a major price shock from the functional closure of the Strait of Hormuz.

That view looks at the wrong price. Commentators remain fixated on crude oil and especially ICE Brent futures. The crisis has arrived further down the barrel, in petroleum products. That is the part of the oil market that consumers actually pay for.

Even crude was hardly calm. Physical Brent, better known as Dated Brent, peaked at $146 per barrel in April 2026. Dated Brent is the global physical benchmark against which almost every crude barrel outside North America is priced.

Add physical differentials of as much as $12 per barrel for grades such as Johan Sverdrup, a medium-sour substitute for crude trapped in the Middle East, and some Asian buyers may have paid close to $160 before freight. Chinese demand then hit the brakes and took some heat out of the physical market.

But crude is not my topic du jour. This episode is about what I call the Great Value Migration in oil: the movement of scarcity value and margin capture from crude production into refining, particularly middle distillates such as diesel and jet fuel.

At the time of writing, front-month European diesel trades at $191 per barrel, or $1,424 per metric tonne. Front-month ICE Brent trades at $96. A barrel of diesel in Europe therefore costs almost exactly twice as much as a barrel of Brent crude.

This is not a European phenomenon. Diesel margins, or crack spreads, are exceptionally strong across the major refining centres.

The diesel crack, usually called the gasoil crack outside North America, measures the difference between the value of diesel and the crude oil used to make it. It is the most important refined-product spread in today’s oil market.

Diesel and its close cousins - jet fuel, gasoil and marine fuel - drive more commercial activity than gasoline. Trucking, shipping, aviation, industry, construction, agriculture and heating all depend on middle distillates.

Part of this strength is structural rather than cyclical.

Since 2020, a persistent shortage of middle distillates has kept cracks above their pre-pandemic norms for long stretches. That shortage explains why refining margins have remained high, why refiners keep pushing throughput and why crude prices can sometimes weaken without delivering much relief to consumers.

Today, diesel cracks sit at the top of the Bloomberg history available to us. In the United States, the front-month spread is roughly $102 per barrel. Add that to WTI at $92 and wholesale diesel costs about $194 per barrel, or $4.62 per gallon. Distribution and retail margins then take the pump price towards $5.60. That is close to the highest level American consumers have ever paid.

So who says the closure of Hormuz failed to cause an oil-price crisis? It did. The war in Ukraine added another layer by attacking the refining system that once supplied a large share of Europe’s diesel. The consequences will reach inflation, interest rates and eventually equity valuations.

The central point is that crude availability and diesel availability have separated. Russia is exporting more crude, in part because it cannot refine as much of it, while China is importing materially less. The residual flow through Hormuz also appears weighted towards crude, although Kpler’s dark-fleet coverage is too weak to measure that leakage reliably.

Together, these factors keep the crude market relatively behaved. The product market has no such cushion: the Middle East, Russia and China are refining less, import demand has not fallen with their output, and the available replacement crudes yield less diesel. The same buyers are chasing fewer suitable barrels, so the adjustment appears in refining margins rather than in crude alone.

Let me show you how the squeeze works, and why it is unlikely to disappear quickly.

Read more

Executive Summary

The Iran war has caused the largest supply shock in petroleum history. Yet even when the fighting was most intense, oil prices only briefly touched the $150/bbl that many analysts and investors, myself included, had predicted at the outbreak of the conflict.

For that, thank a handful of governments. Soon after Iranian munitions made the Strait of Hormuz impassable, trapping 14mbpd of crude inside the Gulf, the petro-monarchs in Riyadh and Abu Dhabi pushed over 3mbpd extra through pipelines bypassing the strait. At the peak, 7.2mbpd flowed around Hormuz, up from under 4mbpd before the war. Washington and Tokyo released a record 2mbpd from emergency stocks. State-led rationing in poorer countries shaved off another chunk of demand.

But disaster would still have struck without decisions quietly taken in another capital: Beijing. Between February and June, China slashed its seaborne crude imports by half, withdrawing 5.5mbpd of buying from the global market, enough to take $30/bbl or more off Dated Brent.1

Let that sink in: more than half the worldwide demand destruction during the Covid-19 lockdowns, when global consumption fell by 9mbpd. With one crucial difference. During Covid, demand disappeared because the world economy shut down. China’s economy today is struggling enormously, but not because it lost access to 5.5mbpd of foreign crude. China chose to stop buying those barrels and, economically, barely felt the difference.

Such resilience is no accident. In November 2003, Hu Jintao warned that “some major powers” were seeking to control the Strait of Malacca, through which four-fifths of China’s imported crude then travelled. Chinese media soon gave the vulnerability a name: the “Malacca Dilemma”. As China Youth Daily put it in 2004, whoever controls Malacca “controls China’s energy route”.

Beijing spent the next two decades building against precisely that vulnerability, raising domestic energy production, electrifying transport, diversifying supply routes and accumulating strategic stockpiles. Under Xi, energy independence went from strategy to obsession.

A system designed with a future conflict over Taiwan in mind has now proved itself in the Strait of Hormuz. That ability to switch enormous volumes of crude imports on and off, at surprisingly little economic cost, hands the world’s biggest oil importer the pricing power OPEC once exercised from the supply side: real power from 1965 to 1989, mostly pretended power after 1990, as we showed in Episode 4.

And with the cartel further weakened by the departure of the United Arab Emirates, one of only three members besides Saudi Arabia and Kuwait holding genuine spare capacity, China’s market power has overtaken OPEC’s - by a country-mile.

As one oil-trading boss put it: “China is the new OPEC.”

Read more

Executive Summary

Today we have a closer look at the Bab el-Mandeb chokepoint and why it matters for global oil and product flows.

The headline is difficult to overstate: Saudi crude exports have crashed. Our August data show crude exports running roughly 74% below the pre-war baseline. That is not simply a shipping-route change. The volume itself is missing. At the same time, some of the barrels that still leave Saudi Arabia are being forced north through Egypt rather than south through the Bab el-Mandeb.

So there are really two stories unfolding at once. First, a Saudi export crash. Second, a Saudi export detour. The second matters because a Saudi barrel that previously reached Asia in 21 days can now require 49 days. The extra freight alone costs upward of $3/bbl, before working capital, insurance and all the additional hassle. Saudi Arabia’s geographic advantage into Asia has suddenly shrunk dramatically.

And that creates the real puzzle. Saudi Arabia is exporting materially fewer barrels, while the barrels that still make it out have become materially more expensive to deliver into Asia. On a shipping map, the obvious conclusion is that Atlantic Basin crude should be flying off the shelf.

It isn’t. That is where this episode gets interesting. A missing Saudi barrel is not automatically replaced by the next available barrel on the water. Crude quality, refinery metallurgy and the willingness of the marginal buyer matter just as much as freight. And right now, that marginal buyer is China.

Deal Or No Deal?

This week, Iran’s Foreign Minister Abbas Araghchi said negotiators are closing in on a temporary maritime transit route for ships entering and exiting the Persian Gulf. Oman said separately, in a post on X (link), that negotiations are progressing in “a positive and constructive atmosphere” and called for a halt on actions in the strait to avoid impacting diplomatic efforts to reach an agreement on navigation.

Are they? From where I sit, nothing has changed — if anything, the recent news flow has confirmed my theory that these two sides are too far apart to ever find a compromise. What do I mean?

First, if we got three lanes — an Omani, an Iranian and an IMO lane, as traffic patterns have established over the past five months and assuming no hot war and shipping attacks — flows could partly or near fully recover. But Iran would effectively lose control over the Strait.

Obviously, that would be an acceptable solution for the U.S. to lift sanctions and, perhaps, even allow the release of frozen funds. It would likely also be an acceptable compromise for the Gulf states, namely the UAE, Saudi Arabia, Qatar, Kuwait and Iraq.

The UAE and the Saudis in particular would likely find an amicable solution with Oman to avoid any fees, while Qatar and Iraq would perhaps be happy to pay Iran some service fee to close this chapter once and for all. Who knows.

However, the Strait has become the life insurance of the regime in Tehran, so it won’t happen. Instead, and according to the WSJ (here and here), Iran insists on asserting control over the Strait. The WSJ wrote:

The parties have agreed on the main points of the draft—which would set up an inbound lane near Iran and an outbound lane near Oman—and have shared it with the U.S., countries in the region and Iran’s top leaders, who still needed to sign off, they said.

According to the draft proposal, which would last for 60 days, ships entering the strait would use the route closest to Iran in coordination with Tehran, and ships exiting the strait would use a route in the Omani waters in coordination with Muscat, the people said.

While the deal would exclude charging ships tolls or fees, Iran might not be prevented from collecting voluntary payments to cover costs like security and search and rescue, the people said.

The above terms leaked on 5 August. The same day, the Hormuz Letter (link) reported that Iran is insisting on a 7% fee on all commercial ships passing the Strait. That demand is not going away — and meanwhile, the IRGC’s Brigadier General Ahmad Vahidi remains in charge of the Strait. Here is what the Hormuz Letter explained:

Iran is demanding a fee of 7% on all commercial ships passing through the Strait of Hormuz, with Chinese and Russian vessels exempt from the fee, while Oman is discussing smaller fees of around 3% and the US wants no fees at all, per Reuters and an Iranian source.

A 7% fee on 20mbpd of oil and products at, say, $75 Brent would mean the regime collects some $40 billion per annum — before counting LNG, LPG and other cargoes — thereby also elegantly circumventing sanctions, as the fee could otherwise not be collected.

Trump won’t approve any of this, but I keep an open mind into the midterms for another “Memorandum of Misunderstandings” — perhaps this time one to which the United States is not even a party, which seems to be what Iran is angling for. But then how would Iran get sanctions relief, among other things?

According to the New York Times (link), Iran’s top security official followed up on the Oman deal rumour by stating that the Strait of Hormuz will not reopen until the U.S. meets a sweeping list of demands:

1. Lifting its naval blockade;
2. Lifting sanctions on Iranian oil, petrochemical and gas exports;
3. Releasing all of Iran’s blocked funds and paying war reparations;
4. Ending the war in Lebanon;
5. Accepting Iran’s right to collect maritime fees on every ship;
6. Withdrawing the U.S. military from around Iran.

Once the U.S. implements these commitments, Iran says, the Strait of Hormuz will reopen — under full Iranian sovereign control.

So forgive me if I don’t hold my breath for a deal to emerge next week. My view is unchanged since March 2026: Iran’s maximalist demands make negotiating a waste of Trump’s time. In reality, there are only two paths to restoring normal oil flows in the Middle East:

a) Regime change;
b) Bypassing Hormuz.

Version (b) remains vulnerable to attacks but is superior to vessel transits through Hormuz. Judging by an interview yesterday, Treasury Secretary Bessent seems to agree with me.

The U.S. administration is playing for time — lower oil prices into the midterms — and has lost faith in a meaningful “deal” with this regime. After that, it will either walk from the conflict and let the oil suppliers bypass Hormuz with pipelines, or it will escalate the war until the regime falls, or a combination of both. As Walter Russell Mead wrote in the WSJ, this war is “irrepressible”.

None of that means Brent won’t trade down on any new headline of an imminent deal, however absurd and detached from reality it may be. But that is precisely the game the U.S. administration is playing here — and for now, it is playing it rather successfully, if WTI is your measure. If gasoline and diesel prices are your measure, well, than perhaps not so successfully.

Now let’s move on to the fundamental part of this episode.

Major Seaborne Crude Oil Flows

Before I explain the Saudi drama that is unfolding right in front of our eyes, let me re-establish some basics.

To think about oil markets in an orderly way, it helps to understand the basic flows of seaborne crude oil.

There are plenty of pipeline flows on land, namely from Russia to China and Eastern Europe, and from Canada to the United States. But the driver of oil markets and oil prices is the 43.6mbpd of seaborne trade. It represents the marginal barrel in the system — and in oil, it is all about the marginal barrel. The last barrel that refineries require or reject is what prices all other barrels.

The route above all routes is Middle Eastern crude to Asia: approx. 14.9mbpd of the 43.6mbpd global seaborne market, or 34% when including intra-regional flows. Of it, China receives — or rejects — 5.5mbpd. In 2015, it was 3.2mbpd.

Asian refiners have specialised in medium-sour barrels from the Middle East: Arab Light and Medium from Saudi Arabia, Basrah Light & Heavy from Iraq, Kuwait Export Crude, Oman Export Blend, and Murban, Dubai & Upper Zakum from the United Arab Emirates. They also source the closest replacements — Urals and ESPO from Russia and, to a lesser extent, CPC Blend from Kazakhstan. I explained in my Crude Quality Matters Substack in some detail why sourcing medium-sour crude matters for Asian refinery yields.

To understand the marginal barrel, however, the seaborne flows of Kazakh barrels to Europe and Russian barrels to India and China are just as important.

And finally, there is the barrel least discussed and often most telling about the state of the oil market: the West African seaborne barrel, known as WAF among oil traders.

Unlike Arab Light or any other Middle Eastern grade, the WAF barrel is auctioned on a daily basis, just like Dated Brent, and the pace of its offtake tells you whether the physical market — not the paper barrel used for hedging, speculation and other purposes — is strong or weak. Keep that in mind; we will need it at the end of this episode.

There is no such price-discovery value in Middle Eastern barrels, LatAm barrels or many others. They are contracted long-term, the contracting is opaque, and only Official Selling Price discounts and premiums give us some indication of what the books at Saudi Aramco et al. look like.

For good order’s sake, and as the above trade flows illustrate: the United States is not crude oil import independent. It imports a total of some 5.9mbpd in June, roughly 4mbpd from Canada via pipeline and the rest via seaborne crude globally, mainly heavy-sour barrels from Latin America (e.g. Venezuela, Guyana, Mexico, Colombia and Argentina) and a few barrels from the Middle East (e.g. Saudi or Iraq).

At the same time, it exports about 4mbpd light-sweet crude from Texas to the world, mainly to Europe and Asia. Those have recently increased to 5.7mbpd, but that was mainly SPR releases that U.S. refiners couldn’t process.

Read more

The Sudden End of The MoU

On 29 June, I explained on X (link) that the Memorandum of Understanding (MoU) was really an exercise in constructive ambiguity — the Kissinger playbook of deliberately vague language so that both sides can claim victory. I called it a Memorandum of Misunderstandings. Why?

On the one hand, the MoU assigns responsibility for reopening the Strait to Iran. It states that “the Islamic Republic of Iran will make arrangements using its best efforts for the safe passage of commercial vessels” and that Iran, in consultation with regional states, will determine the future administration and maritime services of the Strait. Tehran wanted that language because it provides a commercial escape hatch from sanctions while preserving its claim to control the Strait.

On the other hand, the MoU also says that military obstacles must be cleared and commercial traffic should resume immediately. In practice that meant merchant vessels should be able to use the Omani shipping lane to escape the Persian trap at once. That is precisely what happened — for nineteen days.

Further below I will show you the day the corridor opened, how hard it was used, and the day it closed. But I am getting ahead of myself.

The consequence of the ambiguous language is that Iran’s regime lost its de facto control over the Strait, which it clearly realised. That is why, the same week, it attacked a commercial vessel in the Omani lane: to reassert control. In doing so it breached the MoU, prompting the United States to retaliate.

As I have said many times on X, neither side really gives a fig about the MoU. Iran’s regime wants time — time to refinance itself and to preserve leverage through control of the Strait. Trump wants lower oil prices going into the midterms. Once that political objective has been achieved, the odds of the shooting starting again are, in my view, materially higher. Well, it turns out we did not have to wait long. We are back in fighting mode.

The ‘Irrepressible’ U.S.-Iran Conflict

I was always convinced that there is no solution to this conflict other than (a) regime change, which requires political will on the side of the United States government, or (b) muddling through from an oil market perspective until the Strait of Hormuz is no longer the choke point it is today, because a new network of pipelines will have replaced seaborne exports through the Strait.

Neither is currently the plan of action on any side involved, which is why I expect the disruption to the flow of oil and other commodities to have significant duration.

After nearly six months of conflict with Iran in one form or another, the Trump administration still appears to underestimate the nature of the regime it is confronting.

U.S. policy should begin with a basic working assumption: Iran is unlikely to capitulate simply because military pressure increases. The key question is what Washington expects to achieve. Does it believe Tehran will abandon its missile programme, give up its right to enrich uranium, or relinquish what it sees as its strategic position in the Strait of Hormuz?

There is little evidence that the regime is prepared to make concessions on issues it considers fundamental to the Islamic Revolution. On the contrary, Tehran appears convinced that, over time, Washington’s political resolve will weaken first.

If that is the underlying assumption behind the current strategy, disappointment is almost inevitable. Iran is not going to plead for an agreement or surrender its core strategic assets. The likely consequences are continued upward pressure on oil prices and growing frustration in Washington as expectations collide with reality.

Retaliation is not a strategy, and a bombing campaign without a clearly defined political objective is unlikely to produce the desired outcome. If the goal is to force Iran back to the negotiating table in a significantly weaker position, that objective looks increasingly unrealistic.

Policymakers should be clear-eyed about the costs. A strategy centred on coercing Iran into submission risks imposing substantial economic and strategic costs on the United States and its allies without achieving its intended political objectives.

Iran is far from invulnerable. It has significant military, economic and social vulnerabilities that can be exploited. But one reality remains unavoidable: confronting Iran requires time, sustained political commitment, a willingness to absorb casualties, major economic costs and enormous military resources.

As long as the administration is unwilling to bear those costs, extending the current campaign — even if it produces tactical successes — is unlikely to bring Iran to the point of strategic defeat. If anything, a prolonged campaign without the resources or political will to achieve decisive results risks strengthening Tehran’s belief that it can simply outlast Washington.

The great Walter Russell Mead, perhaps the sharpest writer in a very good WSJ line-up, said it best in a recent article when he called this war “irrepressible”. He wrote:

The core problem is that the conflict between the United States of America and the Islamic Republic of Iran is of the kind that Adams’s boss, Secretary of State William Seward, once famously characterized as “irrepressible.” Iran wants to control the energy resources of the Persian Gulf, and there is no local coalition of states powerful enough to block it without American support. The U.S. has believed for 80 years that ensuring the free passage of Middle Eastern oil and gas to world markets is essential to America’s prosperity and security.

Mr. Trump, a man for whom energy dominance is a central foreign-policy theme, is unwilling to give Iran control of the Strait of Hormuz. Iran may make temporary concessions under grave duress, but it will not, perhaps cannot, abandon the drive for control. Both sides seem to see more reasons to keep fighting than to throw in the towel.

I fully agree, and I expect this conflict to last far longer than markets are currently willing to price.

What Can We Learn from the Crossings of the Strait of Hormuz?

We have assessed every vessel crossing of the Strait of Hormuz reported by Kpler since the war began. We now have a sample large enough to report patterns rather than anecdotes.

Kpler gives us the vessel IMO number, the cargo, the loading port and destination, and the route taken. All of it matters for pattern recognition. We join on IMO number and never on vessel name, because names change and a rename would corrupt the attribution.

Yes, Kpler is an AIS tracking service. AIS stands for Automatic Identification System. It is an automatic tracking system used on ships and by vessel traffic services to prevent collisions and to monitor maritime traffic in real time. It also carries important information on the vessel itself, its cargo and its shipping route — which is exactly what makes it valuable for pattern recognition, and exactly why a ship that does not wish to be seen turns it off.

If a ship switches its AIS off, Kpler cannot see it in real time; only a handful of state players can do that. But Kpler updates daily and backfills history as it becomes clear that a vessel crossed dark to avoid detection by the Islamic Revolutionary Guard Corps (IRGC), the military unit that tries to control the Strait and attacks ships. So the dark crossings are in the data — we simply learn about them late.

One honest caveat before we start. Kpler tells us who owns and who commercially operates each hull, but it does not tell us who insures it. Every statement below about ownership is auditable; nothing below is a claim about insurance cover.

Let us look at the data together.

Total Vessel Crossings

Kpler reported 2,143 crossings — dark or with AIS on — for all commodities, meaning crude oil and petroleum products (together, liquids), LPG, LNG and dry cargo such as fertiliser, between 28 February and 24 July 2026.

That is 147 days, so 14.6 crossings per day across all commodities. On 28 February 2026 alone there were 77. Before the war broke out there were well over 80. In other words, the Strait has been running at roughly 19% of its pre-war rate for five months.

Between 19 and 24 July, as both sides intensified their attacks, Kpler identified 34 crossings in total. That is 5.7 per day, or 7% of the pre-war rate.

Even the most risk-friendly commercial shipping managers do not like it when rockets and drones fly over their heads. So the first learning from this excercise is straightforward and hardly surprising: when the war is hot, expect only symbolic crossings — not commodity trading.

Read more

Executive Summary

Russia’s fuel crisis has entered a new phase. What began as a logistics problem has become a volume problem. According to our analysis, refinery throughput has fallen from roughly 5.0mbpd in 2022 to just 2.9mbpd in early July 2026, leaving the country unable to meet domestic demand for gasoline, diesel and other petroleum products.

Gasoline shortages are now nationwide. Diesel has followed. Exports have effectively ceased. Rationing has spread across much of the country, and the Kremlin has quietly shifted from denial to damage control. The scale of the deterioration is best illustrated by our weekly refinery output estimates below.

Naturally, these numbers are a moving target. Some disruptions prove temporary and refinery throughput can recover sharply once damaged storage tanks, utilities or other peripheral equipment are repaired.

That distinction matters because refinery strikes have become part of the information war. A spectacular fire at a remote storage tank on a five-million-square-metre refinery site may dominate social media, yet have little lasting impact. Crude processing can often resume within hours or days.

