climb aboard my submarine, uranium aquanauts
*record scratch * I was going to write about HALEU this week, but even I’m not a big enough jerk to withhold the goods after the week that was.
Two weeks (and a few trumpet blasts) into the Age of Sprott and speculations are flying high. $55 by next week! $200 by year-end! Retail guys screaming about collecting utility scalps like they’re Aldo f***ing Raine! [Thankfully this particular scalp is hiding deep inside flyover country – make sure to stock up on provisions first before venturing into the heartland.]
I have done by best to stay off the bird app this week, for my sanity and for yours. I have been screaming about big capital entering the physical space since 2020, and it’s not very much fun being right about it. So I’ve tried to tune out the noise a bit and focus on what I can contribute to the rapidly-increasing discourse. Except for SnowyWindows, whose handiwork I present to you below:
The Fukushima accident triggered a freefall in uranium price which lasted (arguably) for over six years until the price bottomed out in 2017. Quite a bit of contracting was done on the way down, too, so combined with the over-contracting of the 2005-2011 period, there was too much uranium. But the pounds kept stacking up – mines produced pounds without a destination, offtakes arrived and clogged up traders’ books, utilities with less-than-expected uranium requirements continued to take delivery and looked for ways to ease the ol’ balance sheet. You can see the inventory among US end-users and suppliers trickle up opposite the declining price from 2011 through 2017 in the chart below. There’s some argument to be had here on how the Megatons to Megawatts Russian imports should be considered, but suffice to say inventories did increase from 2011 to 2017 and there is a myriad of reasons why.
A better uranium historian than I would know when exactly the carry-trade first came to uranium, but the necessary combination of ingredients came about in the years after 2011. Uranium was cheap, interest rates were historically low, and desperate sellers wanted to get unclaimed supply off their books – all of which combined to create the forward curve, a spot price-driven formulation of forward uranium prices. [You can see an example of Numerco’s forward curve in the embedded tweet below.]
Bidding into any term contracting for end-users would require competition on the forward curve (carry-trade’s inventory plus financing), so as long as pounds were available and the utility’s credit was good, the forward curve set a harsh reality for miners. In fact, as European interest rates slipped into sub-zero territory, the carry-trade became even more competitive as certain financiers sought high-credit receivables to garner better returns on restricted funds. Call this market dynamic inventory-driven. Money was made in the trading of pounds, sure, but also in the margin that traders took on the financing itself.
A funny thing happened in late 2020, though – the miners pushed back. It wasn’t on every request for proposal and it varied depending on the term of supply, but producers began to win contracts again at prices they could live with. With the COVID-driven pop in uranium price, the forward curve’s spot plus financing (say, three years at 5%) started looking profitable for some sell-siders. Had Sprott not entered the fray, this dynamic may have had the chance to play out, incrementally bringing the market back into balance one transaction at a time. Think of this market dynamic as balanced, at least in terms of the forward curve driving supply-side competition, and supply-side competition pushing back on the forward curve. Still not at incremental production costs, but heading that direction.
The third dynamic, most evident near the highest highs around 2007, I call production-driven. At $140/lb or thereabouts, many aspirational producers volunteered to deliver long-term, price-protected supply at lower prices than the going spot rate. Supply disruption drove the gotta-have-it-now spot price, but in the 2008-2011 period there was no shortage of future supply at acceptable prices. Not every producer believed that price would come back to earth (some leveraged their books entirely to spot and experienced the full downside post-2011), but the quoted “long-term” rates during the 2006-2011 period suggest that deals were to be had if you were willing to wait. It’s possible that such a dynamic is ahead of us.
