Adam Rozencwajg — Oil Crisis Not Over Yet, This is the Real Bottleneck
"That to me is not the sign of a very weak demand market. That's the sign of a refining problem." Why 150 days of a closed Strait of Hormuz produced no visible crisis — and why the missing barrels are hiding in the one inventory nobody can measure.
Recorded 2026-SEP-01 ("we talked today on September 1st"); published 2026-SEP-03. The figures come from G&R's quarterly letter released 2026-AUG-31. The YouTube page also carried an earlier title variant, "Oil Crisis Hiding in Plain Sight, What to Watch Now".
One-line take: The consensus reads 150 days of a closed strait without a crisis as proof the oil market was always loose. Rozencwajg's counter is a measurement error: the reported 5 mb/d of "lost demand" is really a refining outage. Roughly 10 mb/d of upstream production has been shut in since March (~1.5B barrels), while global refinery throughput is down ~6 mb/d — Gulf refineries idled because product would be trapped in the region, China's ~2 mb/d export refining halted in favour of its domestic market, and Russia's complexes war-damaged (~1 mb/d). Because demand models are "essentially a function of GDP and refinery runs," a war-driven run cut is booked as demand destruction — a loss "2x the global financial crisis" that airline miles, vehicle miles and general activity flatly contradict. The falsification test is the crack spread: normally $10–20, "they hit 100 bucks. And they're staying there." So end demand is intact and the shortfall is being paid out of refined-product inventories — the one series nobody instruments, because product tanks have fixed roofs and defeat the satellite shadow method that works on floating-roof crude tanks. The IEA says non-OECD product stocks are down ~50M barrels; "if we're right, it might be down 500 million" — an order of magnitude, with Bangladesh, India and the Philippines already generating the headlines. Both exits are bullish crude: a product crisis in the global south into Cushing at operational minimums and heavy speculative shorts, or a reopening in which idled refineries and drained SPRs all bid for crude at once. Positioning is unchanged from January — sold gold on the "silver sell signal," rotated into crude — expressed through Canadian oil sands, selective US shale, and offshore drilling. Elsewhere: gold's ~20% pullback is likely only half-done (typical is ~40% over one to two years, and ETF length is still elevated), uranium's term price is at an all-time nominal high of $95.50/lb while the equities fell 30-odd percent on no news, and global coal is the "four-letter word" nobody owns into a European gas shortfall and Indonesian export curbs.
1. Stocks & names mentioned
| Ticker | Name | Research | View | What he said | At |
|---|
| Crude oil | Crude oil (the molecule) | — | Positive | "Where do prices go from here? I suspect higher. I don't think we're out of the woods. I don't think we can write down the probability of tank bottoms to zero." Crude in the low $90s never exceeded its April–May high even though "unequivocally the global petroleum situation is much worse today, meaning much tighter" — because the only thing moving price all year "has been what your risk team is allowing you to short." | 32:23 |
| Cdn oil sands | Canadian oil sands producers | — | Positive | Unchanged since the January rotation: "we continue to favor Canadian oil sands names. I like good long-lived assets in a friendly jurisdiction." He is Canadian, has lived in the US 24 years, and still rates both "good places to do business" measured against the rest of the world. | 34:08 |
| US shale | US shale producers (selective) | — | Positive | "US shale producers, where you can find them, still offer really attractive assets" — but "we're depleting them rather quickly and it's hard to find really good quality reserves anymore… not enough to build a whole portfolio around the way it was a few years ago." The same slowdown is why G&R bought crude in January: the growth engine that "carried all of global supply growth for the last 15 years was slowing sharply." | 34:34 |
| Offshore drilling | Offshore drilling / drillers | — | Positive | "We also favor the offshore drilling sector… they're in a massive bottoming process here," with "very very favorable green shoots developing, consolidation in the industry" and balance sheets wiped clean by COVID bankruptcies — "very very attractive risk-adjusted returns." The price of admission is volatility: the offshore names "will either be our biggest contributors or biggest detractors to performance in any given month or quarter." | 34:54 |