The opposite is true when critical crude distillation units (CDUs) or downstream processing units are damaged. Modern refineries are built around these processing trains, often with only two or three at a site. Knock one out and a substantial share of the refinery’s capacity disappears. Repairs become measured in months rather than days, during which the refinery is entirely offline - a big deal.

That is precisely Russia’s problem today. Over the past twenty-five years, Russian refineries became heavily dependent on Western technology and specialised equipment, among others, to produce Euro-5 fuels. Replacing sanctioned Western technology with Chinese alternatives has proved difficult, if possible at all, while procurement lead times are always exceptionally long. Refineries are bespoke industrial systems, not mass-produced machinery.

The obvious solution is to import petroleum products. That sounds simple. It is not. Importing an occasional 50-100 thousand barrels per day (kbpd) of gasoline for a few weeks is manageable. Sustaining imports of 200-300kbpd every single day for months, perhaps years, to supply a domestic gasoline market consuming roughly 900kbpd is an entirely different challenge.

At 300kbpd, Russia would need to source roughly 7.5% of the entire global seaborne gasoline trade while simultaneously overcoming sanctions, unfavourable economics, limited export availability, shipping constraints, port logistics and, increasingly, maritime security risks. That is a “champions league”-type undertaking, not minor league.

Belarus and India can provide partial relief. But Belarus is also the most obvious refinery target outside Russia for Ukraine. President Zelensky has already warned Lukashenko, expressis verbis, that Belarusian refineries could be next if they continue supplying fuel to the Russia. China, meanwhile, is unlikely to come to the rescue for reasons I explain below. None can replace the lost domestic production at the required scale.

Imports may cushion the shortages. They cannot eliminate them. The consequences extend far beyond empty filling stations. Fuel is the bloodstream of a modern economy. In a country spanning eleven time zones, every missing barrel eventually ripples through trucking, agriculture, food prices, freight markets, inflation, military logistics and, ultimately, political stability.

In the final chapter, What’s the Trade? (paywall), I explore the most obvious ways to position for this new, structural imbalance in global fuel markets.

This report also builds on my earlier Substack from October 2025, Running on Fumes: How Refinery Strikes and Urals Economics Threaten Russia’s Oil Output – and Its Regime. I recommend reading it first if you want a thorough Russian oil industry warm-up (paywall).

I will keep this report free until the Investment Section. I have spent more than twenty years investing in commodities and believe deeply that capital markets ultimately rest on something far more important than markets themselves: freedom, private property and the rule of law. Those are precisely the values Ukraine is defending today.

If this research helps journalists, OSINT researchers, policymakers, investors, or anyone else better understand what is happening inside Russia's refining system, while separating fact from propaganda, then making it public has been worthwhile. I hope you agree.

If you find it useful, please share it. Reliable, timely data on Russia’s fuel system are expensive to obtain, painstaking to verify and increasingly difficult to assemble. I hope this report makes a small contribution to understanding how the refinery campaign is reshaping Russia’s economy, military logistics and, ultimately, its ability to wage this war.

Enjoy.
Alexander

Share

Unpopular Methods

You do not have to take the President of Ukraine’s word for how bad the Russian fuel supplies situation has become. Take the Kremlin’s.

That same day, Vladimir Solovyov, the Kremlin’s wannabe Goebbels, who has spent four years insisting night after night on Russia’s main state television channel that everything was under control and victory inevitable, finally admitted reality. The shortages are real. Deputy Prime Minister Alexander Novak’s assurance that there is “enough fuel” is a lie. “Everyone,” he conceded, “understands the real situation.”

For once, the propagandist and the petrol queue agreed. Then he showed us how the Kremlin thinks: “War is not about popularity,” he declared. “It is about victory.” In the very next breath he warned his own people that “unpopular methods” were coming. That single sentence is the key to the entire regime.

Listen to it yourself. I believe this is the single most important clip to emerge from Russia since the full-scale invasion of Ukraine began in February 2022, back when it was still called a “special military operation” and supposedly measured in days.

Today Russia is running short of something far harder to replace than missiles: men. Ukraine is reportedly inflicting casualties faster than Russia can replenish them, roughly 35,000 killed or wounded every month. Sooner or later, the Kremlin will have to reach for the most unpopular measure of all: general mobilisation. That, too, is what Solovyov means when he speaks of “unpopular methods”.

Not since reservists were first mobilised in September 2022 have Russians felt so uneasy. On 2 July, the Public Opinion Foundation (FOM), a pollster closely aligned with the Kremlin, reported that 55% of respondents believed their relatives and colleagues felt anxious, up from 40% a year earlier (link).

Russia’s post-Soviet social contract, whereby citizens stayed out of politics and the authorities largely left them alone, has broken down. The war has come home and become everybody’s problem. Drone attacks, once confined to border regions such as Kursk and Belgorod, now reach deep into the Russian interior (link). On 6 July, Ukrainian drones struck Russia’s largest refinery in Omsk, some 2,500 kilometres from the front line - a most remarkable achievement.

The consequence of Ukraine’s refinery campaign over the past sixty days is that petrol is now being rationed across much of Russia. Drivers queue for hours to purchase no more than 20 or 30 litres before being turned away, while many filling stations have simply run dry. Both the rationing and the closures can be verified in real time on gdebenz.ru, a crowdsourced platform launched in late June to help Russian motorists navigate the crisis.

As one would expect in a highly centralised state, however, the shortages are far less visible in Moscow than elsewhere in the country. Mediazona has done the heavy lifting for us (link). It collected and analysed anonymised payment-card transaction data from three of Russia’s largest banks—Sber, T-Bank and Alfa-Bank—to estimate which filling stations were actually dispensing fuel.

Their work shows that between 13 and 14 July, more than 85% of Moscow’s filling stations sold fuel at least once during the preceding 24 hours. Across Russia as a whole, however, only around two-thirds of filling stations were active over the same period.

Meanwhile, the Kremlin’s preferred explanation is “panic buying”. To be sure, once shortages emerge, panic buying amplifies them. That is a feature of virtually every supply crisis. It is not the cause. The shortages themselves are real and reflect a structural deterioration in Russia’s fuel balance, as the data throughout this report will demonstrate.

The Shortages

Mikhail Khodorkovsky, the man who once ran Yukos, Russia’s largest oil company, and whose former Samara refineries are now among those now burning, puts his finger on an important part of the story in a recent post on X [link]. By volume, he argues, this is not yet a catastrophe.

The breakdown, he continues, lies less in the barrels than in the brain: logistics and decision-making. Moving fuel across eleven time zones requires pipelines, rail tankers, strategic reserves and, above all, a state capable of deciding quickly what goes where. That state no longer exists. The people who once knew how to run it have emigrated, been sidelined or simply refuse to touch it. That, he argues, explains why shortages have spread to Chita and Irkutsk, thousands of kilometres from any front line.

Logistics are undoubtedly an issue. Decision-making is too. But they are no longer the binding constraint. Volumes are.

According to our data, Russian refinery throughput has fallen to just 2.9mbpd, down from our 2022 baseline of 5.0mbpd. At those levels, petroleum-product exports are effectively off the table, replaced instead by higher crude exports as Russia ships abroad the barrels it can no longer refine at home.

Against domestic petroleum-product demand of roughly 3.2mbpd, encompassing gasoline, diesel, jet fuel, naphtha, fuel oil and everything in between, the Russian petrostate, as of the first week of July 2026, can no longer satisfy its own market.

Treat the above numbers as a moving target.

We do not, and cannot, know precisely how severe the damage is at every refinery. We simply do not possess the refinery-by-refinery high-resolution satellite imagery produced by firms such as Vantor. Even then, it would often remain impossible to determine whether a crude distillation unit was hit directly or merely caught fire as a consequence of damage elsewhere. Fortunately, we do not have to rely on satellite imagery alone.

Besides our refinery balances, we watch Russia’s exports closely. They provide an elegant back-test. If a drone destroys the CDU, the crude distillation unit, the large distillation tower that sits at the front of every refinery, the plant can no longer process crude oil. Refinery throughput falls and those barrels leave the country as crude instead. We observe exactly that.

Kpler data show seaborne crude exports exploding from 3.4mbpd in January 2026 to 4.6mbpd in June. At the same time, Russian crude discounts have widened sharply, with August-loading Urals into India trading at discounts exceeding US$10/bbl, among the widest since before the Iran conflict.

Taken together, the conclusion is straightforward. Our export back-test confirms that at least 1.2mbpd of Russian crude can no longer be processed domestically. Naturally, exports lag refinery outages by roughly two to three weeks, making the July and August export data particularly important following the latest wave of Ukrainian strikes.

Back to the refinery failure mode and its implications for product processing.

Hit the reformer and gasoline production falls while, assuming the CDU continues operating, naphtha exports rise. Hit the fluid catalytic cracker (FCC) and gasoline production falls further. Hit the hydrocracker and diesel output collapses. Many Russian refineries remain relatively simple facilities with low Nelson Complexity Index (NCI) ratings, making these relationships comparatively easy to identify through changing export patterns.

Two important additions: First, secondary processing units take far longer to repair than a CDU. Their components are substantially more complex and, under sanctions, far harder to replace. Second, restarting crude processing does not mean motorists receive more fuel. Crude throughput may recover while the gasoline and diesel units remain offline for months. Rising refinery runs therefore tell us surprisingly little about what ultimately reaches the filling station. That is precisely why we prefer measuring actual product balances rather than headline refinery throughput.

The verdict is clear. Gasoline is short. It was already heading in that direction before the latest drone campaign. Exports fell from 136kbpd in 2023 to 102kbpd in 2024, then to just 46kbpd in 2025, before collapsing to effectively zero under the 2026 export ban.

The more interesting story is diesel. Russian diesel exports have now collapsed as well. Combining diesel and gasoil, exports have fallen to roughly 400kbpd from around 910kbpd as recently as January. Those exports are now effectively banned too.

Naphtha has proved the most resilient product so far, declining only modestly from 427kbpd during 2025 to 397kbpd in June. Even there, however, exports continue to move lower. The damage is therefore no longer confined to one refinery or one product stream. It is broad-based across the entire clean-product barrel.

Freight Volumes Fall Off a Drone Cliff

To triangulate the severity of the shortages, we performed one final cross-check: rail freight volumes in June.

Over the decades, Russia’s oil industry has evolved into a highly specialised system. Transneft owns and operates the country’s vast pipeline network, stretching well beyond Russia’s borders. It moves crude oil from the oilfields to refineries and, just as importantly, to the major export terminals on the Baltic, Black and Pacific seas.

Petroleum products are different. Pipelines are poorly suited to transporting gasoline, diesel and other refined products across Russia’s enormous distances while handling dozens of different fuel specifications.

Instead, that task falls largely to Russian Railways. It distributes refined products from refineries to regional wholesale terminals, from where specialised road tankers complete the final journey to filling stations. Rail also transports a smaller volume of crude oil.

That makes Russian Rail freight an excellent proxy for the physical movement of petroleum products across the country. IPEM, the Institute for Natural Monopoly Problems, tracks these rail volumes. Its June data complete the picture of a refinery system in distress.

In June 2026, Russian Rail transported 16.7% less crude oil and petroleum products than a year earlier. Since crude typically accounts for only around 15% of these rail shipments, the overwhelming majority of that decline reflects collapsing volumes of refined products. The fuel simply is not there.

Share

Gasoline - A Volume & Behavioural Issue

As at 7 July 2026, Russia’s refinery system was short roughly 320kbpd of gasoline against domestic demand of around 880kbpd.

The number of Russian federal subjects experiencing shortages and fuel rationing has climbed steadily since mid-June. Regional governments and filling station operators responded by imposing gasoline and diesel purchasing limits in an attempt to contain the inevitable panic buying. Here is what that looks like. Social media is full of similar reports, some quite funny, others less so.

According to ISW, approximately 78 of Russia’s 83 federal subjects have now experienced fuel shortages, while authorities in 48 have imposed official gasoline or diesel purchasing limits.

There is no doubt panic buying is now amplifying the crisis. But that is not a bug of a supply crisis. It is a feature of every supply crisis. Once households lose confidence that fuel will be available tomorrow, they buy more today. No government has ever managed to legislate that psychology away.

But the key takeaway is this: The panic is a consequence, not the cause. Russia has a gasoline shortage because it lacks gasoline.

Back-test - GdeBENZ: Where Is The Benzin?

In June 2026, Russian programmer Evgeny Chudov launched GdeBENZ, a crowdsourced “people’s map” of Russian filling stations. Drivers report whether stations are open or empty and estimate queue lengths in five categories, ranging from fewer than five vehicles to more than one hundred.

The site exploded in popularity, reportedly attracting more than 812,000 unique visitors and generating around 100,000 station reports within forty-eight hours, according to Vedomosti Ural. Naturally, the data are imperfect. No registration is required and Ukrainian activists quickly began submitting false reports through VPN connections. This is, after all, an information war as much as a conventional one.

Individual station reports should therefore be treated with caution. The aggregated city-level picture, however, is surprisingly robust. Janis Kluge scraped the data, or at least published the aggregated results, and they tell exactly the same story as our refinery balances. The problem is volume.

Other services are now emerging as well, including Yandex. They would not exist if, as the Kremlin continues to insist, everything were under control. Feel free to try them yourself.

Notice the 30-litre purchase limit shown in the screenshot above. Every station I checked displayed the same restriction. That independently confirms what local media have been reporting for weeks.

We have now launched our own automated data scraping across roughly 200 Russian cities and will report the results as they become available. Whether these services remain accessible from Western IP addresses is another question entirely.

For now, the picture is already remarkably clear. AI-98 is effectively unavailable across most major Russian cities. Diesel is becoming increasingly difficult to obtain as well, despite what various Russian commentators on X continue to claim.

The situation is not “manageable.” It is exceptionally bad.

Diesel - A Volume & Behavioural Issue

Diesel has now become the more complicated story and, arguably, the more dangerous one.

Historically, Russia has been a major diesel exporter. In a normal year, its refinery system produced roughly 1.8mbpd of diesel against domestic consumption of around 0.9mbpd, leaving a comfortable export surplus. That is no longer the case.

So far in July, diesel production has averaged just 1.0mbpd, while our latest weekly estimate stands at only 886kbpd, marginally below domestic consumption. Again, these numbers may prove temporary as damaged refineries recover. But for now, that is the physical diesel supply available to the Russian economy.

Moreover, unlike gasoline, diesel is not a single market. It splits into low- and high-sulfur grades serving very different parts of the economy. Low-sulfur diesel powers most modern trucks, agricultural machinery and heavy equipment. High-sulfur diesel still finds its way into older machinery, locomotives, power generation and marine applications. Understanding that distinction matters because Russia is now running short of both.

As the table below shows, our latest balances indicate that domestic diesel supply has now fallen roughly 30kbpd below demand. Beneath that headline number, however, the picture is considerably worse. Low-sulfur diesel is running a deficit of around 93kbpd, only partly offset by a surplus of roughly 63kbpd of high-sulfur grades that cannot simply be substituted.

Diesel has therefore crossed the same line gasoline crossed weeks ago. The shortage is now real. Manageable for the time being, perhaps, but real nonetheless. That assessment also excludes any seasonal uplift in demand, where our visibility is effectively nil.

The current diesel supply of 886kbpd looks manageably close to average domestic consumption of roughly 920kbpd. Set priorities, economise a little and, at first glance, the system copes. No big deal.

The problem is that diesel demand, unlike gasoline, is highly seasonal. Harvesting is now beginning across southern Russia, precisely when agricultural fuel consumption accelerates.

Once adjusted for seasonality, Russian diesel demand may well exceed 1mbpd during July and August, perhaps even into September. It could be higher still. We simply do not know because the Ministry of Energy stopped publishing monthly consumption data years ago.

On that straightforward assumption, the market could already be running a physical deficit of roughly 200kbpd (see the weekly diesel supply chart below, currently at just 813kbpd).

We assume inventories accumulated during the winter are now bridging much of that gap. But this is Russia. It is equally possible that those inventories were more virtual than physical, as traders maximised export revenues while few anticipated shortages on today’s scale.

Either way, the margin for error has disappeared. Even if inventories prove sufficient to carry Russia through this harvest season, I struggle to see how they will be rebuilt before the next one if refinery throughput remains anywhere near current levels.

Meanwhile, tightening creates incentives. And in Russia, incentives matter even more than in the West. Once diesel becomes scarce, behaviour changes. Traders begin withholding supply. Middlemen ration access. Buyers pay whatever it takes because they have no alternative. Farmers panic. And so on.

Add the usual layer of corruption and coercion that has become a defining feature of Putin's Russia, and the diesel shortage may be very real indeed, not merely psychological or a problem of “logistics”.

Nowhere is that more visible than in the agricultural regions of southern Russia. According to the independent Russian outlet Pole Zreniya (”Field of View”; link), farmers increasingly report that diesel exists, but only for those willing to pay the panic price. One representative of a federal farming association summarised the situation succinctly:

“There is diesel. But they want to sell it dear. Purely on panic. A cartel agreement. They withhold volumes and release them through selected commercial intermediaries. Just like masks during COVID. Just like buckwheat. Just like sugar.”

As expected, the state’s response has been largely theatrical and as impotent as that of every corrupt post-Soviet regime. The Ministry of Energy insists supplies fully satisfy agricultural demand. The Federal Anti-Monopoly Service recommends “responsible pricing”. President Putin says he understands the difficulties and orders delivery schedules to be respected. None of those statements puts diesel into a tractor.

And the economics deteriorate rapidly. Fuel costs that previously amounted to roughly 5,000 roubles per hectare have doubled towards 10,000, while grain prices have barely moved. As one farmer observed, production costs are rapidly converging with selling prices. Yet they continue buying. They have no choice. Harvest waits for nobody.

Russia is simultaneously experiencing genuine physical deficits and the behavioural distortions that inevitably accompany them. The two reinforce one another. Every liter withheld by intermediaries deepens the shortage.

Rick Stainton, a friend on X who spent a decade farming and operating businesses through Zimbabwe’s fuel crises, summed it up best: “It stops being a fuel shortage and becomes a thriving black market full of middlemen.”

Exactly. The tragedy is not merely that farmers pay more for diesel. It is that they gradually stop being farmers. They become fuel traders.

When Size Becomes the Problem, Not the Advantage

Russia’s vastness is usually portrayed as one of its greatest strategic advantages: eleven time zones, almost limitless natural resources and enormous strategic depth. In a fuel crisis, however, those advantages work in reverse. Nearly everything in Russia moves by road across enormous distances, and roads run on diesel. A fuel shortage in a country this large is not merely a queue at the local filling station. It becomes a tax on every kilometre of the economy.

The trucking market is where the damage appears first. According to Kommersant [link], spot freight requests quadrupled within a single week as shippers scrambled for available capacity, while the supply of trucks contracted because drivers increasingly refused routes they could no longer guarantee to complete. Freight tariffs jumped roughly 10% in a week alone, and market participants expect further increases. The binding constraint is no longer price but availability. Carriers are reaching regional fuel purchasing limits, forcing them to reroute, postpone or cancel deliveries altogether.

The structure of the market is beginning to buckle as well. Roughly 70% of Russian road freight normally operates under long-term contracts. Those agreements are now being reopened despite cargo volumes remaining broadly unchanged. Shippers increasingly consolidate multiple customers’ goods into a single truck, while contracts that once ran for a year are renegotiated monthly simply to lock in prices for a few weeks. That is what a market looks like when it loses the ability to plan.

The disruption is spreading rapidly through the wider economy. Public transport, taxis and aviation are all feeling the strain. In some regions taxi drivers are simply leaving the road because they either cannot obtain fuel or would operate at a loss. Freight rates into occupied Crimea have reportedly increased by three to six times, reaching as much as RUB1.2 million per truck. Since transport costs are embedded in virtually every product sold, economists interviewed by The Insider [link] warn that the fuel crisis will feed directly into inflation across almost the entire economy.

Will This Normalise Quickly?

I doubt it.

What strikes me most about modern drone warfare is not the drones themselves but the economics they create. Ukraine no longer needs to destroy Russia’s refining system. It merely has to damage it faster than Russia can repair it. That is a profoundly different type of war.

A chemical engineer recently made an observation on X that caught my attention [link]. Watching footage from Omsk and several other refineries, he pointed out that the facilities were clearly operating normally when the drones arrived. Columns and pipelines were still full of hydrocarbons and the entire system remained under pressure. Under those conditions, even a relatively small drone can trigger secondary explosions and fires that destroy equipment worth hundreds of millions of dollars and take months, sometimes years, to replace under sanctions.

That should not have happened. Every refinery has emergency shutdown procedures. In the event of an imminent attack, operators can isolate processing units, stop hydrocarbon flows and depressurise the system by routing hydrocarbons to the flare. An emergency shutdown is disruptive and costly, but it dramatically limits the damage if a strike follows.

Omsk raises an obvious question. Ukrainian Fire Point FP-1 drones cruise at roughly 200km/h. A flight from the Ukrainian border to Omsk, around 2,500 kilometres, therefore takes well over twelve hours. Russia’s air-defence network does not need to shoot every drone down. It merely has to detect them early enough to warn the target refinery that an attack is coming. Yet the repeated footage suggests exactly the opposite. Refineries continue operating until impact.

Whether that reflects failures in radar coverage, command and control, communications or simple complacency hardly matters. The result is the same. The Russian state is allowing billion-dollar industrial assets to absorb maximum damage from drones carrying warheads that cost only a tiny fraction of that amount.