So, to summarize, the transition between an inventory-driven market and a balanced market should occur at some point around production costs (for at least some producers). There’s a lot of room in between balanced and production-driven, so we can think of that transition occurring when there are more available producers than forward demand. Spot price is critical to price formation in the former two, while production economics are critical to the latter.
interlude: bohr’s model of the atom
Modern physics advanced at a blistering pace in the late 19th and early 20th centuries. In a period shorter than Law & Order has been on the air, scientists hypothesized the neutron, discovered the neutron, and then harnessed neutrons to split other atoms to create an atomic blast at Trinity.
Most high school chemistry classes will spend some amount of time discussing the development of atomic theory – the slow climb from a loose concept of the indivisible universal building block, the Greek atomon, to oddities like the plum pudding model, to more complete quantum mechanical models. We understand today that there is a fundamental uncertainty that bounds the accuracy by which we can measure the position and momentum of subatomic particles, and this formulation results in three-dimension electron orbitals rather than circular paths. This was a late-breaking insight when compared to many of the experiments that (in hindsight) form the basis of our understanding of the quantum world.
Bohr and Rutherford created one of these intermediate models – close enough to the truth be notable but not completely accurate. With planet-style orbits of electrons at specific distances from the nucleus, and a quantum (e.g. a fixed quantity of energy) in between neighboring orbits, Bohr’s model explained the discrete spectral lines observed in 19th century spectroscopy experiments. For simple atoms like hydrogen, the energy levels explained by Bohr’s model are pretty accurate (even in the modern context), which is why Bohr’s model of the atom is still broadly taught to students.
But if you want a complete understanding of quantum mechanics, especially for larger atoms, Bohr’s model is lacking. But even without the theory of quantum mechanics, these scientists managed to calculate many key features later explained by more advanced theories and lay the groundwork for decades of research to come.
As Bohr himself said, “Anyone who is not shocked by quantum theory has not understood it.”
Two things cross my mind when I think of these early quantum physicists, lacking the last eighty years or so of scientific achievement to go from:
- They didn’t need the full picture – the deepest of understandings – to accurately explain the experimental data they had at hand. Being wrong about the “complex” cases (and the horror of f-orbitals) doesn’t diminish the convenience of Bohr’s model. So too do retail investors mostly have it right about the uranium market. I might split hairs about what “undersupply” means and minutiae about the uranium requirements of initial core loads and other such 400-level nonsense, but the dude yolo’ing his 401k into SPUT on the way up doesn’t need an advanced knowledge of marginal uranium production economics to see which way the wind is blowing, to try to make sense of the things he sees without understanding the full context. And the broad strokes do carry a certain elegance…
- But on the other hand – don’t underestimate the power of smart dudes. The big utility fuel groups and trading groups are full of experts with significant resources and many tools at their disposal to enforce their will on the market. Three huge state-owned fleets (France [and the UK], Russia, and China) won’t take this one on the chin without some sort of response. SPUT changed the dynamic faster and with greater magnitude than most predicted, but there are ways out of this puzzle box and I fully anticipate market makers to find solutions to today’s complex problems. It just might not happen tomorrow. Most of the puzzle box’s solutions don’t even summon the cenobites!
“Do I look like someone who cares what [the utility rate board] thinks?”
contracting for uranium
So what does a long-term uranium contract actually look like?
A contract is built around a promise of performance – that the seller will deliver a quantity of pounds on a fixed schedule, and that the buyer will pay the agreed-upon price. Long-term contracts span multiple deliveries and will often have built-in flexibilities – allowing for slight deviations in delivery date or quantity (e.g. +/- 10%).
Then there’s a pricing formula:
- Fixed price – the parties agree to fixed unit costs for each delivery in the contract
- Base price-escalated – the parties agree to a real-dollar cost for future deliveries, then escalate that base price to the date of delivery (using a macroeconomic indicator tied to inflation, most often)
- Market price – the parties choose one or more indicators at or near the time of delivery to set the price. These contracts usually carry a slight discount to market price in exchange for locking in the supply, and the free-floating market price often has “caps,” minimum and maximum prices that will be paid even if market prices undershoot/exceed the outer limits. Think of these caps like backdoor fixed prices.