| Uranium | Uranium / nuclear fuel | — | Positive | The term price — "which is where 90% of the market transacts" — made an all-time nominal high in Q2 at "95.50 a pound," breaking the '08 peak by 50 cents, while uranium equities fell "30-odd percent, on absolutely no news whatsoever." Utilities are "still very under covered in their long-term contract books" and there is "not much in the way of new mine supply to bail the market out." "Just buy it and put it away and enjoy the uptrend." | 42:30 |
| Coal | Global coal (seaborne thermal) | — | Positive | "The other thing that people don't look at at all is the global coal industry. It's obviously a four-letter word, completely starved for capital, no one cares." Two near-term catalysts: Europe "seems very unlikely to be meeting their gas requirements going into the winter" so coal stockpiles get rebuilt (the 2022 playbook), and Indonesia — "what the US has been… to oil and natural gas" for coal — is looking to ban or severely limit exports. "There's some good attractive opportunities there." | 44:15 |
| Refined products | Refined products (diesel, gasoline, jet) | — | Neutral | The real bottleneck, and the diagnostic: crack spreads that "normally might average between 10 and 20 dollars… hit 100 bucks. And they're staying there. That to me is not the sign of a very weak demand market. That's the sign of a refining problem." He believes non-OECD product inventories are down ~500M barrels against the IEA's ~50M — "an order of magnitude difference" — but the position he owns is crude, not the product. | 17:29 |
| NXE | NexGen Energy (Rook I, Athabasca Basin) | QT · SA · STK · FA | Neutral | Named, not rated: "there's been some interesting news rumors around NexGen and BHP in the last couple weeks that there might be something going on there." His point is supply, not the takeover — "that project still remains a number of years away… that's a big project, the Rook I, to bring online," so it does nothing to relieve a term market already at record prices. | 42:52 |
| BHP | BHP Group | QT · SA · STK · FA | Neutral | Mentioned once, as the counterparty in the same rumour: "some interesting news rumors around NexGen and BHP in the last couple weeks." No view on the company is offered — it is cited as evidence that major diversified miners are circling development-stage uranium while new mine supply stays years away. | 42:52 |
| Gold | Gold & precious metals (incl. gold equities) | — | Negative | Still out, still early: G&R sold down gold in January on their "silver sell signal" and bought oil stocks. The pullback is "about 20%, but it's about half of what we've seen in the past" and has taken 6–7 months against a typical one to two years; ETF holdings "are still quite elevated"; "the gold to oil ratio is still very much in oil's favor." "I don't think that the sell-off is over" — long-term bulls, but a re-entry needs either a deeper flush or central banks doubling their purchases. | 38:06 |
"View" is Adam Rozencwajg's stance in this conversation (Positive / Neutral / Negative), not a price rating. Research links: QT Qualtrim · SA Seeking Alpha · STK Stock Analysis. Row scope: this is a physical-market and asset-class conversation — crude, refined products, gold, uranium and coal are discussed as commodities, and Canadian oil sands, US shale and offshore drilling as sectors he owns without naming a single holding, so those rows carry no ticker (per the hub convention, a sector or commodity is never given an invented symbol). NexGen Energy and BHP are the only listed companies he names in the whole interview, and both only in passing, as an M&A rumour.
2. Talking points
0:45 The question: 150 days of a closed strait, and no crisis
- McLeod's framing: since the mid-March conversation about how disruptive the Iran war and a Hormuz closure would be, "the impact hasn't been as strong as many expected."
- Rozencwajg concedes the framing is fair and warns the answer is arithmetic-heavy: "this is one of those ones where there's quite a few figures and facts."
1:10 The source document — G&R's quarterly letter, published August 31
- "We just released our latest quarterly letter… put out just yesterday morning on the 31st of August. You can find it on our website" — it carries the charts and tables behind every figure in the interview.
- "Don't try to scribble down anything. Just take the story for what it is… it's one that has really profound investment implications as we go through the rest of the year."