Ukraine has cracked the code how to hurt Russia were it hurts and Moscow remains completely unprepared. Ukraine is destroying refining capacity faster than Moscow can repair it. Why would it stop? That is why I do not expect this crisis to disappear anytime soon.

Remember Khodorkovsky’s comment about “decision making” (link)? Russia’s brightest minds either left long ago, were imprisoned, silenced or killed. For twenty-six years this regime has systematically rewarded loyalty over competence. That may be an effective strategy for preserving power and loot Russia’s natural resource wealth. It is a terrible one for managing a national industrial crisis and defend the country against drone swarms also far inside Western Siberia.

In All the Kremlin’s Men, Mikhail Zygar describes today’s Russian state not as an administration but as a medieval court, rewarding loyalty and proximity to power rather than competence. That system works remarkably well when distributing patronage. It performs rather less impressively when coordinating rail tankers across eleven time zones. My point? Don’t expect the crisis to suddently resolve unless Ukraine stops hitting these refineries.

Imports to the Rescue?

Khodorkovsky (link) is right that imports can cushion Russia’s gasoline deficit. But I argue they cannot solve it.

To close the gap, Russia needs gasoline that clears four hurdles simultaneously. It must be unsanctioned, surplus to the exporter’s own domestic market, available in meaningful volumes and cheap enough to make the trade commercially viable. Clear all four at once and the list of candidates becomes remarkably short, both on land and at sea.

Let’s start at the top of the market and watch the options disappear.

The world’s two largest gasoline exporters, the EU-28 and the United States, are immediately ruled out. Europe remains by far the largest net gasoline exporter globally, while the United States supplies not only Mexico but much of Latin America. Both have imposed comprehensive petroleum-product sanctions on Russia (here & here). That removes the largest part of the global export market at a stroke.

Even if one entertains the fantasy that Brussels lifted sanctions as part of a future peace agreement, the trade still would not work. Russian domestic gasoline prices are simply too low.

That brings us to the second filter, and arguably the more important one. Russia’s domestic gasoline is cheaper than almost anywhere else. Cheaper than China, the only country with genuine spare refining capacity. Cheaper than India, the world’s third-largest gasoline exporter. Cheaper than Turkey, Pakistan, Malaysia and Mexico. Only a handful of Middle Eastern and African petrostates sell gasoline below Russian domestic prices.

No rational exporter voluntarily ships gasoline into a market priced below its own domestic parity.

The Middle East is often presented as the obvious solution. It is not.

Iran, despite being one of Moscow’s closest allies, exports virtually no gasoline. Decades of sanctions and chronic underinvestment have left its refining system unable to produce a meaningful export surplus. Iraq, Kuwait, Qatar and even the UAE are not major gasoline exporters either. Producing export-grade gasoline is considerably more difficult than exporting crude oil, requiring sophisticated secondary refining units that many hydrocarbon-rich countries still lack, even decades after nationalising the assets of the Seven Sisters.

That leaves Saudi Arabia and Oman. From Moscow’s perspective, those are the best-case scenario. Saudi Arabia exports gasoline from the Red Sea, safely outside the Strait of Hormuz, while Oman sits outside the Strait altogether. Their export infrastructure is among the most secure in the world. Unfortunately for the Kremlin, that is where the good news ends. Combined, Saudi Arabia and Oman exported roughly 180kbpd of gasoline during June 2026. That volume would certainly help Russia. It simply is not available. Both countries already supply higher-paying, lower-risk customers across Singapore, the UAE, Egypt, Jordan, South Africa, and elsewhere. Every buyer pays more than Russia can.

With domestic gasoline prices among the lowest anywhere in the world, Russia has effectively priced itself out of the Middle Eastern gasoline market. It cannot outbid Karachi, let alone Singapore. Of course, the Kremlin may chose to subsidise those imports for its consumers at scale. But that is hardly Kremlin-style. The siloviki prefer it the other way around.

Meanwhile, South Korea, Singapore and Taiwan already supply premium markets such as Australia, New Zealand, Indonesia, the Philippines, Malaysia and even California on the US West Coast. Those are mature, low-risk, dollar-denominated trades supported by decades of investment, efficient logistics and reliable counterparties.

Russia cannot compete. Not on price. Not on payment risk. Not on sanctions. Not on logistics. And certainly not with a clean-product tanker fleet it barely possesses.

None of this means Russia cannot import gasoline. It can. Belarus will help, as will the opportunistic cargoes from India or even a cargo from Nigeria’s new refinery Dangote.

But they will likely prove to small to solve the problem. They merely buy time. The underlying deficit remains exactly where it started: inside Russia’s own refining system.

India - A Painkiller, Not a Cure

Now let us turn to India — the most-cited candidate, for good reasons.

India has a vast refining industry, capable of processing around 6mbpd of crude each day (excluding maintenance), with plenty of export capacity. That is what Russia requires. What is not ideal is that India carries a domestic gasoline price around 20% above Russia’s. So prima vista, the arb is closed.

However, India became the largest single buyer of Russian crude starting in 2022 — mostly Urals — besides China with imports well above 2mbpd and closer to 2.7mbpd in June 2026 alone, according to our real time data.

Therefore, the above set-up almost writes itself: a swap — Russian crude for Indian gasoline. It could let one or several Indian refiners buy at a discount to the already-discounted Urals-to-Brent, more than cover the cost of refining and freight, and hand Russia the finished product it can no longer make at home. Prima facie, an elegant fit.

And the trade is apparently already under way, according to this source which cites that Russia seek to import 400,000 metric tonnes of gasoline per month (111,000kbpd) from India - I assume for the time being, not as a one off. That would fit our June data.

That is roughly a third of India's gasoline exports to divert away from existing clients. A stretch — but perhaps a workable compromise. India's refinery complex runs at around 85% utilisation and may be able to push 5% harder to squeeze out an extra 50kbpd for the export trade, while Russia may be able to outbid some barrels from India's existing customers. After all, it too has a bit of leverage.

And let us not underestimate the pull of it: Indians like a good deal, and a subsidised, captive buyer at the other end of a short-ish sea lane is exactly the sort of trade Indian refiners are built to exploit.

However, let us now count the frictions — because this is not a story of one wall, but of many small ones stacked on top of each other. Death by a thousand cuts kind of thing. Who knows, but here are the issues I see coming:

The first cut is the crude leg itself. To force Russia toward the negotiating table, the US tightened sanctions on any trade in Russian crude, and Urals purchases at Indian refiners duly fell. The local champion, Reliance Industries — India’s largest refiner by a distance — stopped buying from Russia altogether in January 2026 (table below), before resuming it through June; most others kept lifting, but at reduced volumes for a brief period of time.

Then, Trump alongside Israel, struck Iran in the spring of 2026 — functionally closing the Strait of Hormuz, sending crude spiking, and giving us a saga I have documented on this platform. The spike changed Washington’s calculus: to keep barrels moving and prices in check, the administration let more Russian oil flow, and handed Indian refiners — expressis verbis — a sanctions waiver to buy the Russian crude that had piled up in floating storage over the preceding months.

Which sets up the live question: does the Trump administration keep waiving? We do not know, and the topic will stay dynamic. The recent signals point to expiry rather than extension — officials, Secretary of State Marco Rubio among them, have stressed that the exemptions were strictly time-limited instruments, meant only to steady global energy markets through the initial Hormuz shock. But Rubio does not call the shots; Trump does, and he is fickle. So I keep an open mind. But crude sanction waivers is the first “cut” meaningful volumes must survive for Russia to fix its gasoline problem.

Now to the cut I think matters most — and it is really two cuts working together: the law and the ton-miles. Take the law first, because it is the one people miss. There is a world of difference between shipping Russian oil and petroleum products out and shipping gasoline in, and it is the whole ballgame.

Sanctioned exports of Russian oil have a legal safe harbour — the G7 price cap. Carry Russian crude or product out below the capped price ($100/bbl for gasoline, diesel and kerosene; moving policy target) and a Greek vessel owner is doing permitted business.

That is how Dynacom, the Onassis group and their peers legally banked close to $4bn over three years hauling Russian oil, as the FT laid out this week. Legal, if not moral — and arguably moral too, in a roundabout way, since the cap was designed to keep Russian barrels flowing so world prices would not spike; the West wanting the oil moved, just not the Kremlin enriched. An uncomfortable trade-off, because nobody should be feeding a war machine. But it is legal, and it is a machine that exists.

Gasoline flowing into Russia has no such machine or, rather, price-cap channel that blesses it — only exposure: Russian ports, Russian counterparties, the EU’s tightening maritime-services net, and the reputational radioactivity that already saw Ukraine brand those same Greek owners “sponsors of war” for the export trade.

So the mainstream, insured, Western-serviced professional fleet — the fleet that does clean-product business every day — has every reason to look at this carefully and make a long-term risk, not a short-term profit judgement.

The tell is in how the cargoes actually move: Russia’s first jet-fuel imports are being laundered from Japan through South Korea “via a chain of intermediary companies,” and the Indian gasoline moves quietly, according to this article. You do not route a legal trade through cut-outs. So that is a first indication of where this is going.

But does that work at scale? Can Russia load enough gasoline it can buy in India each week to bring it back home? I don’t know. What I do know is that there is no established India-Russia gasoline import trade. Creating one means asking tanker owners to dedicate clean-product tonnage to a corridor with limited backhaul optionality.

That raises the freight required to make the trade economically attractive, and I mean substantially. It currently costs $5.34/bbl (about $3.4cent per liter or about 3-4% on current domestic quoted prices) to ship gasoline from Sikka in India to Chiba in Japan. Affordable. But what if that number goes up 10x because of dark-fleet economics and longer routes?

Back to the volume issue. One line of argument says Russia’s own gasoline and diesel export trade is now drying up fast, freeing tonnage for the import run. But that assumes the two are the same trade, legally. I do not think they are — and that difference is the whole point.

People will quickly point out that there is the dark fleet. True that but let’s look at the details more carefully: In February 2026, some 80% of the total value of Russian seaborne CRUDE oil was transported by ‘shadow’ tankers, while tankers owned or insured in countries implementing the price cap accounted for 19%. However, ‘Shadow’ tankers transporting PETROLEUM products such as gasoline handled only some 37% of Russia’s total volume of products. The remaining volume was shipped by tankers subject to the price cap policy.

Russia has now introduced an export ban for diesel too, not just gasoline, which may free up part of its dark fleet capacity used in the product trade. But Russia also continues to export other products such as dirty products. I guess my answer is, we will have to see how this will play out.

The most obvious line of argument is probably truer: among thousands of middlemen in the shipping industry there will always be an idiot keen to make a few dollars, and if that means scrubbing out a dirty dark tanker to load fuel into it, so be it.

Personally, I think size matters most here. 100kbpd — roughly 11,900 tonnes a day, about 360,000 tonnes a month — is about ten MR cargoes a month in pure volume terms. Onto that goes the ton-mile multiplier. Indian gasoline must sail from Gujarat to Novorossiysk on the Black Sea (about 4,200 nautical miles via Suez), or at times to Ust-Luga on the Baltic (some 7,600) — a round-trip cycle of roughly 40 and 65 days respectively, once loading, canal and discharge are counted.

Stretch each cargo over that long a voyage and you no longer need ten ships; you need on the order of 15 held permanently on the Black Sea run, or north of 20 to Ust-Luga — call it something like 6% of the world’s spot-available MR tonnage, just for 100kbpd (see table below). That number is soft, but the direction is not: the distances are punishing, and they multiply every other constraint.

Because now combine the two. The ton-miles say you need a standing fleet; the law says the obvious fleet — the legal, London-insured, Western spot market — will not take the cargo. So you fall back on dark-fleet tonnage and the middlemen, as explained above. A few illegal loadings here and there. No problem.

But 20 MRs day in, day out, for months? I am less sure. Go to my high case in the table below and the requirement climbs to 40-65 tankers — perhaps as much as 38% of the MR spot fleet. That is a stretch. Move it to bigger vessels? Of course — but the same dynamics apply there too, in terms of availability.

Roughly half of Russia’s seaborne oil already moves on the shadow fleet, the G7-serviced fleet carrying most of the rest. Every shadow tanker pressed into carrying gasoline in is a tanker not lifting Russia’s crude out, at least in theory. The two fleets aren’t all interchangeable but you see my point.

Look, it took Russia 18 months to re-shuffle the crude and product trade back in 2022 when the tanker market was loose. Now it isn’t so much. Back then, the interests of the West and Russia were to give it time to adjust were aligned. Now? Not at all.

I do not believe anybody possesses enough knowledge of the dark fleet to answer that question with confidence but let me put it carefully: I’d be surprised if the global (dark) fleet can easily accomodate a 200,000 barrel per day petroleum product trade for months, let alone 300,000 - which is the direction of travel right now. Time will tell.

And another cut: the economics only work because the Russian oil companes pay for it (let’s call it the siloviki crooks). Imported gasoline lands above Russia’s capped domestic pump price, so Moscow is drafting Tax Code amendments to subsidise the importers — the Kremlin paying its own oil companies to bring in fuel it cannot sell at a profit.

Savour the circularity: India buys record volumes of Russian crude, refines it at Jamnagar, Vadinar and elsewhere, and ships a slice of it back to Russia as finished gasoline, which Russia then subsidises on arrival. The self-styled energy superpower, run by hyper greedy siloviki, paying a premium to buy back its own molecules, cracked in Gujarat. That looks risky to me too.

Today, the Financial Times came out with a remarkable report [link]: an unprecedented wave of Ukrainian drone attacks on Russian shipping during July. According to the paper, Ukraine struck more than 110 Russian vessels in just nine days in the Sea of Azov, making it one of the most concentrated maritime attack campaigns ever recorded.

The objective is straightforward. Rather than chasing individual ships, Ukraine is systematically targeting the logistics chain sustaining Russian military operations in occupied southern Ukraine and Crimea. According to Tomas Alexa, senior analyst at maritime security firm Ambrey, even the Iran-Iraq Tanker War of the 1980s saw roughly 450 attacks spread over seven years. Ukraine has now exceeded a quarter of that total in little more than a week.

“I cannot stress enough how unprecedented this is,” Alexa told the Financial Times. “We have never seen anything so concentrated on a global scale.”

Ukraine’s Unmanned Systems Forces announced that a further eleven Russian vessels had been struck overnight, including five tankers, five cargo ships and a tugboat, bringing the reported total to 116 vessels in just nine days.

“The goal of the operation is to systematically disrupt the enemy’s logistics chain,” the force stated. “Disabling tankers, cargo ships and auxiliary vessels complicates the export of oil and petroleum products, limits maritime transport capabilities, and reduces the enemy’s ability to supply fuel to its forces and occupation grouping in temporarily occupied Crimea.”

This matters because it adds yet another friction to any future gasoline import program. Russia would not merely have to find the gasoline, the ships and the subsidies. It would also have to move those cargoes through an increasingly contested maritime environment.

So can India help? Absolutely. The economics mean that it almost certainly will, at the margin. But margins are exactly the point. India can shave the top off Russia’s gasoline deficit. It cannot eliminate it. Stack the constraints together: uncertain sanctions waivers, the legal grey zone surrounding import cargoes, punishing ton-miles, limited clean-product tanker availability, a shadow fleet already working close to capacity, subsidies required to make the economics work, and India’s own domestic market. None of those constraints is fatal on its own. Together they become formidable.

My base case is that Russia can probably sustain imports in the neighbourhood of 50kbpd, perhaps 100kbpd with determined political support and generous subsidies. Three hundred thousand barrels a day, the sort of volume required if Ukraine permanently disables another third of Russia’s gasoline production, looks far more like a strategic sealift than a commercial trade. At that point the question is no longer where the gasoline comes from. It is whether enough ships, terminals, rail capacity and willing counterparties exist to move it every single day for months, perhaps years.

Could that change? Of course. Sanctions evolve, fleets expand and traders invariably find creative ways to make money. But as matters stand today, imports are a painkiller, not a cure. The gasoline can be produced. The ability to deliver it into Russia, economically, reliably and at scale, is the real scarcity.

China To Rescue?

Which brings us to China, now the world’s largest refining system with roughly 17–18mbpd of installed capacity — about three times India’s and marginally larger than that of the United States. Actual refinery runs, however, currently tell a very different story.

At first glance, the set-up resembles India’s: domestic gasoline prices are materially higher than Russia’s, China imports large volumes of discounted Russian crude under its strategic partnership with Moscow, and plenty of idle refining capacity sits ready.

China is also the world’s only major importer of Iranian crude. The Trump administration temporarily relaxed sanctions during the Strait of Hormuz crisis, but withdrew that waiver this week after repeated Chinese breaches of the Memorandum of Understanding.

Unlike India, however, China remains trapped in a prolonged economic slowdown — for reasons I have discussed extensively on this Substack — and domestic fuel demand is exceptionally weak. That weakness forced Beijing into a remarkable balancing act during the Strait of Hormuz crisis. The objective was simple: keep domestic fuel markets fully supplied while avoiding any meaningful drawdown of strategic crude inventories.

The solution was extraordinary. China slashed seaborne crude imports from around 12mbpd to roughly 6mbpd, cut refinery runs from about 15mbpd to 12mbpd, shifted refinery yields towards gasoline at the expense of naphtha, maximised coal-to-gasoil production to offset the lower refinery output — and effectively suspended petroleum product exports, prioritising domestic consumption over export margins.

The consequence: crude inventories barely moved. The adjustment ran instead through petroleum product inventories — which, unhelpfully, cannot be monitored independently by satellite, since they sit largely in closed-roof tanks.

None of this was particularly profitable. Refinery margins were squeezed between government-controlled retail fuel prices and crude purchased at above US$100/bbl for almost four months. Then again, most of the large refiners are state-owned enterprises (SOEs) and profitability was never the priority for the CCP’s SOEs.

With crude prices retreating, China has lifted seaborne crude imports from roughly 6mbpd to around 8mbpd last week. Where this ultimately settles is too early to call. Beyond renewed Saudi volumes following the brief reopening of the Strait of Hormuz — note that the Strait is again functionally closed — much of the incremental crude remains unidentified in Kpler’s tracking system and is most likely discounted Iranian barrels.

Even so, I do not expect China to return to pre-war seaborne crude imports of roughly 12mbpd anytime soon - certainly not in 2026. The country has already likely achieved substantial fuel substitution with a workable equilibrium not to draw too much crude stocks.

Meanwhile, export quotas - a source of future higher runs - remain tightly controlled by Beijing. Yes, Bloomberg reports that authorities have approved around 1.3 million tonnes of product exports for July — roughly 335kbpd at a typical gasoline/diesel/jet split — to at least three refiners, including private processor Zhejiang Petroleum & Chemical and several state-owned names. But I read such articles as a distraction rather than a hint of where this is going. Let me explain.

China isn’t the clean-product exporter Russia needs. Neither gasoline nor diesel nor jet was regularly on the export menu in size, meaning in the volumes Russia requires. Where it exports, products get shipped to high-margin Asian clients such as Australia or New Zealand. More importantly, not that gasoline and diesel exports were essentially shut down by the CCP at the beginning of the war.

For years, China’s product trade was essentially naphtha, mainly with South Korea’s giant chemical industry, alongside other Asian buyers. Readers who have studied my China episodes on this Substack will recognise the pattern: this is about managing China’s own industrial overcapacity — in chemicals as much as anywhere — not about exporting petroleum products to optimise refinery margins.

Indeed, throughout this decade China has remained a net importer of both clean and dirty petroleum products. All of this suggests the CCP is managing its own energy independence, not the utilisation — and thus economics — of its refining SOEs. As David Baverez’s book argues, the CCP remains in the mindset of a “war economy”, and the only thing that truly matters to it is geopolitical independence.

Moreover, refinery runs are unlikely to return to previous levels unless three conditions are met: the Strait functionally — and sustainably — reopens, export margins turn exceptional, and Beijing allows its refiners to chase them. None of these holds today. For the purpose of oil flows, consider the Strait of Hormuz closed again as of this week — not denying some dark-fleet shuttling, which by definition cannot be measured, only inferred in hindsight.

I have warned repeatedly here, here, here and here, and also in a recent interview (in German), that this administration cannot fix a “deal” with a terrorist regime — and that the Memorandum of Understanding was in truth a Memorandum of Misunderstanding. What do I mean by that?

The MoU is an exercise in constructive ambiguity — the Kissinger playbook of deliberately vague language so that both sides can claim victory. On the one hand, it assigns responsibility for reopening the Strait to Iran. It states that “the Islamic Republic of Iran will make arrangements using its best efforts for the safe passage of commercial vessels” and that Iran, in consultation with regional states, will determine the future administration and maritime services of the Strait. Tehran wanted that language because it provides a commercial escape hatch from sanctions while preserving its claim to control.

On the other hand, the MoU also demands that military obstacles be cleared and commercial traffic resume immediately. In practice, that means merchant vessels use the Omani shipping lane to escape the Persian trap. The consequence: Iran loses its de facto control over the Strait — something the regime appears to have realised over the past 24 hours. That is why it attacked a commercial vessel in the Omani lane: to reassert control. In doing so, it is now in breach of the MoU, prompting the US to retaliate — somewhat symbolically — with overnight strikes against Iranian military assets near the Strait.

But as I have said many times, neither side gives a fig about the MoU. Iran’s regime wants time — time to refinance itself and to preserve leverage through control of the Strait. Trump wants lower oil prices going into the midterms. Once that political objective is achieved, the odds of the shooting starting again are, in my view, materially higher.