- Re-basers – especially for the most long-term of long-term contracts, price formulas often include some sort of language to ensure that fixed-price components are in the neighborhood of then-market prices (protects utility on severe price upside, and producer on extreme price downside).
Most contracts are going to shake out somewhere in the middle – hybrids between different pricing formulae to create something that both parties can live with. End-users are price-protected, and producers trade maximum price upside for a higher floor.
There are a lot of market indicators to choose from – monthly prices, weekly prices, daily prices, broker average prices, long-term indicators, and mid-term indicators; formulas are going to vary quite a bit based on the beliefs and preferences of the counterparties.
Then there’s everything else:
- Are there options in the contract for additional supply? At what cost?
- What if the reactors shut down early? What if the mine closes?
- What does force majeure really mean, anyway?
So imagine that all of the forward deliveries signed to date – prior to the COVID and SPUT price pops – exist in a superposition of the price at date of signature, the sentiment about the forward price at the date of signature, and the actual market price at the time of delivery. This weighted-average price (as seen in places like the EIA Uranium Marketing Annual Report) is the “real” average price of in-year deliveries, a product of the past and the future written onto the present uranium market. Through the late 2010s, deliveries were made under the most bullish of pre-2011 contracts and the most bearish of discount bin carry-trades, with a fair amount of depressed spot price mixed in for taste. The delivery price will look different from here on out – encompassing low prices of the past, higher market prices at time of delivery, and perhaps a more bullish sentiment on future prices.
interlude 2: the perfect manhattan
Prior to my turn in the nuclear industry, I was a bartender. Two dashes of hopelessness after “leaving” (euphemism) my PhD program and one dash of the post-Fukushima nuclear industry job market malaise put me behind the stick 70 hours a week, splitting time between a local steakhouse and a white table cloth Italian joint. I miss it more than I would have expected.
There’s a certain serenity in beginning your day juicing dozens of limes, lemons, and grapefruits for evening service rather than logging on to check my email. In taking the chairs down and removing plastic wrap from bottles. A few moments of silence before you unlock the doors and the 30 or so inoffensive songs on the playlist begin their inexorable march forward. Of ice chests filled to the brim and of ingredients freed from overnight storage, standing ready for duty.
American cocktail culture underwent a bit of a renaissance in the mid-aughts, with bartenders resurrecting century-old recipes and growing big mustaches to accent the speakeasy vibe. If NYC’s famous cocktail haunt Death and Co. (opened 2007) is Metallica’s Reload album in this analogy, cocktail culture had already progressed to its St. Anger phase by the time I picked up the tins in 2014. Cocktails started popping up on the menus of fast-casual places and beer bars; five-ingredient derivative monstrosities haunted the nouveau américain restaurants of land, with garnishes looking more like a Calder mobile than an edible arrangement. Most of the time I’d be happy to settle with the cocktail equivalent of Death Magnetic, strong and well-made if not particularly creative. But in the joints I worked at, under the tutelage of a bon vivant bar master who looked a lot like Zangief from Street Fighter, we had a lot of fun.
The best way to become a better bartender, I was told, was to become a better drinker. So I went on a quest to find the perfect Manhattan recipe (but not the Perfect Manhattan recipe, which is itself a riff on the original). Not an easy task…especially given the massive list of variants invented in the late-aughts before the famous Brooklyn variant and its signature ingredient of Amer Picon (or Bigallet China-China) was available in the States. In homage to the legendary Brooklyn, there are like two dozen of New York neighborhood-named Manhattan variants…Greenpoints and Red Hooks and Bushwicks…and so I combed through the recipe books to create a comprehensive list. It was a helluva undertaking (both in the consumption of reading materials and the consumption of brown liquids).