1:58 The original arithmetic — 20–25 mb/d through the strait, ~15 impacted
- The strait carried "20 to 25 million barrels passing through it depending on whether you included NGLs and products or just crude oil," and "that essentially ground to a halt."
- Saudi Arabia, Iraq and the UAE found alternate egress fast — most famously the Saudi East-West pipeline "ramped up above nameplate capacity" to export from the Red Sea — recovering ~5 mb/d, leaving ~15 mb/d impacted.
- 2:47 "If you did that for 75 to 100 days, you should have drawn global inventories by over a billion barrels, up to 1.5 billion barrels even."
3:23 Why 6–7 billion barrels of global stock isn't spendable
- Counting everything — strategic reserves, commercial tanks, oil on the water — the world holds 6–7 billion barrels, but "most of that oil is actually trapped… it's not so much inventory like a savings account that you can draw down, rather it's more like working capital that you need to run your business."
- The components: ~1.5B barrels must sit in pipelines ("you can't drain oil out of a pipeline, otherwise it stops operating"); ~1.6–2B barrels must be afloat to sustain an 80 mb/d seaborne market at ~20 days per voyage; plus tank heels, "minimum operating capacities that the tanks can't actually get below."
- 4:24 Net of strategic reserves the drawable pool is "about a billion barrels of easily mobilized crude and product," against an OECD commitment to release ~400M — versus a 1.5B-barrel loss from fields shut in "because they had nowhere to put the oil."
4:50 Tank bottoms — the mirror image of minus $45
- "If this lasts much longer, we're going to potentially get to tank bottoms, which is a very, very catastrophic event… we essentially go from having a buffer to being just in time. We've never actually done it on a global basis."
- The only precedent is the opposite end: COVID filled inventories "to the brim," and at tank tops "we got the minus $45 oil print and you lost $70 in three or four sessions. Presumably something just as dramatic might happen on the flip side."
- 5:40 That alarm was rung in April. "So, here we are… we're over 150 days into this crisis," with the strait "certainly impaired, if not absolutely closed" bar one brief exception — "and yet, we don't seem to have a problem."
6:09 The bears' inference — and why he rejects it
- "The oil bears will tell you that this is proof that the market was in a much greater excess than we had anticipated… far looser and more resilient than anyone, including ourselves."
- His answer: "we think that there's things that are happening perhaps just below the surface, just out of sight of most analysts, that is in fact extremely alarming" — the bridge between modest US and European draws and "the potential catastrophe that I think is looming just around the corner."
8:12 What actually happened: 10 mb/d for 150 days
- Comparing January–February to March–July, and including the brief June reopening, "it now in retrospect seems as though we lost about 10 million barrels of production. That's upstream — at the field, from the wellhead."
- "We've lost 10 million barrels per day over 150 days. So, you're back at your 1.5 billion lost barrels that we had worried about a few months ago."
8:56 The published balance — and the 5 mb/d of "lost demand"
- The headline offset, per the IEA and the banks that start from it: demand also fell 5 mb/d, plus a balancing item that loosened by 1 mb/d — a net tightening of 4 mb/d.
- Off a claimed ~1 mb/d starting surplus, that is a 3 mb/d deficit over five months ≈ 400M barrels, which reconciles neatly against ~50M drawn from OECD commercial stocks, ~200M from strategic reserves and ~100M elsewhere.
- 9:34 "But the problem that we see is with that 5 million barrel a day demand figure, because when we look out at the world, we don't see 5 million barrels a day of lost demand."
10:03 Why that demand number isn't believable
- A 5 mb/d loss "is worse than the oil demand loss that we experienced during the global financial crisis… kind of like 2x the global financial crisis," and on par with Q1 2021 under COVID mobility restrictions.
- "When I look today at airline miles, when I look at vehicle miles traveled, when I look at general economic growth and activity, absolutely none of those things are consistent with 5 million barrels a day of demand loss. In fact, most of the data that I look at is consistent with year-on-year demand growth."