Either way, stable crude supply remains a prerequisite for the CCP to raise runs — and that condition remains absent. Brent is back at US$85/bbl, and OVX — oil’s implied volatility — is on the move again, now at 62 against a “stable” reading of 35 or below. The market continues to search for oil’s equilibrium price, and if you read my last Substack on the history of the oil price carefully, you will know it always will. We remain firmly in the Boom-Bust III era.

"But surely China has to help its ally Russia?" Not at all. Beijing has been consistent for years: China's own energy security comes first. Whenever geopolitical uncertainty rises, the CCP becomes more conservative, not less — the priority shifts towards protecting domestic supply, not maximising refinery utilisation or export profits. Time will tell. But from Russia's perspective, China is far from the obvious solution to its crisis.

The Near Abroad Won't Save Russia

Having ruled out most of the seaborne market, let us turn to Russia's immediate neighbourhood: Belarus first, then Kazakhstan. The rest of the former Soviet Union cannot help.

Belarus - the best Address but for How Long?

Belarus is, without question, Moscow’s best remaining option. It is also no more than that.

The country operates two modern refineries, Naftan in Novopolotsk and Mozyr, both ideally positioned to supply western Russia by rail with exactly the gasoline Russia is now short of. Despite what the Bloomberg map above may suggest, both refineries remain operational. Better still for Moscow, Belarusian pump prices sit broadly in line with Russia’s, eliminating the pricing problem that plagues almost every other potential supplier.

Naftan and Mozyr each process roughly 160-180kbpd and together run almost exclusively on Russian Urals delivered through Transneft’s Druzhba pipeline. That gives the Kremlin considerable leverage. There are few more uncomfortable positions in geopolitics than being indispensable to a regime that values power above everything else.

In June 2026 the two refineries produced roughly 117kbpd of gasoline against domestic Belarusian demand of just 24kbpd, theoretically leaving around 90kbpd available for export. Mozyr, in particular, has been pushed well beyond its 2024 operating levels and is effectively running flat out.

Given those numbers, there is little doubt what happened next. The Kremlin almost certainly instructed Lukashenko to divert as much gasoline and, increasingly, jet fuel as possible into Russia. The data suggest he is doing exactly that. According to Pavel Latushka [link], Belarusian gasoline shipments to Russia between 1 and 22 May increased fifty-eight-fold from a year earlier, while jet-fuel exports quadrupled.

The problem is that Kyiv understands those numbers just as well as Moscow. President Zelensky has already warned Lukashenko publicly that continued fuel deliveries make Belarus part of Russia’s military logistics chain. By our numbers, Lukashenko has ignored that warning so far.

That raises an uncomfortable question. How long does Mozyr remain off limits? Unlike Omsk, which lies roughly 2,500 kilometres from Ukraine, Mozyr sits barely sixty kilometres from the Ukrainian border. It is effectively within Ukraine’s backyard. That makes it one of the easiest major refineries in the Russian supply chain to reach, whether by long-range drones or, if matters were ever to escalate dramatically, by other means.

I would therefore be cautious about treating Belarusian gasoline as a permanent solution. It may prove to be the next target instead.

All the Evidence Points to Kyiv
Der Spiegel‘s investigation into the Nord Stream sabotage, English edition, August 2023 (link)

The vessel was almost comically unremarkable. The Andromeda was a Bavaria 50 Cruiser — a production sailing yacht seafarers sneer at as the “Škoda of the seas” — and by 2022 a tired one: dented topsides, a 75-horsepower diesel that rattled like a farm tractor, a broken autopilot, a head that stank. In early September a crew of six — five men and a woman, travelling on forged papers, among them a doctored Romanian passport — chartered her from a small rental outfit in the Warnemünde marina at Rostock. No other sailor on the Baltic gave her a second glance. That was the point.

Over the following weeks the Andromeda wandered the southern Baltic — a stop at Wiek, where the harbourmaster obligingly topped up her fuel tank, a quiet detour south to the Polish resort of Kołobrzeg, and back to Rostock by 23 September, three days before the seabed shook. On the night of 26 September 2022, divers from that unremarkable little yacht went down and blew holes in three of the four steel pipes of Nord Stream 1 and 2. It was the largest act of infrastructure sabotage in modern European history — in the words of Germany’s own federal prosecutors, an attack on “the internal security of the state.” Forensics teams would later find traces of the very explosive used, aboard the boat.

A year on, after investigators from a dozen countries had chased every thread — the yacht, the DNA, the forged passports — Der Spiegel weighed the trail and concluded, flatly, that “all the evidence points to Kyiv.” And here is the tell that matters more than any single clue: nobody wanted to say so. Sweden and Denmark quietly wound up their inquiries without naming a soul; Germany locked its findings away on orders from the Chancellery. Because the answer was radioactive. Had it been Russia, Article 5 of the NATO treaty was in play. Had it been Ukraine, German public support for Kyiv might have collapsed overnight. Silence was the safest option for everyone.

For a year the comfortable assumption had been that reaching a pipeline on the deep Baltic seabed demanded a state’s toolkit — a submarine, naval mines, specialist deep-sea equipment — the sort of thing a country like Ukraine supposedly could not muster. That assumption is precisely why so many casual observers rushed to insist it must have been the Americans, sabotaging the pipeline to sell Europe more of their own LNG.

It was conspiracy rubbish, peddled by armchair commentators who had never served — anyone who had would have known you do not need a submarine to reach seventy-odd metres of water. Any offshore diving professional could have told them the same in a minute: commercial divers work far deeper than that every day, installing and maintaining oil-and-gas equipment. And the motive never held up anyway — America’s LNG is sold by private exporters, not the White House, and Biden would go on to freeze new export approvals altogether.

The truth was cheaper, and entirely of a piece with today’s assault on Russia’s refineries: a handful of divers and a rented yacht. No superpower required. The mission then was the mission now — cut the income to the Russian war machine. Occam's razor.

Kazakhstan - A Distraction

That leaves Kazakhstan.

At first glance, it looks promising. It is a major oil producer, a former Soviet republic and one of Russia’s biggest neighbours. Look more closely and the case quickly falls apart.

The Soviet past does not make Kazakhstan an ally. Since independence, the world’s largest landlocked country has spent more than three decades rebuilding its language, national identity and statehood, fully aware that Russian nationalists, Vladimir Putin among them, have openly questioned whether Kazakhstan should exist as an independent country at all.

Its foreign policy is deliberately “multi-vector”: maintaining commercial ties with Europe, China and Russia alike while becoming strategically dependent on none of them. President Tokayev demonstrated that approach publicly in 2022 when, sitting beside Putin at the St Petersburg Economic Forum, he refused to recognise the so-called Donetsk and Luhansk People’s Republics. Since then, Astana has quietly complied with Western sanctions while carefully avoiding direct confrontation with Moscow. It has played the hand exceptionally well.

Kazakhstan imports no crude oil from Russia. None. It produces and exports its own. The CPC Terminal shipped roughly 1.9mbpd during May 2026, while Kazakhstan’s refineries run entirely on domestic production. The Kremlin therefore holds virtually no leverage over Kazakhstan’s fuel system.

It does retain some influence over crude exports. The CPC pipeline crosses Russian territory before loading at Novorossiysk on the Black Sea. But that leverage cuts both ways.

Moscow would simultaneously antagonise Europe, Turkey and Beijing over export volumes it can ill afford to politicise. In my view, CPC keeps Astana polite. It does not make it obedient.

Nor can Kazakhstan export gasoline in any meaningful quantity. Its four refineries, Pavlodar, Atyrau, Shymkent and Aktau, produce roughly 138kbpd of gasoline against domestic demand of around 115-125kbpd. That leaves little more than the occasional cargo available for export.

Accordingly, Kazakhstan’s Deputy Energy Minister recently dismissed reports that Astana could provide Russia with 50,000 tonnes of gasoline as humanitarian assistance, calling the reports “unfounded” [link].

“We have received no official request from Russia,” he said. “Our priority is supplying Kazakhstan’s domestic market. If surplus volumes exist and there is a commercial opportunity, we will consider it.”

That statement reveals far more than it first appears. Kazakhstan’s principal concern today is not helping Russia. It is stopping Russians from buying Kazakh fuel.

According to local media [link], authorities have established new police checkpoints along almost sixty roads leading into Russia to combat so-called “gasoline tourism”. Border controls have been tightened, while vehicles are now generally limited to a single crossing each day.

That rather sums it up. Russia, the world’s second-largest crude exporter, is increasingly relying on neighbours whose immediate priority is preventing Russians from crossing the border to fill their tanks.

Kazakhstan is not Russia’s solution. It is another reminder of how few solutions remain.

Three Doors, None of Them Good

According to Mikhail Khodorkovsky, the Kremlin has only two “doors” out of the fuel crisis.

The first is the market solution: free gasoline prices and let them rise until demand falls back into line with supply. The second is the administrative solution: coordinate the railways, strategic reserves and oil companies well enough to move fuel across eleven time zones to where it is needed.

The Kremlin has chosen neither. Instead, it has reached for a third option that Khodorkovsky merely hints at: quietly lowering fuel standards and transferring the cost of the crisis onto ordinary Russians - Kremlin style.

Take them in turn.

The market solution is economically straightforward. Let prices find their equilibrium. In a dozen or so regions, gasoline might climb towards US$5 per gallon equivalent, demand would fall, fuel would flow to where it pays best, and within a few months the shortages would largely disappear. Economically, it works. Politically, it is impossible. The Kremlin cannot afford nationwide headlines showing Russians paying European fuel prices while fighting what it insists is a successful war.

Yet, in practice, Russia is already running exactly that experiment, albeit unofficially. The only question is where the market clears. Judging by numerous reports on social media, somewhere around RUB170-250 per litre, roughly US$2-3 and judging by anecdotal evidence posted on social media. That would barely raise an eyebrow in much of Europe. In Russia, it is politically explosive.

Which is why official prices remain anchored far below market-clearing levels. AI-92, the country’s benchmark gasoline grade, sells at an average of just RUB67.72 (US$0.88) per litre at most filling stations. The gap between the official and the market price is exactly where shortages, queues and corruption begin.

The second option is the administrative solution. On paper, it sounds perfectly reasonable. The state simply instructs companies to coordinate rail capacity, release strategic reserves and move fuel from surplus regions to deficit regions. That assumes a government capable of managing a continental logistics network under wartime conditions. If Russia still had such a government, it would not be in this crisis.

The oil industry itself illustrates the point. Khodorkovsky was imprisoned. Vladimir Yevtushenkov was placed under house arrest before losing Bashneft. Lukoil chairman Ravil Maganov died after falling from a hospital window. One by one, capable managers disappeared while their assets were consolidated under Igor Sechin’s Rosneft. A corrupt state can seize an oil company. That does not mean it can run one efficiently.

Catherine Belton reaches goes a step further in Putin’s People. The system built by the siloviki was never designed to administer efficiently. It was designed to extract wealth. Its organising principle resembled the obschak, the common cash pool of organised crime, where loyalty determined access to resources and the distinction between public and private wealth deliberately disappeared. That is the machine now expected to solve Russia’s fuel crisis. It was built to siphon. Not to supply.

Which leaves the third option. Quietly lower fuel quality and let motorists absorb the consequences. It too is field tested, right now. Take Galina from Krasnoyarsk. One morning she stopped at a Gazprom filling station on her way to work and, delighted to find fuel available with hardly any queue, filled thirty litres of AI-92. Eight hundred metres later her engine died.

She was not alone. Roughly twenty-five vehicles reportedly broke down after refuelling at the same station. Some failed before leaving the forecourt. Others managed a few hundred metres. Drivers filmed visibly contaminated fuel flowing from the pumps. Rust, debris and other contaminants clogged injectors and damaged fuel systems. After cleaning, the cars ran normally again. The filling station continued operating. Gazprom had little to say.

Every crisis eventually reveals what a political system was built to do. Russia's was built to extract, not to govern. Faced with a fuel shortage, it does not create more fuel. It raises prices unofficially, rewards corruption, lowers quality standards and asks ordinary Russians to pay the bill.

That is precisely what this system was designed to do. Its response function will not magically change simply because, this time, the problem really matters.

What’s the Trade?

Read more

1. The UAE Left OPEC — Now What?

After fifty-nine years of membership, the UAE walked out of OPEC on 1 May 2026. The decision is about far more than production quotas, recent rifts with Riyadh, or wartime disruption. It reflects structural changes in global energy and in the world economy — and a clear-eyed Emirati view of where the country stands and where it is going.

OPEC was built for oil-dependent states. Abu Dhabi joined in November 1967, four years before the UAE existed as a nation. As a young federation, its economy depended entirely on oil revenue, and OPEC’s framework — collective production management, shared discipline, co-ordinated pricing — offered the expertise, stability and leverage that a small, newly independent country could not generate on its own. But that country no longer exists.

Less than a quarter of UAE GDP is now tied to energy. Aviation, logistics, advanced manufacturing, AI, tourism, and life sciences are its fastest-growing sectors. Just as telling, the Barakah nuclear plant — the Arab world’s first — has been operational since 2020 and now generates clean baseload power across four APR-1400 units totalling 5.4GW of nameplate capacity. For context: Finland has 4.4GW, Switzerland 3.0GW.

Meanwhile ADNOC has built crude capacity to 4.85mbpd and is targeting 5mbpd by 2027 — most of the growth from the Upper Zakum development in partnership with ExxonMobil. Pre-war UAE production sat at roughly 3.5mbpd, the level set by OPEC quota. That is 1.35mbpd of spare capacity sitting idle — on the same order of magnitude as Saudi Arabia’s, and tens of billions of dollars of upstream investment held hostage to a system the country has outgrown.

Leaving OPEC, therefore, is not a commercial calculation. It is a responsibility towards Emirati citizens — to develop the economy beyond oil rather than write down sunk capital deployed for a future the cartel will not allow them to deliver.

The exit was the plan all along. Oil revenue was always a means to an end; the goal was never to be an oil state. It was to build something more enduring — a diversified economy, a knowledge society, a country deep enough and partnered widely enough to thrive in whatever the world becomes next. One can only hope it stays the course, especially in these testing times. Without the UAE as a credible model, which Arab state can the others realistically emulate, for the long-term benefit of all of its people?

The decision may have surprised casual observers and even the Saudis in the moment. But it surprised no long-time oil market insider, and the timing was perfect. Flooding the market — the threat Crown Prince Mohammed bin Salman, whose ambitions and grievances outrun his wisdom, used in March 2020 against OPEC+ member Russia — requires barrels that can leave the Gulf, and right now nobody’s can, Saudi Arabia’s included.

Nor is there genuine uncertainty about what it means. The UAE exit marks the end of OPEC as we know it — Saudi Arabia now loses one of only two production allies with substantial swing capacity. Yet beyond the blow to Riyadh’s global political reach, it will have surprisingly little practical consequence for oil markets, if any at all.

It is worth stating the inconvenient truth upfront. Despite decades of US presidents picking up the phone to Riyadh for every passing oil anxiety — from Nixon in 1973 to Biden’s fist-bump in 2022 — OPEC has never actually performed the price-stabilising role it claimed to perform. As we will show in detail below, the cartel’s monthly realised volatility under its watch has consistently run two to four times higher than the pre-OPEC Twin Cartel Era’s. The reverence Western capitals paid to the cartel for half a century rested on a function it never delivered. The UAE has now noticed.

As our volatility and quota-decision analysis below will show, OPEC has been a poor buffer of price swings and a dismal steward of oil markets for most of its history. Losing its smartest member changes the optics far more than the mechanics — beyond the 1.5mbpd the UAE will gradually add across 2027-28 once the Strait of Hormuz crisis normalises.

Let me explain.

2. Why Stable Oil Prices Matter

Oil sits at the core of the real economy, particularly in transport, where substitution is slow and scale effects dominate. When prices swing, they don’t just move profits around; they disrupt decision-making everywhere. Producers hesitate on long-cycle investments, airlines defer fleet choices, households rethink vehicles, and logistics firms second-guess fuel strategies. The uncertainty translates into underinvestment, misallocation of capital, weaker employment outcomes, and at the system level it propagates into inflation, financial conditions and policy responses.

Oil moves quickly into headline inflation through fuel and transport, then bleeds into the core via freight costs, food prices, airfares and wage negotiations — as the two 1970s oil crises illustrated and as the March-April 2026 US CPI and PPI prints have re-illustrated for anyone who had forgotten. Central banks are then forced to respond to a supply-driven shock they cannot control, trying to hold inflation near 2% while a critical input price is gyrating.

Stable oil prices dampen these shocks, anchor expectations, and give policymakers cleaner signals. The result is more predictable consumer behaviour, steadier wage setting, and more rational capital allocation. Seen in that light, efforts to smooth oil price volatility — including those by OPEC and earlier regulatory regimes — had a logic beyond simple cartel self-interest. Coordinated supply management could in theory reduce boom-bust cycles, limit fiscal crises in producer economies, and lower the risk of broader geopolitical spillovers.

But the historical record, which the rest of this piece walks through, is that dampening one of the largest sources of macro volatility remains illusive in the 21st century. Among other things, the world simply does not possess the supply-side spare capacity, the institutional discipline, or the political alignment to do so.

3. Historic Oil Price Volatility Regimes

In his book Crude Volatility, Bob McNally did the historical heavy lifting every serious oil analyst now stands on. We borrow his framework of distinct volatility eras and amend it where we disagree — his “Texas Era” becomes our “Twin Cartel Era”, because Texas alone never stabilised the world price; the Seven Sisters did half the work. With modern compute, we have also rebuilt the underlying price series and recalculated each era’s realised volatility from monthly, annual, and — where the data permits — daily returns.

Seven eras across oil’s 167-year industrial life: Boom-Bust I (1859-1881), the Rockefeller Era (1882-1911), Boom-Bust II (1912-1934), the Twin Cartel Era (1935-1970), OPEC Price Setting (1971-1986), OPEC Volume Setting (1987-2003), and Boom-Bust III (2004-present, where we currently sit).

The data delivers a verdict that should make every OPEC apologist uncomfortable. Across 167 years, oil’s calmest stretch by a wide margin was the Twin Cartel Era of 1935-1970, when the Texas Railroad Commission and the Seven Sisters jointly throttled supply — σ monthly of just 10.5%, less than half of any other era. Saudi Light held between US$1.30 and US$1.80 for more than a decade. Those were the days of cheap international oil.

The moment OPEC wrestled the pricing pen away in 1971, monthly volatility nearly quadrupled to 41.1% and annual swings reached 46.5% — the most violent regime since the Pennsylvania speculative collapses of the 1860s and 70s. So much for the cartel’s self-styled role as market stabiliser.

OPEC’s forced pivot from price-setting to volume-setting after the 1986 price war brought relative calm (σ monthly 27.4%, σ annual 21.7%), but only because the cartel surrendered genuine price discovery to ICE Brent futures, launched in 1988. And here is the contrarian finding most analysts miss: across 1990-2003, oil’s annualised realised daily volatility averaged 36.5% — essentially identical to the 36.7% that has prevailed in the Boom-Bust era since 2003. The eye sees a sleepy decade in US$20 oil; the math sees a regime as turbulent as anything that came afterwards. Violent percentage moves on a US$20 base look invisible on a linear chart but compound to the same vol math as much larger dollar moves on a US$100 base.

Our current Boom-Bust III, now in its 22nd year, has delivered σ annual 29.5% and σ daily 36.7% — shale’s productivity revolution, the GFC demand shock, and COVID’s negative-price moment together rank this era as the third most volatile in oil’s history, behind only Boom-Bust I and the OPEC Price Setting decade.

The pattern is unambiguous: tightly held supply arrangements (Rockefeller, Twin Cartel) deliver low volatility; competitive supply shocks and OPEC’s heavy-handed pricing experiments deliver chaos. OPEC’s brand as the world’s swing producer survives more on press releases than on the standard deviation of returns.

Let’s look at these eras now.

4. Boom-Bust I (1859-1881)

Right out of the gate of oil’s commercial discovery, the problem of having too much or too little of it established itself.

From December 1859 to January 1862, oil prices in the United States crashed from US$20 to as low as 10 cents per barrel, forcing many oilmen to close their operations. That is a 99.5% bust in two years, and it was the first of many epic price collapses in oil’s history. With each new field discovered, the market was instantly oversupplied — crude oil production exceeding storage and transportation capacities — and prices collapsed before the next discovery did it again. Unlike soft commodities, oil cannot simply be destroyed or left to spoil; it sits in the ground until someone builds the infrastructure to handle it.

From 1859 to 1875, average monthly volatility ran at 111% (and would have been higher with daily data), according to The Derrick’s Hand-Book of Petroleum published in 1889. The greatest stretch of price chaos the commodity would ever know.

5. The Rockefeller Era (1882-1911)

Drillers tried to cartelise themselves out of the chaos. They failed.

The Oil Creek Association of 1861, the Petroleum Producers’ Association of 1869, the various shutdown movements through the 1870s — all collapsed for the same reason any voluntary supply pact does: the more it works, the bigger the prize for the cheater and the holdout. With thousands of independent operators, low barriers to entry, and a new gusher always around the corner, upstream discipline was a lost cause.

Enter John D. Rockefeller. His insight was that the choke-point was not at the wellhead but downstream and midstream. Refining required real capital and operational skill, unlike wildcatting. Transportation — railroads first, then long-haul pipelines — was a genuine oligopoly, unlike oil resources, which in the United States (unlike most places globally) belonged to whoever owned the land. There were countless such owners.