You drink enough rye whiskey, and soon you won’t want to drink anything else (water helps, though). Dozens and dozens of Manhattan riffs later, I think I finally understood the essence of the cocktail. Its degrees of freedom, if you will. Even heresy like the Cleveland Leader, a Gilded Age misprint which called for a Manhattan made with Dutch gin instead of whiskey, has its own place in the canon.
But the best Manhattan is the simplest one – when you find a new spot and a new bartender, when you order the classic and they make it correctly (two of rye, one of sweet vermouth, a few heavy dashes of angostura, stirred on rocks for a reasonable amount of time and strained into stemware). In that moment, anything is possible.
divination for the atomic age
So what is the essence of the uranium market? When things are chaotic, what are the key ingredients and familiar flavor of the good stuff?
The uranium market is perhaps unique among commodities in that there is only a single use – to make power in a nuclear reactor. Uranium is an oddball, a borrower of the same language that cool kids like iron ore and crude oil use but with less financial complexity. An energy-dense yellow powder traded over-the-counter like it’s still the 80s, a pawn in the strategy between great powers but often hailing from locales far afield from the end-users.
But the core essence? Miners dig up the stuff to make a profit.
At $30, are miners making a profit? At $40? At $50? The “right” number will ultimately descend from a real supply/demand analysis and the cost of that incremental pound – and the uncertainty in the future of nuclear demand makes this a compelling question. Greed is an important dynamic in the short term, but competition from marginal producers can shape the long-term price in unexpected ways. Is the spot price disconnected from the long-term price now? Can we rely on the forward curve as the cheapest attainable forward price? If we have transitioned into a production-driven market, what’s next?
The Sprott rock (a meteor, really) has dropped into the lake and the ripples move outward. Here are a few disparate thoughts I’ve had this week:
- Uranium is more expensive now – including the market component of many existing contracts, although because many of those market components have price caps, it’s unclear what effect this will all have on delivered price in the upcoming years.
- Uranium price is more dynamic and volatile than it was a few months ago – it’s hard to pin a price on a constantly-developing story, especially across days of offer validity.
- How much capital will enter the physical market? When will it enter? And when will it leave? It would be folly to predict how this will play out over weeks (let alone months and years), except that I think that the larger and faster SPUT’s AUM increases, the greater the likelihood that it will eventually become a (limited?) source of supply in the future if/when the squeeze trade loses its glamour.
- It’s important to remember that utility requirements are pretty covered over the next few years – and one of the purposes of having inventory at all is to provide price protection in times like these. The heat is on – but the house isn’t yet on fire. It’s certainly a different negotiating environment than it would have been in a slow bleed up rather than this SPUT-driven spike, sure, but there’s a long-term price to be found out there and the next few months will be interesting in this regard.
- The uranium market might be small, but the utility sector is not. And there are truly titanic state-owned utilities in Russia, China, and France who are (to some degree) “losers” in a spot price spike (to say nothing of big investor-owned utilities elsewhere in the world). It’s too early (and very much not my job) to speculate on how these entities might response to a fully-matured supply squeeze, but I don’t want to underestimate the ability of these groups to find solutions to a supply shortage.
- Many non-producers, big and small, have discussed the need for increased prices to bring their “production-ready” or “shovel-ready” sites into production. If the price gets there, will we see announcements of mine (re)starts? How will the market perceive those announcements?
I don’t think there are great answers to any of these questions, except that it will be intriguing to watch it all play out over the upcoming months and years. I don’t want to underestimate the next (and subsequent) phases of Sprott and others raising capital to buy physical, but I also don’t want to understate that we haven’t yet seen a major response from the demand side. “May you live in interesting times” isn’t well-wishing, and here we are in interesting times.
[Programming note: it might seem like I am producing content on a weekly basis, but alas, this is not the case. I am not nearly that prolific, even in this relative lull in activity as I spend more time inside to avoid Delta. I will follow up with something on HALEU in the next couple of weeks, but it might be a while before I dip a toe back into long-form. Your feedback is always appreciated. -808sandU3O8]