11:23 Supply and demand are two markets — with a refinery in between
- "Supply" is what comes out of the ground (crude, condensates, NGLs, some biofuels). "Demand" is what consumers and industry actually burn (gasoline, diesel, jet, petrochemicals). "Those aren't the same things and in the middle sits the refinery system" — ~85 mb/d of crude runs plus NGLs, biofuels and refinery gains to reach ~105 mb/d of product.
- 12:46 The convention that hides the problem: "usually what ends up happening is the industry will report how much oil went into the refining system." Nine times in ten that is safe, because refiners run to the market's price signal and refining "is rarely the bottleneck… that whole refining system happens offstage."
- 13:31 "We don't make the bridge into these intermediate steps that could be important at times of dislocation. And today we have the times of dislocation."
13:55 Where the 6 mb/d of lost throughput went: the Gulf, China, Russia
- Middle East — refineries offline "because they can't get the products out of there." Even during the June reopening, running Gulf crude through a Gulf refinery meant product "would be trapped in the region for another 2 weeks" — "and in fact, that was a good call because the MOU… didn't last much longer than 2 weeks."
- 14:47 China — "a very curious thing": it shut off its refined-product export market entirely, importing only enough crude for domestic needs and ceasing to be "the outsourced refinery for a huge part of the world, particularly the global south." That is ~2 mb/d less through its system.
- 15:54 Russia — down ~1 mb/d on "very, very extensive war damage to their refining complexes." Total: refining demand for crude down 6 mb/d — against headline petroleum demand down 5 mb/d. "The models that everyone typically uses… essentially a function of GDP and refinery runs" therefore book a war-driven run cut as a demand collapse.
17:06 The falsification test: $100 crack spreads
- If demand really had collapsed, crude and product prices "would be moving together" — both weak. Instead, when oil round-tripped to its January–February lows in June, "product prices never budged."
- "The crack spread… which normally might average between 10 and 20 dollars, they hit 100 bucks. And they're staying there. That to me is not the sign of a very weak demand market. That's the sign of a refining problem."
- 17:49 The implication: 6 mb/d less crude run with unchanged end demand means the world is "short between 5 and 6 million barrels a day of refined product," and "refined product inventories have been bridging the gap… collapsing around the world."
18:38 The blind spot — floating roofs, fixed roofs and satellites
- "The one area that we really have absolutely no visibility into is refined product inventories." OECD data is "pretty directionally accurate"; the non-OECD world "gets a lot murkier."
- 19:52 How crude gets estimated: satellite imaging works because crude tank roofs float on the oil, so "by looking at the shadows of the side walls against the roof at certain times of the day, they can impute how full or empty those tanks are."
- 20:39 Product tanks are fixed-roof: "there's really no way from space that you can say how much inventory is inside… but it doesn't stop people from estimating."
21:03 50 million versus 500 million barrels
- The IEA carries non-OECD product inventories down ~50M barrels. "And if we're right, it might be down 500 million barrels. So it's an order of magnitude difference. And that's essentially we've run out of inventory of refined product in most of the world."
- The corroboration is already in the news flow: "the headlines that we're getting from places like Bangladesh, from places like India and the Philippines… there's major, major problems developing in these markets because we've simply run out of refined product."
21:49 Complacency, and the eye of the storm
- "The complacency that investors have today is really based on the idea that we don't have a crisis yet… I think that feeling of calm is really misguided." Inventories "are materially lower than anyone thinks. They're pushing on that billion barrel mark."
- 22:56 The overlooked admission: when Trump and Vance were criticised for a weak MOU, "they both said, look, we had to do something because even four more weeks of the Strait of Hormuz being closed would lead to a disaster." That was five or six weeks ago — "well, here we are 7 weeks later and there's really no end in sight."
- 23:28 "We might be in the eye of the storm. In fact, I think very, very likely that we are in the eye of this storm and we still have the other half of it to get through."