Rockefeller squeezed both. Through aggressive railroad rebate deals, predatory pricing, secret acquisitions and the 1882 Trust structure devised by his lawyer Samuel Dodd, Standard Oil controlled more than 90% of US refining capacity by 1879 and effectively dominated the pipeline grid through the 1880s to 1911. He never directly set the crude price. He did not have to. By adjusting storage tariffs, refusing to lift excess barrels through his pipelines, and at times curtailing his own refinery runs, he could reach back down the supply chain and force drillers to throttle. When pressure failed, he occasionally just cut a deal — the 1887-88 “Great Shutdown” being the most successful voluntary supply cut of the era.

The results speak for themselves. Average annual price volatility collapsed from 78% in Boom-Bust I to 23% under Rockefeller — the first sustained period of price stability the industry had ever known. And it was not just stability for stability’s sake: kerosene retail prices fell from roughly 45 cents per gallon in 1863 to about 6 cents by the mid-1890s. Stable and disinflationary, which is largely why the public tolerated the monopoly as long as it did.

On 15 May 1911, Standard Oil was found guilty of violating the Sherman Antitrust Act and was forced to break up within six months. Rockefeller retired but retained his original share ownership in every one of the new spin-offs — anti-trust did not mean expropriation. The asset-light architecture of the breakup made him wealthier still: the combined value of the resulting Standard Oil descendants rose fivefold over the next decade.

6. Boom-Bust II (1912-1934)

The 1911 dissolution didn’t just remove the stabiliser — it coincided with a complete rewiring of the demand side. Gasoline overtook kerosene as the most prized refined product in 1910, the Model-T was three years old, and oil’s strategic role for navies and armies was about to be cemented by the First World War.

On the supply side, the discovery treadmill never stopped: Spindletop (1901), Glenn Pool (1905), Cushing (1912), Healdton (1913), the Los Angeles basin in the 1920s, and finally East Texas in 1930. Bigger industry, same pathology. The price pattern repeated almost mechanically as it had in Boom Bust I. WWI demand sent mid-continent crude from US$1.30 in March 1915 to US$6.10 in March 1920 — a 452% jump that would not be exceeded for fifty years. Then California’s new fields came on, supply overwhelmed everything, and prices nosedived from US$6.10 at the start of 1921 to US$2.25 six months later. The full boom-bust cycle, compressed into a single business cycle.

Stabilisation efforts ran on two tracks. Domestically, state officials tried to step in where Rockefeller’s trust no longer could: the Oklahoma Corporation Commission issued the first “proration” orders on the Healdton and Cushing fields in May 1914, Texas followed in 1919 with a Texas Railroad Commission given limited authority over physical waste — but the early quotas were toothless.

Internationally, in August 1928, Walter Teagle of Standard Oil of New Jersey, Sir John Cadman of Anglo-Persian, and Sir Henri Deterding of Royal Dutch Shell met for two weeks of scotch, grouse-shooting and negotiation at Deterding’s Achnacarry Castle in Scotland, and signed an “As-Is” Agreement freezing global market shares at 1928 levels. The Red Line Agreement that July had already carved up the former Ottoman oil territories among the same group plus partners. Both were private treaties between the majors. Both held — for a while.

Average price volatility came in around 26% — much better than the 78% of Boom-Bust I, but only because the industry was now bigger, more interconnected, and faster at moving barrels. The structural problem was unchanged: massive new finds, sticky demand, the rule of capture, and no domestic entity with the authority or scale to choke flow.

Stability would have to wait for the Texas Railroad Commission to discover what it was really capable of — which happened sooner than anyone expected when, on 3 October 1930, a 70-year-old wildcatter named Columbus “Dad” Joiner brought in his Daisy Bradford No. 3 well in Rusk County and triggered the East Texas field, the largest oil discovery in American history to that point. It would produce more than 5.4 billion barrels between 1931 and 1980.

7. The Twin Cartel Era (1935-1970)

The problem of price volatility never really went away — to this day — except for one 35-year stretch when oil traded essentially range-bound between US$2 and US$5 a barrel, with annual volatility as low as 10% between 1935 and 1971. McNally calls this the “Texas Era,” which is half-right. From a global perspective it is better described as the Twin Cartel Era: US stability administered by the Texas Railroad Commission (TRC), international stability administered by the Seven Sisters, with Eisenhower’s import quotas welding the two systems together. Each cartel needed the other to work.

The TRC version began with the crisis Dad Joiner had just delivered. The “Black Giant” (the East Texas Field) was a 220-square-mile contiguous reservoir that within nine months was producing one million barrels per day from thousands of wildcat wells, and East Texas crude duly crashed from US$1.00 to US$0.10 a barrel by July 1931. Between 1926 and 1936, wildcatters discovered 4 of the 10 largest conventional oil fields in the United States. That required drastic measures. Texas delivered them.

On 17 August 1931, Governor Ross Sterling declared martial law and ordered 1,200 men of the Texas National Guard’s 56th Cavalry Brigade into the field to physically shut wells in. They stayed there until the courts forced them out, by which point Austin had figured out how to do it with regulators instead of soldiers. The 1935 Connally Hot Oil Act criminalised interstate transport of oil produced over quota, the Interstate Oil Compact extended the model across the major oil states, and from then on the TRC ran a monthly “allowable” system: forecast demand, set quotas, swing US output to fit. It worked for 35 years.

Internationally, the Seven Sisters did the same job through different means. Achnacarry and the Red Line had already laid the framework in 1928 — frozen market shares, jointly-owned concessions, “self-denying” clauses preventing independent development. The Seven Sisters — namely Anglo-Persian Oil Company (BP), Royal Dutch Shell, Esso (Exxon), Mobil, Chevron, Gulf, and Texaco — between them controlled around 85% of global reserves outside the US, USSR and Mexico, and they didn’t need quotas because they owned the wells outright. Their concessions across the Persian Gulf, Indonesia and Venezuela meant they could decide how much each host country produced, post a price benchmarked to US Gulf crude, and call it a market.

Their political reach was on display in 1953, when the CIA and MI6 jointly engineered the coup against Iran’s Prime Minister Mossadegh to reverse his nationalisation of Anglo-Persian Oil Company (BP) and force the British monopoly open into a consortium that included the Americans.

The result was the calmest oil market in history. Annualised volatility collapsed to 10.5% between 1935 and 1971 — roughly 8x quieter than the boom-bust eras either side of it. Between 1948 and 1970, US crude prices held a US$3-5 band. Monthly Arab Light, quoted in the World Bank’s Pink Sheet between January 1960 and December 1971, hovered between US$1.21 and US$1.74 — less than half its American equivalent. For a brief period in history, oil had become a utility — for the benefit of the post-war development of Western economies.

Why two prices? In 1959, President Dwight D. Eisenhower established the Mandatory Oil Import Quota Program to restrict the volume of foreign petroleum entering the United States at 12.2% of US demand. The primary goal was to safeguard national energy security by preventing dependence on vulnerable foreign supplies, while also protecting the domestic oil industry from cheaper global competition — competition the Seven Sisters had been discovering at an unprecedented pace.

It is almost impossible to grasp the market power the Seven Sisters held in their prime. A staff report by the Federal Trade Commission on “The International Petroleum Cartel” in August 1952 explained the scale. They controlled substantially all production and reserves outside the United States and the USSR — what I call the Free World Reserves. By 1944, with the world’s first publicly disclosed country-by-country reserve figures laid before Senator Truman’s committee, the seven companies already controlled 59.3% of Free World reserves while pumping just 17.4% of Free World output. By 1955 their share of reserves was 78.5%, ex United States.

Perhaps the most interesting observation is what I call the spread between their reserve-to-production (R/P) ratio internationally versus that observed for the United States: it peaked at 68.2x versus 12.1x in 1955. The measure is a way of illustrating the Seven Sisters’ spare production capacity mainly in the Middle East, what I like to call “latent spare capacity”. The Sisters could have easily tripled their output within a relatively short timespan with modest capital. But why repeat the mistakes made during Boom-Bust I with every new giant field discovery in the Middle East and with a set rule of global market share in place?

In the four decades they sat astride the global oil concession, the Seven Sisters added roughly 351 billion barrels to their reserve base — a thirty-one-fold expansion against the US industry’s three-fold — while quietly absorbing the largest sequence of giant-field discoveries in human history: Kirkuk (1927), Burgan (1938), Dammam (1938, the discovery well that pointed to Ghawar), Ghawar itself confirmed in 1948, Rumaila (1953), Safaniya (1951). Five of the ten largest oil fields ever found were brought into the Sisters’ concession portfolio in a span of twenty years, and almost everything that still matters to global supply in 2026 — Ghawar, Burgan, Safaniya, Kirkuk, Rumaila — was on their books before the first OPEC quota was ever set.

As McNally documents from the 1928 Achnacarry minutes, the Sisters’ own internal estimate was that “world shut-in production amounts to 60% of the production actually going into consumption.” This was the deliberate management of an oil glut: an era when annual demand growth ran below the pace of new discovery, and where the cartels’ job — both the TRC in Texas and the Sisters abroad — was not to find more oil but to keep what had been found from drowning the price.

The system did exactly what it was designed to do, and the Western consumer was the beneficiary: 35 years of cheap and stable oil underwrote the post-war industrial miracle. But the wealth transfer that followed its collapse deserves to be named plainly. The nationalisations of 1972 (Iraq), 1975 (Kuwait), 1976 (Venezuela), and 1980 (Saudi Aramco) expropriated concession assets that had been developed under contracts running 60 years or longer, financed and de-risked entirely by the majors, with host-government royalties already restructured to a 50/50 profit split since 1948.

By any standard of contract law these were illegal seizures and forced sellouts, and in any other decade Washington or Westminster would have responded as they had in Mexico in 1938 or Iran in 1953. But the host governments were spectacularly lucky in their timing. Britain had been politically broken by Suez in 1956 and economically broken by sterling’s 1967 devaluation; the United States was distracted by Vietnam, the August 1971 closing of the gold window, and a domestic political crisis that would end in Watergate. The pricing baton did not so much pass as get dropped — and the producing countries picked it up.

Three things guaranteed price stability during the Twin Cartel era: (a) annual demand growth was smaller than new discoveries, so there was always a glut to manage; (b) the TRC monopolised US supply while the Sisters cartelised the rest, with Eisenhower’s 1959 Mandatory Oil Import Quotas welding the two systems together; and (c) the duo together held enough global spare capacity to absorb any shock without breaking the price stability.

Three things guaranteed price stability during the Twin-Cartel era: (a) annual demand growth was smaller than new discoveries, so there was always a glut to manage; (b) the TRC monopolised US supply while the Sisters cartelised the rest, with Eisenhower’s 1959 Mandatory Oil Import Quotas welding the two systems together by walling off the domestic market from international competition; and (c) the Cartel duo together held enough global spare capacity to absorb ANY kind of shock without breaking the price stability.

Read more

What’s the News?

As I was preparing the next episodes on crude oil and what lies ahead, President Trump, as so often, interrupted the process with a rapid stream of high-intensity posts on the negotiations with Iran.

According to the latest headlines, the Strait is “fully open”, mines are being removed, and Iran has agreed never to close it again. Markets reacted immediately. Oil and product prices dropped sharply on the announcement.

And yet, scratch the surface and the picture becomes far less clear. The reopening is conditional, traffic is still routed and controlled, and a U.S. naval blockade on Iran remains in place.

In other words, we are not looking at clarity. We are looking at the fog of peace.

Let’s break it down

First, the reopening of the Strait of Hormuz is part of a two-week ceasefire agreed on 8 April 2026. It is contingent on the cessation of hostilities in Lebanon and subject to “technical limitations”, which in practice means a controlled and limited flow of vessels, as outlined in Episode 1. Link to agreement.

Second, even under generous assumptions, the exit rate remains constrained. If we assume up to 10 vessels per day, it would take roughly three months to clear the backlog of some 800–900 vessels of all sizes currently trapped in the Persian Gulf, ranging from crude tankers to LNG, LPG, product carriers, dry bulk and containers.

So far, I cannot observe a meaningful increase in transit activity that would challenge this assumption. The high-frequency data from Kpler, which captures vessel movements across all ship types, supports that view as of the time of writing.

Read more

Crude Quality Matters

There is no such thing as “oil”.

Instead, the world produces hundreds of different crude grades, almost as many as there are oil fields, although some fields yield very similar qualities.

The chart above illustrates this. Think of it as a periodic table of crude. The one below is interactive and works well on a desktop (not so much on a mobile phone), so take a moment to hover over or tap the individual bubbles. Each represents a crude grade, positioned by API gravity on the x-axis and sulphur content on the y-axis. The size of the bubble reflects production volume.

What matters is not just how much oil exists, but what kind of oil it is.

The chart shows 152 crude grades. In reality, there are closer to 700 grades produced across roughly 6,400 extraction areas and more than 2 million active wells. These 152 alone account for roughly 56mbpd of the 85mbpd produced globally each day and are sufficient to make the point.

To begin, focus on the Medium-Sour quadrant, highlighted in orange. You will quickly notice that most of these barrels originate from the Middle East. This is not a coincidence. It is the backbone of the global refining system.

Take Arab Light as an example. It sits squarely in that cluster, with around 1.96% sulphur and an API gravity of 33.3. Production is roughly 4.6mbpd. It is one of the most important barrels in the world.

P.S. Please use the interactive chart on a desktop. It does not scale well on mobile.

So what does this mean? Roughly 75–85% of Middle Eastern crude exports are shipped to Asia in any given year, as shown in the table below. Refinery systems across Asia have been optimised over decades to run on that specific crude diet.

Refineries are not flexible machines. They can be adjusted at the margin, but they cannot be reconfigured quickly to process entirely different crude qualities while maintaining the same product yields.

And this is the key point the chart above illustrates in simple terms. The missing barrels are not just volumes. They are specific grades, predominantly medium-sour crude, produced at scale in the Middle East.

Those barrels cannot be replaced in the quantities required. Not by the United States, not by Latin America, not by anyone.

Read more

Guidance & Thoughts Upfront

Over the coming episodes, I will try to explain, to the best of my ability, different aspects of this Hormuz crisis, share a significant amount of data, most of it hard to obtain and rather expensive, and assess the implications for the global economy. The aim is to stay grounded in probabilities and avoid bias, as markets do not care about our narratives.

As this will take time and there is a lot to cover, each episode will include updated oil balances, as introduced in the Executive Summary. This should allow you to observe the dynamics in real time and act accordingly.

I begin with the Strait of Hormuz itself. For global oil and gas balances, normalising transits through the Strait has no alternative. Without it, a recession arrives in 2026, not 2027. That is unavoidable, as outlined in the Executive Summary. Avoid commentators who describe the situation as “manageable” or “replaceable”.

It is not, at least not in the medium term. U.S. exports cannot solve the problem. They may increase slightly, for instance as SPR barrels are sold due to domestic crude quality mismatches. But fundamentally, U.S. exports do not create new supply. They merely redistribute existing barrels.

Likewise, U.S. shale, often described as a fast-cycle response to price, will remain marginal. The constraint is not time, but geology. The remaining Tier 1 inventory is limited, and the industry no longer has the locations to scale meaningfully where it matters. Think 1-2mbpd. That is highly unlikely. But at least higher prices will further delay the decay of U.S. output. At least I suspect so.

Nor can domestic re-routing solve the problem in the Middle East itself. Beyond the East-West pipeline in Saudi Arabia and limited flows via Fujairah in the UAE, alternatives are constrained, vulnerable and take years to be in place. That is not a near-term response.

And yet, unless the Strait is reopened quickly, closing it will prove a long-term strategic mistake for the regime, should it survive. Why? Because shutting it for more than a few weeks inevitably triggers behavioural change. The pain inflicted by this economic weapon of mass destruction is simply too severe.

That, in turn, accelerates long-term workarounds that marginalise Iran’s influence. New pipeline infrastructure will be developed, and upstream investment will be redirected, at scale.

The 1970s oil crisis offers a clear precedent. It triggered the development of the North Sea basin in the U.K. and Norway, where major discoveries in 1969–1970 led to production ramping from the mid-1970s onwards, ultimately adding millions of barrels per day to global supply.

Venezuela could become part of that adjustment. In 2026, it exports roughly 0.9–1.1mbpd, far below historical levels, and meaningful expansion requires years of investment and infrastructure repair. The world needs to replace up to 14mbpd. This is not the solution, but it is a start.

Venezuela’s relevance lies in crude quality. Its reserves are dominated by heavy and extra-heavy sour crude from the Orinoco Belt, the type of barrel currently missing from the Middle East, although extracting and upgrading it remains capital-intensive and time-consuming .

This Venezuela does not possess “spare capacity” in the IEA sense (higher production in 30 days for 90 days), but it is one of the few scalable sources of the right barrel with potential well above the 2mbpd it exported in 2014. Think 5mbpd and beyond.

And yes, of course, the local situation requires years of investment, some with quick yields, others much longer, depending on the field and infrastructure. None of this happens without rule of law. Without it, capital will not return. I discussed this in detail a few weeks ago. So this too requires this administration to finish the job.

Anyway, in the coming episodes, I will cover a range of topics, from crude quality and barrel counting to unintended consequences for Russian exports, the role of LPG in India, the importance of sulphuric acid and other chemicals, and the impact on LNG and even gold. It will be quite a journey.

Today, we begin with what I call the Persian Trap.

Negotiation Failed, Now What?

Almost a full day of negotiations ended with a terse press conference and no deal. Talks between the United States and Iran began on Saturday afternoon and continued through the night, almost 21 hours in total.

The sun had risen in Islamabad by the time Vice President J.D. Vance emerged to brief reporters: “We have not reached an agreement. We’ve made very clear what our red lines are and they have chosen not to accept our terms.” The press conference lasted less than three minutes.

If the Administration believed that, after 21 hours of negotiations, this regime would give up uranium enrichment, as Vance implied, it misread both the moment and the Iranian power structure, in particular the IRGC.

On a lighter note, they did not spend much time studying the past 80 years of conflict and the rather poor track record of reaching lasting agreements, let alone durable peace.

Below is an extract on the Korean War and how it ended, which I published on this Substack. Not many read it. Nor my other pieces on war and peace. Perhaps now is as good a time as any to prepare mentally for what lies ahead?

The Death of Peace Treaties and Why Wars Often End Without Justice

The armistice signed on July 27, 1953, between the United Nations Command and North Korea reaffirmed the 38th parallel as the border between North and South Korea and created the four-kilometer-wide “Demilitarized Zone” (DMZ). The Military Demarcation Line bisects the “Bridge of No Return,” used for prisoner exchanges until 1968, when the crew of the USS Pueblo crossed it into South Korea.

The Korean War lasted just over three years, from June 1950 to July 1953. While fought on a divided peninsula, it was in fact the first hot war of the Cold War—a violent proxy confrontation between the democratic-capitalist West, led by the U.S., and the communist East, led by the Soviet Union. South Korea was supported militarily by U.S. and UN forces; North Korea by China, which sent hundreds of thousands of troops, and by the USSR, which supplied matériel, strategy, and legitimacy.

Because the war was a proxy clash between two worldviews, there was never a foundation for trust, let alone a peace process. Yet negotiations began relatively early, just over a year into the conflict. Formal talks opened in Kaesong in July 1951 and later moved to Panmunjom. Despite stalemate on the battlefield, they dragged on for more than two years, paralysed by disputes over prisoners of war and the demarcation line.

It took more than 700 meetings and nearly two years to negotiate the agreement—almost as long as the war itself. The signing of the armistice took just 12 minutes, conducted in silence at two separate tables so the delegations did not face one another.

The armistice contained operational provisions, notably prisoner exchanges: Operation Little Switch and Operation Big Switch staged the return of sick, wounded, and eventually all remaining POWs. But beyond this, the agreement was strikingly shallow. After 700 days of talks, the sides had failed to address any core political questions: no peace treaty, no normalisation, no territorial settlement.

Many North Korean and Chinese prisoners refused repatriation, fearing persecution or worse at home. Their resistance became a major stumbling block and symbolised the deeper ideological divide. We see similar dynamics today in Ukraine, where defections and shifting loyalties complicate any prospect of peace.

The Korean settlement was a minimalist framework—closer to a UN resolution than a true bilateral accord. It reflected the smallest possible denominator: not a shared vision, but mutual exhaustion. Violence stopped, but the war never truly ended. To this day, there has been no Korean equivalent of the Tokyo Trials—no reckoning, no justice, no closure for victims. Remember this when you hear calls for a “just peace” in Ukraine.

Since 1953, the armistice has been violated more than 100,000 times. Even during global détente—such as the 1975 Helsinki Conference—tensions on the peninsula deepened. The frozen conflict became a “continuous nightmare”: artillery duels, assassinations, espionage, hijackings, and later, nuclear and missile tests. What was achieved in 1953 was not peace, but stalemate.

And that is why the Korean Armistice matters today: it offers a cautionary lesson to avoid false expectations as the world debates a “just peace” in Ukraine.

From the Iranian side, officials indicated that progress had been made on several points. Two or three outstanding issues, including the handling of the Strait of Hormuz, prevented a breakthrough, according to Foreign Ministry spokesperson Esmail Baqaei. After 40 days of war, mistrust remains deeply entrenched. “Naturally, no one could expect us to reach an agreement after a single meeting. We will continue working to bring Iranian and American positions closer together.”

Baqaei added that the talks took place amid “mistrust and suspicion”, while state media blamed American “excessive demands” for the lack of progress. Both delegations have now returned home without setting a date for further negotiations.

The talks ended in Islamabad on Sunday morning. The path forward remains unclear. Pakistan’s Foreign Minister, Ishaq Dar, urged both sides to adhere to the ceasefire. Iranian officials signalled no urgency to resume discussions. Until the United States presents what Tehran considers a “reasonable agreement”, there will be no change regarding the Strait of Hormuz.