23:46 Path one — an emerging-market product crisis into a heavily short market
- "Either we hit this wall and the global south runs out of refined product inventory and there's a crisis and that spills and propagates into other parts of the world very quickly." Meanwhile "there's a ton of short interest on the different crude contracts, whether Brent or WTI."
- 24:09 Cushing — the WTI delivery point — is "basically at their operational minimums right now," and diverting oil there in a squeeze runs into "lots of logistical bottlenecks… Immediate doesn't happen immediately."
- The 2020 analogy runs in reverse: when oil went negative "it wasn't that every tank overflowed, it's that those tanks risked overflowing and you had to settle your contract there."
25:09 Path two — the reopening is itself bullish
- Reopen the strait and ~2.5 mb/d of Middle East refining returns immediately; China "would likely come roaring back because you can make a tremendous amount of money right now… with $100 crack spreads"; the rest of Asia is "struggling to find the feedstock." Only Russia's damaged plants stay down.
- "So, we would have a huge bid for crude to go and refill that refining system. And there would be a massive, massive bid for refined product" — everyone incentivised to run above nameplate and defer maintenance.
- 26:24 Add a third bidder: governments whose strategic reserves are now "at essentially almost operational minimums as well," OECD and China alike. "If there's one lesson here, it's that you probably want to have a strategic petroleum reserve. I can't imagine how that's not the takeaway from the last 4 months."
- 27:15 "Ironically it's going to take maybe a getting back to normal in order to truly manifest and see all these bottlenecks." Either that, or a brick wall first — "in both cases we're pretty dangerously close."
28:03 The repricing trigger is a data release
- "If that data were to be released to the market tomorrow, that would result in a pretty big panic, certainly a big short covering." Whether it arrives via crisis or via normalisation is unknowable — "but it's not nearly as calm as the price would suggest."
28:43 Price and positioning — a $90 print nobody is bullish on
- Crude is "trading in the low 90s" as of September 1, after touching ~$70 on the June reopening announcement. "It hasn't made a new high. In fact, the first high was done back in April and May" — even though the situation today is "unequivocally… much worse, meaning much tighter, much more bullish."
- 29:54 The war spike was not a bull move: "the physical price spiked because people started hoarding," forcing "a mass liquidation of the short positions, notably at hedge funds." Tellingly, "right after the crisis started, most of them fired their energy teams… I suspect the reason you would fire your energy team at the outbreak of a war is if they were really short going into that war."
- 31:19 The shorts came straight back — by the second MOU "the gross short positioning of all these speculators exceeded where it did at the beginning of the year" — and were covered again when it fell apart, this time because "their risk teams came to them and said, look, this happened once already this year."
- 32:03 The conclusion: "the only thing that's really affected the price has been what your risk team is allowing you to short. So, that's going to flip… And you're at a $90 price without any of that having happened yet."
33:05 Positioning — sold gold in January, bought crude, and nothing has changed
- "We did sell down a lot of our gold exposure back in January. We felt that it had run. We remain long-term gold bulls, but that doesn't mean you can't get a fairly dramatic correction."
- The crude buy was not a war bet: "obviously not expecting a war to break out, but rather because the big growth engine, which is the US shales… was slowing sharply" into a market "the market was unduly pessimistic towards" that "was going to slip into deficit."
- 34:08 How he owns it now, "maybe tweaked a little bit around the edges": Canadian oil sands ("good long-lived assets in a friendly jurisdiction"), US shale where quality reserves can still be found ("not enough to build a whole portfolio around the way it was a few years ago"), and offshore drilling — "in a massive bottoming process," cleaned-up balance sheets, industry consolidation, and the biggest swing factor in the portfolio each quarter.
36:17 Gold — the contrarian test, and the silver sell signal
- The sentiment check: "when we announced in January we were selling our gold and buying oil stocks" there was "a loud audible groan from everybody. And today if I were to sell all my oil stocks and buy nothing but gold and copper, people would just put me on their shoulders and give me a parade."
- 37:28 The signal that got them out: "the silver sell signal, which is that silver lags and lags and then it stages this huge catch-up rally. And when it does that, it's usually time to step away from both." Historically precious metals then "fall about 40% and it takes about two years to bottom."
- 38:06 Where we are: gold is down ~20%, "about half of what we've seen in the past" and "on par with the shallowest pullbacks after a silver sell signal" — one instance in 50 years. "If you go off the average, it's closer to 40%," and this one has run 6–7 months, not one to two years.
38:54 Why the gold correction isn't finished — and what would change it
- The fundamental that hasn't reset: Western speculative ETF demand piled in during January and "the holdings of the different ETFs… are still quite elevated. They've come down some, but not nearly as much as they have in past corrections. So there's more metal to be sold there."
- 39:32 The rates worry, inverted: gold peaked "with everyone expecting Warsh to come in and deliver a bunch of cuts," so a cut could have sold off on the news. Today the reverse setup is not yet in place — "if Warsh were to come out today and announce… a surprise 25 basis point hike, I don't think that gold would rally." The buy point is when the market prices two or three hikes and even a hike triggers short covering.
- 41:39 "The gold to oil ratio is still very much in oil's favor." The single override: "if central banks were to double or triple their gold purchases, that could obviously overwhelm ETF liquidation… but I don't see it happening just yet."
42:30 Uranium — a record term price against a 30% equity drawdown
- "People were really willing to call the uranium rally over earlier this year. There was a big pullback in uranium stocks, 30-odd percent, on absolutely no news whatsoever." Spot fell — but "the term contract price, which is where 90% of the market transacts, made an all-time high in the second quarter. It's 95.50 a pound," breaking the 2008 nominal peak by 50 cents (still below it in real terms).
- 42:52 On the M&A chatter: "some interesting news rumors around NexGen and BHP in the last couple weeks that there might be something going on there. But that project still remains a number of years away… that's a big project, the Rook I, to bring online" — so "there's not much in the way of new mine supply to bail the market out," and utilities remain "very under covered in their long-term contract books."
- 43:31 The shape of the trade: a long-term uptrend "punctuated by these periods of hedge fund and retail enthusiasm that pushes prices up and then they just pull all their money out." His advice at the end of one of those flushes: "just buy it and put it away and enjoy the uptrend… very, very bullish and very powerful for the next 10 years."
43:57 Coal — the four-letter word with two near-term catalysts
- "The other thing that people don't look at at all is the global coal industry. It's obviously a four-letter word, completely starved for capital, no one cares."
- 44:37 Europe's winter gas shortfall — "very unlikely to be meeting their gas requirements going into the winter," so coal stockpiles get rebuilt as a reserve, exactly as in 2022 when "European coal burn rose that year to try to save and ration the more scarce and precious gas."
- 45:09 Indonesian export curbs — Indonesia "has been to coal what… shale has been to oil and natural gas," the swing producer of the last 10–15 years. Curtailing exports "would be no different than if the US said we're not going to export shale anymore… a big blow to the seaborne market."
3. In plain English
A jargon-free summary of the thesis behind each call — what it actually is and why he holds that view. (Plain-language companion to the table above; renders on each name's consolidated page.)
Crude oil — the molecule Positive
The puzzle he is answering is simple: about a tenth of the world's oil production has been switched off since March because tankers can't leave the Persian Gulf, and yet nothing has broken. Most people conclude there must have been far more oil sloshing around than anyone thought. Rozencwajg's answer is that the shortage is real, but it has been recorded in the wrong place.
Here is the mechanism. The world reports "demand" not by measuring what people burn, but largely by measuring how much crude goes into refineries — because normally refineries only run when there is a customer, so the two match. This year they came apart: refineries across the Gulf, in China and in Russia stopped running for reasons that have nothing to do with customers (war damage, trapped cargoes, a policy decision). Six million barrels a day of refining vanished, and the models dutifully reported five million barrels a day of "lost demand." Meanwhile people kept driving and flying. So the missing fuel is coming out of storage tanks nobody counts.