A breakthrough was always unlikely

Reaching an agreement on such a complex and multi-layered dossier over a single weekend was never realistic. The more relevant question was whether a framework for continued negotiations could be established. That would have required both sides to recognise the altered realities after six weeks of war and move away from maximalist positions.

For Iran, this centres on uranium enrichment, a red line for Trump. During the last round of talks in Geneva, Iran was reportedly prepared to make meaningful concessions. Its enrichment facilities in Fordo, Natanz and Isfahan are believed to have suffered significant damage, although it remains unclear how quickly operations could be restored.

The last nuclear agreement between the United States and Iran took two years to negotiate. The more likely outcome now is an interim arrangement that addresses part of the dispute and buys time. Iran would need to relinquish control over the Strait of Hormuz and potentially limit its near-weapons-grade uranium. The United States would need to offer economic relief, for instance by unfreezing Iranian oil revenues held abroad under sanctions.

The dilemma is as much about sequencing as substance. Iran views uranium as leverage, to be traded for a comprehensive agreement, not a temporary one. The United States prefers to resolve the issue early, rather than allow it to remain in Iranian hands during a prolonged negotiation.

Each side expects the other to concede. The Iranians believe they hold the upper hand, having endured six weeks of bombing by far stronger opponents while inflicting significant damage on the global economy. Time, in their view, is an ally. To them, Trump appears increasingly eager to end the war before elevated energy prices weigh on the Republican Party ahead of the midterm elections.

The United States sees it differently. While the war has fallen short of several objectives, it has weakened Iran’s leadership, its armed forces and its economy. A further escalation would likely be more painful, particularly if the United States and Israel intensify attacks on infrastructure and economic targets.

My own view is that the U.S. focus on the nuclear file risks granting Iran excessive leverage. Based on my research, the regime lacks both the financial and industrial capacity, as well as the political time, to rebuild its nuclear programme meaningfully before internal pressures mount. This raises the question whether the nuclear issue is primarily a negotiating tactic: insist on it now, only to accept a more ambiguous compromise later in exchange for reopening the Strait.

Meanwhile, Iran expert Robert Malley highlighted what we outlined in the Executive Summary: time favours Iran, at least in its own perception. The closure of the Strait generates increasing marginal returns for Tehran, while additional strikes deliver diminishing marginal returns for the United States.

The Persian Trap

Another sticking point is the Strait of Hormuz, and what I call the Persian Trap, the Persian Gulf is easy to enter, but not necessarily to leave.

Iran has pledged to keep the waterway open during the ceasefire. In practice, only a handful of ships have been able to transit the Strait. According to U.S. officials, one reason is that Iran itself may no longer be able to locate all the mines it has deployed.

Iranian officials have indirectly acknowledged these constraints. Foreign Minister Abbas Araqchi noted that any reopening would take place “taking into account technical constraints”. That may reflect genuine operational limits, or simply a mechanism to control flows and extract rents.

Following the U.S. and Israeli strikes on 28 February, Iran asserted de facto control over the Strait. As part of the ceasefire, it signalled to mediators that crossings would be limited to roughly a dozen vessels per day, subject to tolls. President Trump, by contrast, has insisted on a full reopening.

In reality, shipowners from Greece, China, India and Pakistan are now negotiating individual crossings with the Islamic Revolutionary Guard Corps. Toll payments reportedly reach up to US$2m for VLCCs. Iranian vessels are exempt.

What we can observe is the routing. Approved vessels are directed north of the standard shipping lane, navigating a narrow corridor between Qeshm and Larak, hugging the Iranian coastline before exiting into the Gulf of Oman.

Pre-war, roughly 130 vessels transited the Strait daily. Since the conflict, flows have collapsed. Between 13 March and 9 April, only 122 crossings were recorded in total, across all cargo types. In other words, one month of traffic now equals a single pre-war day.

The majority of these movements are bulk carriers and tankers carrying petroleum products, primarily linked to Iranian exports into markets such as China and India.

Technically, the northern route is viable, but only just. The Strait’s depth ranges from 30–90 metres, while the northern corridor offers roughly 25–30 metres. That is sufficient for some fully laden VLCCs with drafts approaching 20–30 metres, but it leaves little margin for error. The route works, but it does not scale.

And scaling is the issue. The northern route cannot replicate the throughput of the standard channel. Even under a ceasefire, expect a structural bottleneck across all vessel classes. Crude tankers will not receive priority.

From a purely technical standpoint, I would expect 5–10 vessels per day to exit the Strait during a ceasefire, assuming it holds. The constraint is not only physical, but operational. For vessel operators, this is an execution nightmare.

The backlog already sits at 800–900 vessels. At 10 per day, clearing it would take 80–90 days, assuming no further disruptions. That is a heroic assumption. Entries are effectively zero, excluding Iran’s dark fleet. Without inbound flow, the system cannot normalise.

The core point is simple: until the trapped fleet exits safely, no rational operator will accept entry risk. You cannot price this level of schedule uncertainty. The opportunity cost alone is prohibitive.

The implication is that this crisis extends well into July and likely beyond, potentially bleeding into 2027, regardless of any temporary SPR releases. Tehran will not relinquish the Hormuz lever easily. It is an economic deterrent, cheaper and, in this context, more effective than a conventional weapon.

Example from a Financial Times article:

Jakob Larsen, chief safety and security officer at Bimco, advised members to “await further guidance from the relevant authorities”, warning that transit currently involves “heightened risk”.

The toll regime introduces an additional constraint. Payments may violate U.S. sanctions. As Lars Jensen of Vespucci Maritime noted, shipowners face a binary choice: keep vessels stranded in the Gulf or pay Iran and risk breaching sanctions.

Shipbrokers said insurance also remained a big challenge. “The situation remains clear as mud . . . the main stumbling block for most parties remains a lack of clarity around insurance,” one told the FT.

Anchan, the chief executive of Safesea, said underwriters were requiring shipping companies to provide proof of approval from the IRGC before providing insurance but “no one has a clue where to get this approval from”.

David Smith, head of marine at insurance broker McGill and Partners, said there had been “a huge volume of requests for quotes since the ceasefire”. But he noted that the strait was still “classified as a very high-risk area” by the insurance market’s Joint War Committee, whose designations influence the premiums set by underwriters.

Cargo insurance rates quoted for shipping through the strait have fallen from about 7.5 per cent of the insured value of goods last week to roughly 3.75 per cent on Wednesday, according to one large broker.

Signal Versus Noise

I still expect more vessel transits to be registered in the coming days, perhaps weeks, even in the absence of a deal. That, however, is not what matters. After all, exits from the Strait resemble the kind of escape one would expect after being trapped for six weeks.

For the purpose of re-establishing the regular supplies of oil flows, what exits the Persian Trap during this ceasefire is largely irrelevant for long-term oil and gas balances, or for products such as LPG and fertiliser. It is, in effect, one-off noise.

Yes, if the above calculations hold, and assuming the ceasefire is extended, roughly 80 million barrels of trapped crude and products could reach their intended destinations. But that remains a one-time release.

Read more

Read more

Europe’s Post-Gazprom Era:…

Read more

The Cannibal’s Dilemma

Let us approach mining using First Principles.

Mining is an extractive industry that consumes itself year-by-year. To wit: every tonne of ore mined is one less tonne in the ground. Every ounce poured shrinks the reserve base. Whether the commodity is gold, silver, copper, nickel, iron ore or oil, the principle is the same. Production is depletion and depletion is destiny, unless the company doing the extracting can continuously replace what it extracts.

In practice, however, most miners are chronically bad in the replacing game. Not just in absolute terms, but in the only terms that matter to shareholders: per share.

That is why I keep coming back to Reserves per Share (as first presented here), in the table below measured as ounces of reserves per 1,000 shares. It collapses all the usual excuses into a single number: geology, M&A discipline, capital allocation, and dilution.

Solving for Reserves per Share yields a brutal insight: as a group, senior gold miners went from 97.03oz per 1,000 shares in 2010 to 38.26oz per 1,000 shares in 2024. That’s a long-term, structural transfer of value away from existing owners.

Put differently, mining is an awkward place to hunt for moats.

Most industries are structured in such a way that an exceptional entrepreneur can build a business where time becomes an ally. They leverage their unique strengths to establish a strategic advantage, which can manifest in novel ways.

Consumer brands, like Coca-Cola or WD-40, embed habit and distribution. Buffett calls this “share of mind”. Luxury houses, like Hermès, embed scarcity and pricing power. Media and software platforms, like Alphabet, grow revenues while marginal costs remain stable, which is the Network Effect in action.

Great financial franchises, like Berkshire Hathaway or Svenska Handelsbanken, compound because they reinvest into systems, relationships, and underwriting cultures that scale. In Berkshire’s case, the growing insurance float became the cheap capital Buffett could redeploy at attractive returns.

Even in industrials, the best operators build moats through installed infrastructures, service revenues, technical know-how, and procurement scale. Think Siemens Energy and its gas turbine business.

Buffett has a clean way of putting it: the best businesses make time their friend. Bruce Greenwald explains the machinery behind this beautifully in Competition Demystified. The essential point is: When a moat exists, compounding becomes natural because the business gets stronger as it grows.

In mining, that kind of compounding is rare.

True, miners can enjoy a supply-side advantage; they own a specific resource. Yet, that advantage comes with a string of sunset clauses, e.g. reserves are finite, licences have durations. The tough truth is that even the best ore body is a wasting asset in a legal wrapper. This is why land optionality matters so much; it is the closest thing the mining sector has to extending the time horizon of a good asset. And it’s why many “high quality” miners still struggle to become true long-duration compounders.

Worse, resource companies are price takers, i.e. they do not set the price of gold, copper, or oil. (Pro tip: When someone in a Patagonia vest utters the comfortable (but false) macro line that ‘commodities protect you from inflation’, it’s best to ask which part of the chain has pricing power. The cleanest inflation hedge is a company with durable pricing power, not “real assets” in the abstract. Commodity producers often have the opposite. Input costs can rise while realised prices fall. Commodity cycles do not care about macro theses. Remember that please when you read the next stupid line like it.

Worse still is above-ground risk, that inchoate discomfort that masquerades as political interference, windfall taxes, permitting delays, social licence, or good old corruption. That’s when you build the perfect mine with one set of rules and then watch it all slip away as the rules change because of reasons. Investors learned that that hard way, in Panama, with First Quantum’s Cobre Mine in September 2023. They were reminded again across West Africa, including Burkina Faso’s more assertive stance in 2025.

And Western jurisdictions are not immune. The UK’s 2022 windfall tax episode in oil and gas was a blunt reminder that “rule of law” does not mean “rule stability”. In the United States, the slow-motion theatre of permitting and NIMBY politics (Not In My Back Yard) has become its own form of de facto expropriation. No wonder generalist investors keep their distance.

So, if lasting moats are scarce in mining, then where are they and when do they manifest?

Spoiler Alter: Usually not from geology alone, even when the geology is spectacular. Lundin Gold’s Fruta del Norte is as close to what consider a geological unicorn. K92 has extremely rare growth qualities. Island Gold is a very special asset with grades steadily increasing as the mine goes deeper. Yet, the depletion clock still ticks percussive effect on all of them, and above-ground risk still applies in each case, particularly in Ecuador and Papua New Guinea.

However, when a mining company builds something approaching a moat, it comes from the firm’s winning culture, operational excellence, disciplined capital allocation — all of which repeated over decades. You see it in all of the best operators, without exception. That’s why I have previously emphasised what culture does for names like Alamos Gold, Agnico Eagle, and historically for Endeavour Mining.

Which brings us to G Mining Ventures (GMV).

Read more

Let me give you a quick take on the events unfolding in Venezuela. Given oil is part of my circle of competence, as most of you know well, I will keep this short, sharp, and open (no paywall) for all…

Read more

High quality asset with a clear pathway to first gold

Eskay Creek is a sizeable, high grade open pit project, grading roughly 3.5g/t gold equivalent. The mine plan targets around 450kozpa gold equivalent production in years 1 to 5 and roughly 320kozpa over a 12 year mine life. With AISC below US$900 per ounce in the feasibility study, Eskay sits in the rare category of high margin projects in a tier one jurisdiction.

What makes this story particularly interesting is that it is not a typical greenfield discovery that still needs to invent its infrastructure. Eskay is a brownfield re start with significant legacy assets, including road access, camp infrastructure, and most importantly an existing permitted tailings facility. This materially reduces both capex intensity and permitting loop risk, and it helps explain why management believes first gold in H2 2027 is achievable.

At today’s price, Skeena trades at a material discount to NPV, roughly 0.55 times, despite the project now having a clear catalyst pathway. Over the next 12 months, we expect further de risking to drive a re rating through three primary levers. First, final permitting milestones, including formal consent under the Tahltan framework, which until recently was the dominant political risk but isn’t anymore. Second, continued de-risking of capex and execution, including engineering progress and early works. Third, the potential buyback of the Orion gold stream, which would increase unencumbered leverage to gold and silver prices and improve equity quality.

In short, the project has moved from being a permitting and governance question to becoming an execution and valuation question, and that is where the market typically begins to pay most attention, and not just institutional investors, also the competition.

Eskay Creek, A Mine the Gold Price Left Behind

One of the cheapest and least risky ways to find a new gold mine is not to explore for gold, but to revisit old gold mines that were placed into care and maintenance because the gold price at the time no longer justified continued production. The industry refers to this as brownfield exploration, as opposed to greenfield exploration.

The advantage of this approach is simple. Historical mines come with data. Technical reports, drill logs, production records, and cut off grades already exist. In theory, if one takes a systematic view across closed gold mines and revisits the cut off grades that defined resources decades ago, it becomes possible to identify ounces that were uneconomic then but profitable at today’s gold prices. There is no magic involved. It is mostly arithmetic, discipline, and patience.

This is precisely what Walter Coles did in 2018. His background as a credit analyst at UBS in New York, perhaps, allowed him to see the forest for the trees at a time when gold prices were still relatively uninspiring. Somebody had to do the work, take a view, raise capital, and act. Walter did. Today, he owns roughly 4% of Skeena Resources, which seems well earned.

Walter identified Eskay Creek as a target within the so called Golden Triangle of northwestern British Columbia. The Golden Triangle is one of Canada’s most prolific mineral belts, host to both historic and current world class deposits. Seabridge’s KSM sits within the region. Brucejack, now owned by Newmont, is located there as well, along with its famously challenging high grade geology, which I have mentioned more than once by now.

Eskay Creek’s history stretches back to 1932, when prospector Tom Mackay first recognised the area’s unusual geological potential. Lacking modern exploration tools, Mackay relied on surface sampling and was unable to locate the core of the ore body.

It took more than fifty years for Eskay Creek’s true potential to be revealed. In 1988, a new exploration syndicate returned to the site under the guidance of legendary Canadian geologist Ron Netolitzky. Consolidated Stikine Resources, together with Murray Pezim’s Calpine Resources, initiated an aggressive drilling campaign. After 108 drill holes, it was hole 109 that changed everything, intersecting 27.2g/t Au and 30.2g/t Ag over 208m.

Following permitting, Eskay Creek entered production as an underground mine in 1994. Over the following 14 years, Eskay Creek produced roughly 3.3Moz of gold and 160Moz of silver at average grades of 45g/t Au and 2,224g/t Ag. It became the highest grade gold mine in the world and, by volume, the fifth largest silver producer globally.

In 2008, in the midst of the Great Financial Crisis, Eskay Creek ceased production and was placed on care and maintenance. Gold prices at the time were roughly US$872/oz. In simple terms, an underground cut off grade of 15g/t Au was no longer sufficient for Barrick Gold to justify continued mining. Despite its exceptional grades, Eskay Creek operated as a diesel and propane powered underground mine, which made it a high cost operation by the standards of the time.

After sitting dormant for several years, Walter Coles, through Skeena, secured an option to acquire Eskay Creek from Barrick Gold in 2017. In 2020, Skeena exercised that option and became the sole owner of the asset. In 2019, Walter also acquired the nearby Snip mine from Barrick, located west of Eskay Creek, for zero consideration plus assumed environmental obligations of roughly US$3m. Snip is another brownfield asset, formerly a high grade underground gold mine, and comes with extensive historical drilling and production data.

Below is a short clip of Walter Coles explaining how he approached the Eskay Creek opportunity, in conversation with Rick Rule in June 2024.

Before turning to the specific permitting framework at Eskay Creek, it is helpful to step back and clarify the geological risk differences between open pit and underground mining. That distinction explains not only why Eskay Creek looks the way it does today, but also why I deliberately structured the order of company discussions in this Substack the way I did.

We now turn to geology.

Read more

Mining veteran Sean Roosen’s message for the industry: “Shut up and drill”

Gentle Reminder

Before diving into Osisko Development, a brief note. The easiest way to extract value from this Substack is to revisit my earlier pieces on underground gold miners, namely Alamos Gold, Lundin Gold and K92 Mining.

They outline some technical foundations and investment principles that matter most in this sector. I won’t repeat those lessons here and in future posts. Instead, I build on them. A quick refresher will help frame the analysis that follows and make the nuances of this post far easier to absorb.

Mining Developers and the Lassonde Curve

Before assessing Osisko Development, it is worth revisiting a framework that has guided my thinking for years: the Lassonde Curve, named after Franco-Nevada co-founder Pierre Lassonde.

The Curve is not a law of nature, but it is one of the most useful mental models for understanding how value is created and destroyed during the mining life cycle. It maps the major inflection points in a project’s journey and helps investors distinguish between geological uncertainty, permitting and funding risk, construction risk, operating risk and above-ground jurisdictional risk. Over time, it has helped me identify where genuine alpha in this sector tends to emerge.

The Curve follows a simple narrative arc. It begins with the concept phase, when a company acquires ground, conducts early mapping and sampling, and tries to raise capital on a geological idea. If the rocks respond, the project enters the discovery phase, where drilling defines the scale and quality of the mineralisation. This is often the moment of maximum excitement and maximum volatility, because the market is forced to update its priors in real time.

If a discovery holds together, the project moves into the feasibility phase, where engineers, metallurgists and financiers replace the geologists. Here, PEAs, PFSs and FSs translate drillholes into mine designs, capital budgets, operating cost estimates and economic models. The romance of discovery gives way to spreadsheets, trade-offs and discipline. A strong feasibility study can re-rate a project; a weak one can collapse it.

From there, the Curve descends into the development phase, the part of the cycle that destroys the most value. This is where permitting, funding, inflation, labour shortages and cost overruns collide with reality. Only once construction is complete does a project reach the start-up phase, where the mine produces its first metal and the market finally learns whether the geology, engineering and cost assumptions were accurate. By then, the risks are fewer, but so are the potential returns.

The Lassonde Curve does not tell you whether a specific project will succeed. It is not a substitute for technical and common sense due diligence. But it does remind us that the market consistently misprices risk as a project moves from imagination to execution, and that the sharpest opportunities, and the most painful losses, tend to cluster around the transition points. Understanding where a company sits on that curve is often half the investment decision.

And here is how Sean Roosen positions the company on the Lassonde curve. He basically pitches a re-rating, i.e. construction start to producing gold miner (development stage), along the lines of G Mining Ventures and Artemis Gold, two of the best gold companies in the sector (I will discuss both these companies in December). Is this realistic?

Ticking All the Development Boxes?

At first glance, Osisko Development appears to tick every box required to catch the next upward leg on the Lassonde Curve.

The company controls roughly 4.5Moz of resources at Cariboo in British Columbia, with a risked NAV at spot of about C$3.4bn. On a fully diluted share count of 377 million, assuming the 19.1 million warrants do not meet their strike, this implies a value of C$10.25 per share, or roughly 108% upside from its close on 30 November 2025.

The headline story seems attractive. Cariboo, the company’s flagship, is now fully permitted at the provincial level and fully financed as of Q3 2025. Construction should begin in 2026, with first gold at the end of 2027 and a ramp-up to 4,900tpd some time in H2 2028. The entire effort is led by Sean Roosen, one of mining’s few modern-day legends. For a gold developer in a Tier-1 jurisdiction, this combination of scale, permits, financing and elite leadership is rare.

Viewed against the broader development universe, Cariboo stands out. Rupert’s Ikkari is high quality but remains years from first gold and still sits in the permitting maze. Skeena’s Eskay Creek is not yet permitted but already sports a US$2.2bn valuation. Perpetua, heading toward a federal Record of Decision, is valued at US$2.5bn. Probe Gold’s Novador has just been taken out by Fresnillo. Snowline faces a long and uncertain permitting risk and infrastructure cost road in the Yukon. Coffee, now owned by Fuerte Metals, has major execution questions ahead. Seabridge and Novagold each require US$8bn+ and remain too large for the usual financing channels for now (think dilution). West African names like Montage Gold or Robex–Predictive are interesting but carry jurisdictional baggage and share overhang risk, if you studied it closely.

Note: we will discuss all the relevant names above over the coming weeks.

On that landscape, Cariboo can look irresistible. A permitted, financed, construction-ready Canadian gold project with multi-million-ounce scale and a household-name operator at the helm. For many portfolios, it feels like the project you “have” to own.

And Sean Roosen sells this vision masterfully. In one interview he even jokes that you can order pizza and McDonald’s from the mine site, proximity that most developers could only dream of. He is an outstanding communicator and, more importantly, an outstanding miner. None of that is in doubt.

But now we must turn to what really matters and what will drive the success or failure of this project far more than permits, infrastructure or charisma. This is where the narrative stops being all blue sky. The central issue, again, is geology.