Why that matters for the oil price: crude at roughly $90 has never exceeded its April high even though the physical situation has got steadily worse, and the whole year's price action has been about how much risk committees let traders short, not about anyone being bullish. When the world eventually has to refill its fuel tanks, refineries will have to bid hard for crude — and there is still a real chance of "tank bottoms," a market with no cushion left at all. "Where do prices go from here? I suspect higher."
Canadian oil sands Positive
Oil sands are exactly what they sound like: deposits in Alberta where the oil is mixed into sand, expensive to get started but, once built, producible for decades at a fairly steady rate. That is the whole appeal — "good long-lived assets in a friendly jurisdiction." A shale well is largely depleted in a couple of years, so a shale company must keep drilling just to stand still; an oil sands project keeps producing without that treadmill.
The second half of the argument is political rather than geological. In a world where a strait can close and cut off a tenth of global supply, where your barrels physically sit matters as much as what they cost to produce. Rozencwajg — a Canadian who has lived in the US for 24 years — argues that despite the recent friction, "Canada and the United States are still good places to do business" when you compare them with the rest of the oil-producing world. This has been his largest oil exposure since before the war and he hasn't changed it.
US shale producers — selective Positive
American shale was the engine that supplied nearly all of the world's oil production growth for fifteen years, and its slowdown is the reason G&R bought oil in January — before any war. The core of the bull case for crude is that this engine "was slowing sharply" while the market was still pricing it as an infinite backstop.
That creates an awkward split for the equities. Individual shale companies can still be attractive: "US shale producers, where you can find them, still offer really attractive assets." But the qualifier is doing heavy lifting — "we're depleting them rather quickly and it's hard to find really good quality reserves anymore," so this is now a stock-by-stock hunt rather than a sector you can simply own. "Not that it can't be done, but that's maybe not enough to build a whole portfolio around the way it was a few years ago."
Offshore drilling Positive
Offshore drillers own the rigs that drill wells at sea and rent them to oil companies by the day. They don't own the oil, so they are a bet on activity: when producers need new fields, day-rates and utilisation rise together and profits move violently. They are the classic late-cycle way to own an oil shortage, because a shortage eventually forces companies back into expensive deep water.
The reason he likes them now is a cleaned-up industry rather than a hot market. "Most of the companies went bankrupt during COVID and so they all wiped their balance sheets clean" — the debt that killed them last cycle is gone — and the survivors are consolidating. He describes them as "in a massive bottoming process," with early positive data and "very very attractive risk-adjusted returns."
He is unusually candid about the cost of holding them: "our offshore names will either be our biggest contributors or biggest detractors to performance in any given month or quarter," and "in drawdowns they definitely pull back hard." A position you have to be able to sit through.
Uranium Positive
There are two prices for uranium. The spot price is a thin market where relatively little changes hands, and it is what the headlines and the stock prices react to. The term price is what utilities actually pay when they sign multi-year supply contracts, and "that's where 90% of the market transacts." Earlier this year spot fell hard and uranium equities dropped "30-odd percent, on absolutely no news whatsoever" — while the term price quietly made an all-time high in nominal terms at $95.50 a pound. That divergence is the whole opportunity.
The reason the term price keeps rising is that utilities are "still very under covered in their long-term contract books" — they have committed reactors and not enough contracted fuel — and no meaningful new mine is arriving to fix it. Even the biggest development project in the world, NexGen's Rook I in Canada, "still remains a number of years away."
His practical advice is about temperament rather than analysis. Uranium is "one of the simpler stories," a decade-long uptrend "punctuated by these periods of hedge fund and retail enthusiasm that pushes prices up and then they just pull all their money out." The way to own it is to buy after one of those flushes and stop watching: "just buy it and put it away and enjoy the uptrend."
Global coal Positive
Coal is the investment nobody will be seen holding — "obviously a four-letter word, completely starved for capital, no one cares." That is precisely the setup he looks for: an industry that cannot raise money cannot build new supply, so any increase in demand goes straight into the price.