Before walking through it in detail, listen carefully to Roosen himself. At Minute 7:15–7:45 of the interview, he says:

“And we know it’s there, but we didn’t do enough detailed drilling to define it. And we’ve averaged about 1.2 million ounces of overall resource for every kilometre that we’ve drilled over the 4.4 kilometres. So the next plan is, so we’ve got the main 2.2 million ounces. 2 million ounces is there, the measure indicate and inferred. That’s in and around. That is the next plan. The ‘at depth target’ is just to keep extending these vein corridors. At depth it is this orogenic system. Very, very predictable, very knowable.”

This is where the story becomes more complex—and where the technical realities matter much more than the marketing sheen.

Read more

Winter is Coming

High time for an update on my beloved natural gas sector.

What should have caught your attention is the relatively muted price action in TTF and JKM as Europe enters the winter with storage levels around 82%. To be precise, Europe begins the heating season with 86.5 billion cubic meters (bcm) of natural gas held across 19 countries that provide storage for 28 European nations — our relevant perimeter for TTF price formation, including the UK, Switzerland and Norway, but excluding Ukraine and Turkey.

That is a low level in the post-Gazprom era. In fact, it’s the lowest since the post-Covid winter of 2021/22, when Gazprom, already preparing for its illegal invasion of Ukraine, stopped filling its European storages in Germany and Austria. That same winter, Europe entered with just 78.7bcm in storage. By 2022, Gazprom — at the Kremlin’s order — had shut off nearly all pipeline supplies to Europe (flows through TurkStream remain but are marginal). The aim was to fracture Europe’s political, military and financial support for Ukraine. The result was the most violent price spike in TTF history, which dragged Asian markets higher alongside it.

Since then, Europe has filled the 180bcm void left by Russia with higher LNG imports and lower consumption. What it has not done is replace those lost pipeline volumes — neither through Azerbaijan or North Africa nor through a revival in domestic output. In fact, production has kept falling since 2018, thanks largely to “green” policies in the Netherlands and the UK. Both now produce substantially less natural gas than they did seven years ago. Think of Groningen, once a top-ten global field, shut down because of tremors so minor they’d make every Japanese smile.

Europe, in short, now depends on LNG from across the world to keep its households warm through winter. Were Europe to import the same 68bcm of LNG between October and March as last season, and given today’s relatively mild weather forecasts, storages would fall below 10% — an emergency threshold. Pipeline systems must maintain a minimum pressure to function; below roughly 5%, the system risks mechanical failure.

And yet, speculative positioning in TTF — the benchmark European gas contract — has been steadily unwinding. From a February 2025 peak of 292TWh in speculative longs, the market fell to 202TWh around this time last year, when storages were completely full. Now, with storages at only 82%, net speculative longs have collapsed to 46TWh and could turn neutral by November. Who knows.

The Calm Before the Winter Storm?

So why is the market so calm?

Some may argue, intuitively, that Europe is sliding into a technical recession and that industrial demand for gas and power will be weak this winter. But the recession is likely already here nor is it particularly relevant. What drives European gas consumption are households, not industry or power generation. And some 237 million European households — including those in the UK, Switzerland and Norway — will inevitably turn up the heating as temperatures fall.

The correlation is stunningly clear, as we’ve shown time and again. From Edinburgh to Rome, from Amsterdam to Bucharest, consumption rises as temperatures drop below 20 °C, and it accelerates exponentially once we fall below 10 °C. Green transition rhetoric aside, the numbers speak for themselves, month after month since 2019. For those of us who care about data, it’s deeply satisfying.

Because of these correlations, we can now forecast European gas demand with remarkable precision. The method is straightforward: country-by-country consumption models built on four independent European weather forecasts, weighted by their certainty. To this, we add industrial demand (adjusted for economic indicators) and power-sector gas use (adjusted for spark spreads and other electricity metrics). The more we refine it, the more accurate and responsive the model becomes — perhaps even a step towards automated trading execution one day.

We’ll see.

Read more

Why is K92 A Special Situation?

K92 is a scrappy operator’s masterclass.

Under John Lewins, a narrow-vein, undercapitalised asset that most misunderstood—Barrick included—has been turned into a high-margin growth engine through incremental geological intuition, focused exploration, careful engineering and disciplined reinvestment. Restarted in 2016 and declared commercial in 2018, it is no longer a promising teenager; it is a tier-one athlete hitting its stride.

What sets it apart is not a slide deck but a production profile and a cost base. K92 is moving into the 400,000 oz Au-equivalent class, with one of the lowest unit costs in our universe—thanks to a superior cost structure (grid power, access, paste backfill, mechanisation, geometry, labour costs, etc.) and copper and silver by-product credits that the market has barely priced.

This is also an execution story, not a promotion: grades and continuity in a “spaghetti soup” of veins—but really golden curtains, as I explain below—access opened and mechanised; and, crucially, the best near-mine resource endowment in our universe, much of it still to be declared as resource over the coming years by one of the largest exploration teams in gold mining. Huge.

All of that comes without financial leverage and with low operational gearing, courtesy of the superior cost structure. Just as important, the current growth leg carries minimal, if any development risk: the key expansion building blocks such as the new processing plant, paste plant and underground development access are already in place. Very special.

The kicker? John Lewins has executed this brownfield revival with the least equity dilution per ounce we know of. K92 may yet prove the most successful mining growth story—across commodities—with the lowest dilution per ounce produced from start to finish. A tier-one athlete on a value multiple, hiding in plain sight.

How Kainantu Found its Pulse Again

When K92 Mining bought the abandoned Kainantu mine from Barrick in 2014, the consideration was intentionally light up front and heavy on success: US$2m cash, a contingent US$60/oz on the first 1Moz produced, and a 2% NSR in Barrick’s favour. It was a classic “pay as you prove it” deal.

The original operation started construction in 2004 and came into production in 2006. Barrick acquired the mine in December 2007 from Highlands Pacific. Under Highlands, output was 27koz in 2006 and 32koz in 2007; under Barrick it managed a mere 10koz in 2008—69koz in total—before the mine was placed on care and maintenance in early 2009. Barrick’s rationale was straightforward: narrow-vein geometry and high operating costs made the mine uneconomic at prevailing gold prices.

Reading Barrick’s 2007 announcement with hindsight and the point is obvious: the acquisition was never really about the struggling mine; it was about the 2,900 km² land position.

By 2015, when John Lewins and team took over, Kainantu wasn’t a turnaround so much as a rescue—albeit one with roughly US$100m of sunk development to access Irumafimpa and Kora and a small processing plant. After seven years of care and maintenance, the tunnels were flooded, the plant was idle, and the highlands jungle had begun to reclaim the site. Not an easy brownfield set-up.

Before the details, look at the site yourself: three Google Earth history screenshots (September 2015, July 2022 and August 2025) tell the story of a mine being brought back to life.

What looked like failure to a major with a 30,000-foot portfolio view in 2009 was opportunity to a team that understood how to treat a small, high-grade system. A firmer gold backdrop helped: gold averaged US$1,160 in 2015, US$1,393 in 2019, then broke out to US$1,770 in 2020.

Crucially, the government renewed Mining Lease ML150 and no state equity was imposed. Kainantu carries no 10% free-carried interest at the mine or company level. Papua New Guinea participates via royalties and taxes, not share ownership: a 2% royalty (NSR/FOB basis), the 0.5% MRA levy on mine revenue, and 30% income tax, alongside landowner agreements.

Important Risk Assessment:

The statutory right for the State to take an equity stake can apply to new mining leases in Papua New Guinea, but K92’s current ML150 carries no such provision. Politically, alignment still comes from operating performance, stable relationships and benefits flowing to landowners—not from a carried interest. But a possible development of Blue Lake within the Exploration Licence 470 (EL 470) may change that conversation. Also, to our knowledge there is no Standstill Agreement in place to legally protect (in practice everything can change) the current status quo, similar to Fruta del Norte, for example.

Geological Comparison

Before drilling into Kainantu’s operation, it helps to anchor the deposit against the underground case studies we have already covered—Island Gold, Young-Davidson and Fruta del Norte—using the same simple comparison table to keep the contrasts sharp.

Kainantu is a high-grade, narrow-vein epithermal system with pronounced variability in thickness both between and within the principal lodes, yet with consistently very good continuity along strike and down dip. In round numbers, working widths sit in the narrow-to-moderate camp, with true thickness commonly in the 1–15 meters range and resource grades in the mid-to-high single digits for gold.

That geometry puts Kainantu near to the Island Gold deposit, not bulk-tonnage underground archetypes: not as high-grading as the best Island Gold panels, but generally thicker and with comparable continuity—exactly the sort of profile that matters for stope design, dilution control and mine scheduling.

Set against Young-Davidson, the differences are first principles. Young-Davidson lives on volume rather than grade: orebody thickness on the order of 80–120 meters enables bulk mining at low single-digit grades, a strategy that optimises unit costs through scale but depends on sheer tonnes moved.

Kainantu extracts value through selectivity. Mining widths are measured in metres, not tens of metres; production leans on long-hole stoping and cut-and-fill to preserve head grade and contain dilution; development is targeted to chase discrete, steeply dipping lodes rather than to sweep a broad, homogeneous panel. It is a different cost and risk equation.

Fruta del Norte, by contrast, is the unicorn that combines thickness with high grade and excellent continuity. A deposit capable of 100–150 meter mining widths while still carrying resource grades around 7g/t for gold is rare, and it compresses both operating and capital intensity per ounce. Kainantu cannot claim that blend of bulk width and high grade; what it does have is a disciplined narrow-vein architecture that delivers continuity over kilometres of strike, allowing value to compound by repeating well-understood stoping patterns at scale.

The Golden Curtains of Kainantu

When geologists speak of a narrow-vein system, most of us non-geologists instinctively picture our own human blood vessels — thin, branching, fragile. At least that’s what happens in what Daniel Kahneman calls System 1 in his brilliant book Thinking, Fast and Slow: when certain words are spoken, intuition takes over.

But in this case, intuition deceives us. That mental picture is a misleading one as it doesn’t do this geological system any justice. The “narrow veins” of the Kainantu Mine — the Kora and Judd systems — are not fine threads of gold, but rather steep, towering curtains of quartz, sulphides, and metal running deep through the mountain.

Technically, the term narrow-vein system is of course correct. But “narrow” refers only to thickness — typically one to fifteen metres wide. Imagine a bird, or avatar with laser vision, flying over the Kainantu mountainous region able to look through solid rock. From his vantage point high above, he would see two glowing lines buried in stone — perhaps like spaghetti laid out beneath the surface.

These “spaghetti” (lines in red above) would be astonishingly long — literally kilometres in what geologists call strike length (strike, by the way, has nothing to do with labour disputes or lightning bolts here, more with a poorly developed English vocabulary). Our laser-eyed avatar would see these narrow veins stretching for two to three kilometres, their thickness expanding and contracting from one to fifteen metres as they undulate through the rock. Huge dimentions.

But the real beauty of Kainantu would only reveal itself to our avatar if he could slice the mountain open into its cross section like a cake, right along the lines of those veins. Inside, he would find mineralized curtains of gold, copper, silver, and quartz hanging down for hundreds of metres — what geologists call their vertical extent (or down dip). Johan Wolfgang Goethe wouldn’t approve of such simple words, but it is what it is.

So don’t imagine tangled spaghetti in stone. Imagine instead long, folded golden curtains, descending vertically through the crust — continuous lodes where mineral-rich fluids once surged upward, cooled, and crystallized. In some places, these curtains swell — forming dilatant zones, where the rock cracked a little wider and trapped more quartz, sulphides, and gold. Those are the rich pockets — the geological knots of fortune, with a mix of gold and copper clitter.

Kora and Judd run side by side, like twin arteries of an ancient geological heart, shaped by the same tectonic rhythm and fed by the same molten pulse. Nor is it a two-vein curtain of Kora and Judd. The 2024 technical report sets out at least nine other golden curtain siblings: Judd (including Judd South), Kora (including Kora South), Karempe, Irumafimpa (Barrick’s original production target), Mesoan, Mati, Bona Creek, Maniape and Arakompa. That is the visible family; the geophysics hints at more.

This breadth flows from John Lewins’s pragmatic approach to exploration. When he and his team reopened the mine from Barrick in 2017, they did not declare reserves. They kept disclosure at resources, fully aware that spending scarce capital to engineer reserves in a narrow-vein system—before underground access and close-spaced drilling—would waste time and money to appease generalists who would not grasp the geometry anyway.

Instead, they restarted with limited but well-aligned capital and a clear aim: let production fund exploration. Cash flow has since underwritten the drill-out of multiple targets, with Arakompa the current standout—growing rapidly and not yet even a declared resource. The company will bring in two more surface diamond rigs to arrive in late Q4 2025, supporting a significant ramp-up in its exploration. That is land optionality at work: acquire a district, open the cash register and let geology compound. It is, within the rules, the clean way for you to front-run Wall Street too.

The menu is already a small but excellent meal—and it gets better. The technical report points to the likely price of this spaghetti soup, and it is large. A helicopter-borne electromagnetic survey, flown in November 2021 by Expert Geophysics Limited, mapped the property at depth and outlined it in a technical report. The programme delivered a far greater depth range than was technologically possible before. Think of it as geology’s analogue to improved seismic tech in oil and gas: higher fidelity, deeper penetration, better targeting. In short, a sharper, bigger image where a smaller, blurred one once stood.

Data were collected along east–west lines nominally spaced at 200 metres, with north–south tie lines at 2,000 metres. The conductivity maps light up the known lodes and, more importantly, illuminate many more noodles plus a few likely bolognese pots. The grey “interpreted structures” (below in grey) are new spaghetti strands yet to be drilled and potentially converted into resources, and there are loads of them. The yellow-circled zones mark porphyry targets—the sauce. We will return to those.

The 2024 report says it plainly. The airborne deep-penetrating geophysics over the entire property shows strong correlation between known deposits and conductive bodies. A swathe of high Apparent Conductivity extends from Kora and Judd into EL470 for several kilometres, and a longitudinal section shows the same highly conductive corridor north and south of the Kora mining centre.

The conclusion is unambiguous: there is extensive untested strike potential to the Kora–Kora South and Judd–Judd South vein systems beyond the A1 porphyry to the south-east for several kilometres. The programme also identifies new vein and porphyry targets on K92’s licences and highlights known mineralised porphyries, including A1, Blue Lake, Tankaunan and Timpa.

Now let that settle and glance at the reserve and resource table for Judd and Kora alone. Gold resources and reserves for those two veins total 6.38 Moz. Add silver and copper and the package equates to 8.8Moz AuEq, supporting a 27-year life at 240,000oz per annum, which the company is set to deliver from 2026. These veins are not bounded; both remain open at depth and are likely to lengthen as development advances and provides ever cheaper underground drill platforms. That is how narrow-vein systems behave when the lodes are continuous and the structure is coherent.

Now layer in the rest of the vein systems already mapped—and the additional structures now outlined by geophysics—and the size of the prize comes into view. The district potential is very large. A 20Moz gold endowment or more is within reasonable geological expectation.

Have I misplaced my marbles? Hardly. The number is defendable. A 20Moz district endowment sits well within reasonable geological expectation—strictly as an exploration target, not a resource—and the distinction matters. Here is the sanity check, built only from what we can see and measure today.

Time will decide what the drill bit delivers, but the investment point is balance. Geologists are paid to be conservative; their craft is humility before geology, because Mother Nature punishes hubris without appeal. At the other pole sit the Silicon Valley promoters—Elon, Cathie and friends—who can conjure trillion-dollar valuations from a sketch and a slide on “TAM expansion”. Hand them land optionality and they would price Kainantu at 100 Moz booked by tomorrow morning. Neither instinct serves capital well. Too conservative and you rack up silent opportunity cost; too promotional and you incinerate cash.

In my humble opinion, our job as resource investors is to sit between those extremes: recognise that land optionality has real, compounding value when the geology is predisposed; insist on disciplined capital allocation and let production pay for proof; refuse to badge exploration upside as reserves or resources until the drill core says so; but allocate capital where land optionality is waiting to happen. Then let time be your friend and invest for the long term. Find a few of those and take a holiday.

This series began with Alamos Gold for a reason, and K92’s spaghetti soup earns its place on the table precisely because it rewards patience without requiring faith. And no, three elephants in a row — those being Alamos, Lundin Gold and now K92 — does not mean they’re as common as mice — you won’t find them on every corner. What you do get in this Substack is the benefit of my pre-selection. If this work proves useful, I’d be grateful if you shared it with a few like-minded investors; it helps justify the time it takes to do it properly.

Share

Read more

“Early on I learned that it takes as much time and effort to do a small project as a large one.” Adolf Lundin

Girlfriend, Not a Marriage

I first met Ron Hochstein at a classic speed-dating session—Berenberg, Zurich, 2018—when they pitched me Lundin Gold. I knew the Lundin dynasty from the oil patch. Lundin Petroleum rode Johan Sverdrup and was the best value for money on the Norwegian Continental Shelf besides Aker BP, by a country mile. But I wasn’t into gold miners; gold trades off real rates and I wasn’t exactly bullish on the metal.

Still, it was a Lundin outfit, so I burned 45 minutes crossing town—almost as big a waste of time as broker speed-dating. My mindset: maybe a girlfriend, not a marriage. Once the dog-and-pony show ended, I asked two things: how do you work with Lukas Lundin (mining side; Ian Lundin ran the oil side), and if you had a silver bullet, which competitor would you take out?

Ron was mid-construction and looked appropriately stressed. His answers: “two to three times a day,” and “Pretium.” Yes, you read that right. Lukas Lundin—who ran many operations, including Lundin Mining, Bluestone Resources (a gold development that was left for dead later), Filo & Josemaria (both copper explorations in Argentina which took off big later; no, they didn’t own NGEX back then)—called Ron multiple times a day for construction updates.

Anyway, Ron delivered on time and on budget. Or rather, the Gignac family did—with their mine-construction firm, G Mining Services. Today I know they’re the best construction team on the planet—mining hall-of-fame material. Back then, I had no clue. Remember the name; we’ll come back to them in this series.

The Lundin Gold stock, meanwhile, gyrated in 2018 like a pole dancer on an oil-slicked stage. Not much of a Lassonde Curve to admire. Only in June 2019 did it finally move—from C$5.10 to C$8. Partly because it was nearing commercial production (declared in February 2020), but also because gold did a 20% shuffle from US$1,250 to US$1,500.

Miners were hated across the board—precious, base, anything metallic. Why go asset-heavy when you could own Google’s asset-light recipe? When my “girlfriend” finally perked up, I happily passed her to the next suitor at C$8. Early 2020, the stock overshot gold as commercial production was announced for a few weeks, then cratered back to C$5 during Covid when Q2 2020 production went dark.

From there Ron delivered—or overdelivered—quarter after quarter, and yet the stock merely tracked gold for roughly 10 straight quarters, just with more whiplash. So no, you’re not gonna win a market-neutral set-up as a hedgy in metals. By July 2022 it was back under C$8, broadly in line with the metal. So much for the “leveraged play” in a quality name; it traded more like a guaranteed put on a slippery earnings print. That 2018 meeting taught me more about gold mining than I realized—less romance, more cash flow; keep the relationship casual until the geology proves it wants a ring.

And for the record, I owned Pretium too. That was the real stomach acid. Why, you ask? Try my last post on Alamos. The answer’s there—assuming you can read. Hint: geology. No, Pretium’s problem wasn’t grade; it was nuggety, discontinuous geology—bonanza pockets stitched through blank rock, the textbook nugget effect. Wall Street hated it. Bottom line: nature and quarterly reporting are rarely a match made in heaven.

Fun fact:

Lundin Petroleum, born in 2001, discovered Johan Sverdrup. Yes, Statoil showed up later, but Lundin made the first strike in 2010 on Avaldsnes with well 16/2-6—later recognized as part of Johan Sverdrup.

Why does that matter? Because Johan Sverdrup is, by far, the largest conventional offshore field in the North Sea today (Ekofisk/Statfjord were bigger in their days): 2.7 billion barrels of recoverable reserves and a plateau capacity up to 755,000bpd—roughly a third of Norway’s current oil production. Let that sink in. And yes, just like the best deposits, Johan Sverdrup only got bigger as the drill bit did the talking.

After the merge with AkerBP in June 2022, their Working Interest in the field is 31.57%. Nemesia, the primary investment vehicle of the family, holds 14.38% of AkerBP as per Bloomberg at the time of this writing, worth US$2.2 billion.

Better Lucky AND Brilliant: The Fiscal Turn That Made Fruta del Norte

Lukas Lundin bought Fruta del Norte (FDN) from Kinross in December 2014 for US$240m—US$150m in cash and US$90m in Fortress common shares (today’s Lundin Gold).

At the time it was one of the largest, highest-grade undeveloped gold projects on the planet: 36 mining concessions over 64,300 hectares in southeast Ecuador, 80 km east of Loja. In Kinross’s book, FDN carried 7.2Moz M&I and 2.5Moz inferred. To prove up the system, Aurelian—the original discoverer—and later Kinross drilled at least 141,000 exploration metres from 2006 to 2012 in proper jungle conditions. In today’s drilling-cost terms, that’s roughly US$50m before technical work.

Kinross walked in 2013 after failing to agree key fiscal and legal terms with Ecuador, including a punitive windfall-profit construct of 70%. The government also wouldn’t grant the stability Kinross wanted. Mining is long term. Hence, good mine operators want a legally binding standstill agreement that is enforceable under international arbitration. Consequently, Kinross took a US$720m write-off (mostly non-cash), including US$20m in severance and closure, and treated FDN as discontinued. Despite sabre-rattling over the licence timeline, Kinross ultimately sold to the Lundin group.

For perspective, in 2014 Kinross produced 2.7Moz GEO (gold equivalent) and carried 34.4Moz reserves, 23Moz M&I and 4Moz inferred, with a US$3.2bn market cap. In other words, the market valued all of Kinross’s resource at US$52/oz with gold at US$1,266/oz. Kinross sold FDN for US$25/oz of resource. Today, markets value 15Moz of FDN resources at north of US$1,000/oz—forty times that. Not bad.

Did Kinross blow it? Not ex-ante. In 2013–2014, Ecuador’s regime under Rafael Correa (notably the 70% WPT) made FDN uneconomic for a major at US$1,300/oz. The mine was in the middle of nowhere and required power lines, roads and plenty of upfront capital. The US$240m exit was rational. One could even argue it was a good price, likely replacing most of Kinross’ cash cost. Who knows.

What we do know is that Latin America had the usual above-ground noise around that time. On 6 February 2016, The Economist was tallying social-licence conflicts across the region. Chile’s Supreme Court suspended Pascua-Lama; Colombia’s courts blocked Mandé Norte. Latin America has the metals endowment across the periodic table, but politics were (and are) a contact sport. And that is why Ecuador remained a mining minow. It had a mere 60 mines operating in 2015 and registered 7 conflicts in the piece. That is 12%—“Lodes of conflict,” as the title of the graph correctly reads.

Only after Lenín Moreno took office in 2017 did Ecuador finally grow up on mining policy—the 2018 reform killed the 70% Windfall-Profits Tax (WPT) for mining and made the regime investable. But when Lukas bought in December 2014, the “reforms” (Decree 475 and a grab bag of incentives) were lipstick on a pig; the 70% WPT still hung over the model. For a portfolio major, FDN screened poorly. And Quito wasn’t offering squatters’ rights: work it or risk the licence. You can’t have your cake and eat it too. Lukas was a risk taker in the Lundin family sense. Adolf is quoted to have said, “no risk no glory”. It lived on in his most capable sons.

By 2016, Lundin announced the adult thing—paper first, rocks second. Call it proper risk management. The Exploitation Agreement locked the project’s fiscal rules and clarified how the WPT would work: it would apply only when the metal price exceeded a stipulated contract “base price,” and only after FDN had recouped its cumulative investment (the Government even floated “payback + four years” during the legislative clean-up). In plain English: the 70% take hit just the price-above-base slice, not every dollar of profit, and not until the mine had paid itself back. The contract also included a US$65m advance royalty—unusual for most mining jurisdictions, but part of the bargain to get bankable certainty for a worldclass deposit.

The Investment Protection Agreement signed later in 2016 layered legal and tax stability on top: a fixed 22% income-tax rate for FDN, ISD remittance-tax relief on debt service, and most-favoured-investor style protections. But none of that repealed the WPT. It just made the playing field knowable while the politics caught up. The actual repeal of the mining WPT arrived in 2018 under Moreno’s Ley de Fomento Productivo. That’s when the windfall tax finally left the stack. That was the game changer. And that was lucky.

Today, Lundin Gold still pays plenty of taxes. If you own the stock, study the fiscal take as closely as grade and capex—it defines intrinsic value. Ecuador mandates a 15% profit share on pre-tax profits in mining: 3% to employees, 12% to the State. It shows up as 3% in operating costs and the 12% in the tax line.

On headcount, Lundin doesn’t list employees in the AR, but 2023 sustainability materials note 1,907 employees (local press pegs 2024 in the same ballpark). That’s a lot of paycheques tied to one mine—a real, tangible resilience mechanism in a country-risk discussion. Our calculation suggests FDN will pay out US$334,000 per employee, assuming a US$4,000 price deck, ceteris paribus.

FDN’s operator, Aurelian Ecuador, pays a fixed 22% corporate income-tax rate (CIT) for the life of the mine under the IPA, even though Ecuador’s headline rate rose to 25% in 2018. So that is good news. It shows some fiscal stability. It’s the little things that matter in judging country risk. Also, Profit Shares are deductable for the purpose of CIT. Sensible.

FDN obtained relief on debt service in the so-called Impuesto a la Salida de Divisas (ISD)—Ecuador’s tax on the exit of foreign currency. Dividends can qualify for ISD exemptions subject to the usual conditions, which we assume too. As social influencer and presidential candidate Gloria Álvarez in Guatemala explains: Latin America knows 50 shades of socialism. Finally, under the Canada–Ecuador treaty, dividends to Lundin Gold carry 5% withholding (treaty conditions apply); without treaty relief, Ecuador’s domestic rule is 10%.

In our intrinsic value work we thus model the stack as 15% profit share + 22% income tax + 5% dividend WHT, with profit share deductible before CIT and dividends paid after tax, the effective take lands below 42%. Not great, but certainly more competitive than many African regimes that layer equity carries up to 15% on top of royalty and income taxes.

In all, Lukas had timing and spine. Better lucky than brave, sure—but they were both. You couldn’t guarantee the political turn in 2017–2018; you could position for it. He sure did. But let’s be frank: the FS 2016—before the turn—suggested US$669m upfront capital to put the mine to work. The gold price in 2016 and 2017 averaged more or less US$1,250.

If we were to apply that price deck in our model with a 70% WPT, FDN would have had a hard time returning its invested capital—we mean ever. Even with the capital-recovery protection agreed in 2016, it still would have struggled to pay dividends later at a US$1,250 gold deck. Water under the bridge.

The IPA, combined with the 2018 repeal of the mining WPT under Moreno, made greenlighting FDN a no-brainer. Without both, Lukas might have walked—just as Jack Lundin (one of the two active superstar sons now running the mining side of the Lundin empire) did with Bluestone Resources in Guatemala in 2022. Adam is the other. Harry and William—also Lukas’s sons—are active on the energy side of the group. We’ll leave other family details aside; this story is about FDN.

The lesson: if you invest long-duration in LatAm resources, you study politics, social issues and taxes like an operator and do not rely on management slide shows. The amplitude of change dwarfs North America, Australia, and Europe—despite the fashionable takes that “our” above-ground risk is just as bad.

Among us girls: it isn’t—except, of course, in the UK, with its so-called Energy Profits Levy (EPL), the local windfall tax that’s been kneecapping the North Sea’s best operators since the post-Russia gas shock in 2022–2023.

Fruta del Norte — The Geological Unicorn

Those of you who’ve followed me on X know the drill: resource investing starts with the deposit (in metals) or the reservoir (in oil and gas). A gift of nature can offset a lot of above-ground pain and price volatility. It’s the insurance policy for corporate survival—especially when it takes real technical skill to lift and sits out of easy reach of bureaucrats and “rebels.” Think offshore Nigeria in oil terms. Same playbook.

So let’s look at what Fruta del Norte actually is. Before the jargon, apply common sense. The ideal underground gold deposit looks like this.

Continuity. The ore body sits at moderate depth and runs laterally—a panel, not a pinprick; a football, not a fishing line. You drive in, not down a thousand-meter shaft. Minimal waste development. A short, straight road to ore. The grade shell barely pinches over mineable distances. Lower risk, lower cost, fewer surprises when it’s time to hit the quarter for Wall Street.

Width and depth. You want a thick, consistent block—again, football not spaghetti. Not stringy lenses disappearing at depth, not broken shoots you can’t find when you need them. Call it 100–150 meters true thickness, at serious grade, near enough to surface that physics and diesel are on your side.

Rock competence. You want competent host rock that doesn’t fall apart, so you can run big stopes without diluting grade. In the plant, you want metallurgy that cooperates—gold that separates cleanly enough to deliver high recoveries, without throwing the periodic table at it. Doré over alchemy; if you must make concentrate, fine, but don’t make it the only route to pay dirt.

That’s the dream. Continuity, simple geometry, competent rock. Simple!

Now, let’s park the daydream and look at this ore body.

Fruta del Norte is an epithermal vein-breccia hosted in sediments. Now, I have no idea what that really means - geologically speaking. But in my resource table below you’ll notice three things that pop immediately if you’ve looked at enough mines. None of it takes a geology degree to comprehend, just commen sense and basic math.

First, grade: M&I averages in the 7g/t neighborhood with reserves around 7.8g/t—serious rock even in underground land. Second, scale: the ore zone is 100–150 meters thick, runs roughly 1,700 meters along strike, and sits around 700 meters below surface. Very special. Third, continuity: it’s a single, laterally continuous panel. That’s “rare earth” material.

The geometry makes the mining plan sing. A true thickness of 100–150 meters means big, mechanized stopes—exceptional at these grades. Access is by side portal and decline; you’re not hoisting from a kilometer down. It’s “drive to the face, mine, turn, drive home”—far simpler than the contortions at Island Gold or Young-Davidson.

Don’t get me wrong: Alamos turned both into industrial beauties. My point is that Fruta del Norte asks for less tech and fewer brain cells to deliver. It’s the kind of asset Buffett wants—a wide moat that even an idiot can run, because one day an idiot may will.

On the plant side, the company reports dilution around 8% and 88% gold recovery. That’s fine—room to optimize, but already very profitable rock. Metallurgy isn’t trivial here; FDN produces a mix of doré and concentrate, and the team keeps nudging recoveries upward. If they crack another couple of points consistently, so much the better—but let’s not get greedy now when geology is doing this much heavy lifting.

Fruta del Norte is the rarest thing in mining: size and obedience. Continuity, thickness, access, grade—four aces in one hand—translating straight into low costs and long duration. She’s a Latino beauty, but without the usual high-maintenance supermodel drama. Look at her. It’s screaming “buy.” Meanwhile 99% of Wall Street can’t read a long-section. A geological unicorn, hiding in plain sight.

If Escondida is the type example for open-pit copper, FDN is the underground gold analogue—an ore body that just gets the big things right. Per kilometer of strike, it’s one of the densest gold endowments per kilometer on the planet. A block of money lying flat at 700 meters, waiting to be mined for the next 20 years—methodically, predictably, without endless technical headaches. The proverbial gift of nature.

Which is why Lukas Lundin didn’t need a séance in December 2014. He’d looked at hundreds of deposits across metals. When a world-class geometry shows up at a price with a pathway to bankable contracts, you take the risk. Pros do.

And you don’t need to cosplay as a geologist to see it. Spend a few hours with technical reports, stare at the long-section, compare vein systems and thickness columns across peers, and let common sense do the work. Mining has so many comps it’s practically a stock picker’s playground. The edge isn’t in sounding smart; it’s in seeing the obvious—continuity, geometry, competence—and then waiting while the cash flow tells the story. Looking at geology and knowing who the real operators are is the legal version of front-running Wall Street.

Land Optionality: Let the Drilling Begin

Does the Latina supermodel have geological upside? I don’t know. Nor can we know. It’s a known unknown, in the words of former Secretary of Defense Donald Rumsfeld. But there are a couple of things we can say.

Let’s start with the country. Ecuador—LatAm drama like the rest of the EM pack—did its best to stay poor by shutting the mining cadastre in 2018 over “irregularities in the concession system,” whatever that meant. Code for corruption, in all likelihood. It reopened in June 2025, finally launching new concessions after seven years to attract projects and curb illegal operations. Dahh.

Between 2018 and 2025, Ecuador didn’t outlaw drill rigs at FDN, but it starved exploration with policy, permitting gridlock, and uncertainty. Closing the cadastre froze fresh ground and slowed anything beyond the mine’s immediate footprint. Add the 2020 COVID stoppage and, later, court fights over new environmental-consultation rules (Decree 754 was provisionally suspended in 2023), and you had a national case of hurry up and wait.

The rational choice for Lukas Lundin and Ron Hochstein was to build the mine first, chase ounces later, which is exactly why the resource history reads like a flatline with barely a pulse: tiny additions, depletion marching on, and not much movement until the drills came back in earnest in 2022/23, with real replacement finally in 2024. In fact, Lundin Gold didn’t even bother to report annual reserve reports for 5 years straight.

Once the mine was running and access improved, Lundin rolled out its largest district drill push since 2007—first near-mine, then regional. By 2021–2022 they were back on pads around the Suárez Basin, a 38km² pull-apart that hosts FDN and—crucially—looks big enough to host siblings. The run of news that followed (near-mine hits in 2022–2023, then a full district program) reads like a company taking the restrictor plate off. See for yourself.

Can the company deliver on this 2025 ambition? Absolutely. Lundin is running 16–17 rigs in 2025, which is massive. At that fleet size, you need only 16–18 meters per rig per day to hit 108,000 meters—well within normal productivity, especially with a mixed fleet (surface + underground). The implied 8 meters per crew per day (two crews per rig) is conservative for diamond coring, leaving buffer for weather, maintenance, or pad moves. Ron and team will deliver—unlike the many junior lifestyle outfits without funding. Hunt elephants, not mice. It pays over time.

It gets better. Listening to Lundin Gold’s Investor Day 2024, studying their approach, and thinking about the industrial probability I get good karma in a probabilistic ballgame. They have a team of 20 geoscientists underpinned by a successful operating model. And much of it is new. VP Exploration Andrey Oliveira joined in March 2022. Since then, he’s hired 18 new geologists (ten women from communities around FDN).

Under his leadership, the company is focusing on near-mine exploration within a system that is open south, west, east, and at depth, plus a regional program—what I’d call a real exploration culture run by people who’ve actually seen this epithermal style before. Best of all, Andrey seems to embrace AI and big-data tools. Tech should let exploration make quantum-leap gains when applied with discipline.

Do I know what that will ultimately yield? No. Known unknown. But here’s a known known: Andrey’s team will likely turn over every single rock on 63,000 hectares of highly prospective epithermal ground—meter by meter, systematically, with no budget constraint and a tech stack unprecedented in Ecuador’s exploration history. That’s real horsepower. Name me a junior that can match it and fund it without equity dilution. There isn’t one.

Here, you get real exploration by pros—in proven geology, with a unicorn already on site, an operating mine ready to absorb any discovery (speed matters), and a rock-solid social contract on top. Keep that in mind when you’re pitched junior stories with nothing but glossy slides, a Lassonde-curve cameo, and “100x” typos. That’s not investing; that’s roulette—most of it noise dressed up as geology.

Here, the drilling is already paying off. Bonza Sur, one kilometer south of the FDN portal, started as a humble soil anomaly and is now a new gold deposit. The mineral envelope starts at surface, runs more than 1.8km along strike, is about 100 metres wide, and extends to at least 500 metres depth in drilling. That’s not a scout blip; that’s scale with geometry.

Lundin Gold is guiding to a maiden resource in H2-2025 and had one rigs chewing on it in Q2 2025. 11 surface rigs were drilling, with four of them at Trancaloma (east to FDN), one at Sandia (north east to FDN), one below FDN depth, one at FDN East, one at Bonza Sur, and three testing new sectors. Whatever they find inside the above white-dotted perimeter converts, by design, to high-margin cash flow—no fresh permits, no new mine build, no strings attached. Call it 1:1 conversion when its slot comes up in the mine schedule. Sexy.

Meanwhile, the regional map remains gigantic. Lukas acquired 64,300 hectares—roughly half the size of the Canton of Zürich with its 1.6 million inhabitants. It’s huge. There’s hardly a soul on it, and the land package has never been systematically explored. So yes, you can be cautiously optimistic that where one supermodel hid, a few mini-models may be waiting their turn. A dream come true for dreamers.

When FDN’s VP Exploration Andrey Oliveira says “epithermal systems are never alone,” he isn’t speaking figuratively. The geological corridor already hosts other mines and projects—Mirador (copper), Warintza (copper), and several prospects. The best mental model might be this: the Carlin Trend began with one orebody in 1961; today it’s a 90-million-ounce hydrothermal city. Pierre Lassonde understood the power of that optionality when he scraped together US$2 million to buy the royalty—every dollar he had.

Epithermal systems come in clusters, and if the geology stays preserved—as it has under the sedimentary cap at FDN for 150 million years—each new vein corridor can extend mine life by decades. I don’t need to and couldn’t lecture you for one minute on the petrology; I just need the history: that’s what nature did. What we’re watching in southern Ecuador could be the early chapters of a modern Carlin-style evolution, written in epithermal script rather than carbonate. With budget, brains, and tech, Andrey’s team is likely to find more gold, copper, silver, and friends. But let’s not outrun the drill bit—or mother nature.

Have we valued any exploration upside yet? No. I’m not going to model fences and micro-structures off a handful of holes, and unless results cluster, regional targets won’t become economic. Our model, however, uses ALL 9.4Moz of resources and risks none—chance of success set to 100% for the in-resource case.

Sell-side analysts typically value reserves in full, then apply a risk factor to measured and indicated, and often assign zero to inferred. They have their reasons, but most are intellectually flimsy and mainly designed to avoid embarrassment when the market is watching. Intrinsic value in resources has to be situational. But resources come in countless configurations: sometimes the extra ounces aren’t worth the paper their block model is printed on; sometimes they’re worth every ounce on paper; most often it’s somewhere in between.

Unfortunately, many sell-siders are juniors fresh out of university—overwhelmed for the first couple of years—and they operate inside a culture that rewards herd behavior over independent thinking. Keep that in mind when you read their price targets (which I haven’t done in years). They struggle to think outside the box on gold and, in a rising tape, basically set targets to match the market. Groupthink at its finest. It’s a total waste of time.

Anyway, my value approach is not timid, as you will learn further down. Nor should it be—with a Latina supermodel. But it’s a shade more conservative than what I did at Alamos’s Island Gold, where we have a no-brainer growth track over five years and on-the-record guidance from John McCluskey. Here, we wait. The unicorn orebody makes the inferred feel borderline a no-brainer (hence the 9.4Moz in the model), but we’ll let the assays, not the adjectives, carry the rest of the day.

The range of outcomes is wide: Fruta del Norte could carry 9+ million ounces for years—if not decades—via steady year-by-year replacement; or the resources could peak in the next few years. What the drilling already suggests is that average grades aren’t improving the way they did at Island Gold—but tonnage (ore volume) may grow. Time will tell.

One last remark on exploration: you read our Alamos piece. You know what to look for on resource upside: continuity that scales, geometry that mines, competence that recovers, metre by metre.

That’s the essence of land optionality. FDN has the address. Now let the drill bit write the invitations. At the very least, near-mine results should keep replacing reserves for years to come, as 2024 already showed. Now let’s watch out when the first maiden resource for Bonza Sur is decleared.

Meanwhile, FDN isn’t just a geological unicorn, it’s an operational case study too.

Read more

“It’s the optionality value of the land, the value of the operator spending money on our land, and the optionality to higher gold prices. And that is worth so much money… it’s what’s given us the Goldstrike billion-dollar royalty, the Detour billion-dollar royalty, the Tasiast royalty, and so on.”

Pierre Lassonde, Founder Franco Nevada

Guiding Principle

Last month I promised a series of gold-miner drill-downs. This is the first, and it will set the tone for the rest. My aim is simple and unfashionable: help you look under the roof of resource companies the way we have done for more than 10 years. This is not trading. It is fundamental work anchored in an industrial view of how mines actually make money, tied to a Discounted Cash Flow that treats the business as if you owned 100% of it. Gold will remain volatile, yet the metal sits in a structural bull I intend to lay out in detail this year. For now, we begin where it matters: the miners.

I’ll start with Alamos Gold because it lets us teach the discipline of underground mining through a live, compounding case: Island District, Young-Davidson, Mulatos pivoting underground, and Lynn Lake moving from permit to construction. You will learn how geology, method, power, people, and capital combine in the only way that counts—cash flow. You will also see why words like “resource,” “reserve,” “cutoff,” “AISC,” and “optionality” are not jargon; they are the grammar of value.

Thereafter and as a preview, this series will continue with Lundin Gold, then K92 Mining and Osisko Development. Each of these underground operations or soon-to-be operations sit at a different point in the life cycle, in a different jurisdiction, with different ore bodies and methods—long-hole stoping versus bulk tonnage, shaft versus ramp, grid power versus diesel.

Those choices explain costs, margins, replacement ratios, and the pace of discovery. They also let us talk plainly about capital structure: royalties, streams, prepay gold loans, and why taxes—royalties, VAT, withholding, windfalls—are not a footnote but a central driver of intrinsic value. Country risk will be treated the same way: not sentiment, but contracts, cash costs, power tariffs, permits, and social license.

Later in the series we’ll move to open-pit gold miners and heap-leach operators, because understanding how oxide flowsheets print cash at scale is as important as knowing how a shaft mine compounds.

If you stay with me through this arc, you won’t need to chase tickers or narratives. You’ll have the tools to read a technical report, translate it into cash, and decide whether time is your friend or your enemy.

That, in the end, is the point of this Substack. Let’s begin…

Alamos Gold and the Value of Land Optionality

Pierre Lassonde bought Franco-Nevada’s first royalty in 1986 on the Goldstrike mine in Nevada’s Carlin Trend for US$2m. At the time it was a small heap-leach run by Western States Mining.

Soon after, American Barrick (now Barrick) bought Goldstrike and drilled deep. The result was a 50Moz system that powered Barrick—and turned Franco-Nevada’s tiny cheque into a franchise.

That is land optionality: let a capable operator spend money on your ground, let geology and time do their work, and let higher gold prices widen the funnel.

John McCluskey at Alamos has done the same in his own way. Few outside the industry realize he’s a living legend—active, disciplined, and still compounding.

If you want a clean, liquid way to be long gold with operating leverage, Alamos Gold is, in my view, the best beta in the space that still trades easily.

Why is Alamos special? It starts—and ends—with people.

Read more