Two specific things could supply that demand in the next few months. First, Europe looks "very unlikely to be meeting their gas requirements going into the winter," and the 2022 playbook after Russia cut the gas was to burn more coal in order to ration the scarce gas — so European utilities start stockpiling. Second, Indonesia is looking to ban or sharply limit coal exports. Indonesia is to seaborne coal what US shale is to oil — the swing supplier of the last fifteen years — so curbing it "would be no different than if the US said we're not going to export shale anymore… a big blow to the seaborne market."
He names no company; the call is that the commodity and the sector are mispriced because nobody is looking.
Refined products — diesel, gasoline, jet Neutral
A refinery buys crude and sells fuel; the gap between the two is called the crack spread, and it normally runs $10–20 a barrel. Today it is around $100 "and they're staying there." That single number does the analytical work in this interview, because it discriminates between the two possible stories. If the world had genuinely stopped consuming fuel, crude and fuel would both be cheap and the gap would be normal. A gap this wide means fuel is scarce while crude is not — "that's the sign of a refining problem."
Follow that through and you get the alarming part. If refineries are processing 6 million fewer barrels a day while consumption is unchanged, the world is short 5–6 million barrels a day of finished fuel, and it must be coming out of storage. But refined-product storage is the one thing that cannot be measured from space: crude tanks have roofs that float down on top of the oil, so a satellite can read their fill level from the shadow the wall casts; fuel tanks have fixed roofs and give away nothing. The official estimate is that non-OECD product stocks are down 50 million barrels. "If we're right, it might be down 500 million" — and fuel shortages in Bangladesh, India and the Philippines are already in the news.
So why is this only a Neutral? Because being right about the shortage doesn't make products the thing to own — the $100 crack already says everyone knows. The gap should close with crude rising toward fuel, so he owns the crude side.
NXE — NexGen Energy Neutral
NexGen is a Canadian development-stage uranium company. Its Rook I project in Saskatchewan's Athabasca Basin is the largest undeveloped uranium deposit in the Western world — but it is a project, not a mine: nothing has been produced yet.
It comes up here as a fact rather than a recommendation. There have been "interesting news rumors around NexGen and BHP in the last couple weeks that there might be something going on there" — market chatter that the world's largest miner may be circling. Rozencwajg doesn't rate the shares and doesn't speculate on the deal.
His point is the opposite of a takeover story: even this, the flagship project, "still remains a number of years away… that's a big project, the Rook I, to bring online." Which is exactly why the uranium term price keeps making highs — there is "not much in the way of new mine supply to bail the market out," so a shortage cannot be solved quickly however much money arrives.
Gold & precious metals Negative
G&R sold most of their gold in January and bought oil stocks with the proceeds, and this interview is about why they haven't bought it back. They are still long-term gold bulls — the argument is about timing a correction, not about abandoning the metal.
What got them out was a pattern they call the "silver sell signal": silver tends to lag gold for a long stretch and then stage a violent catch-up rally, and when it does, that has historically marked the end of the move for both. What typically follows is a fall of about 40% taking one to two years to bottom. So far gold is down about 20% over six or seven months — "about half of what we've seen in the past," and far faster.
The confirming evidence is who still owns it. Western investors piled into gold ETFs in January, and those holdings are "still quite elevated. They've come down some, but not nearly as much as they have in past corrections. So there's more metal to be sold there." Sentiment says the same thing: selling gold to buy oil in January drew "a loud audible groan," while doing the reverse today would earn "a parade."
The other worry is interest rates. Gold pays no income, so it competes with government bonds; the market topped when everyone expected Warsh to cut, and the buy point comes only once investors have swung all the way to pricing in hikes. One thing could override all of it — central banks doubling or tripling their gold buying would swamp the ETF selling — "but I don't see it happening just yet."
Summary & timestamps derived from the public YouTube video (transcript in transcript.html) for personal study. Not investment advice. © Investing News Network / Adam Rozencwajg & Goehring & Rozencwajg for source material.