Description

This is my application idea, written in June. I updated it a little. I write on a substack (www.marketweasel.substack.com) where I cover uranium and uranium equities more in depth, as one of my ongoiing longs.

1. Introduction and Overview:

The decade post Fukushima is littered with the skeletons of investors calling for the uranium bull market. Keeping that in view, I will give my take on why a strong move in the metal and equities is close, and why Global Atomic is an undervalued way to play it. I will first explain some industry basics. Then, to justify my view that uranium is in a bull market, I will break down the supply and demand situation and show, with a margin of safety, the case for a significant and sustained deficit. I will then discuss my view of the market over the next year. Finally, I will value Global Atomic.

2. Uranium industry basics:

Uranium is primarily used as fuel for nuclear reactors.

2.1 The fuel cycle: To get from mine to reactor, uranium ore (U3O8) is mined and milled, then converted to UF6 (conversion), which is separated into 2 streams of higher and lower concentration (enrichment), and higher concentrate is fabricated into fuel. Whole process takes 18-24 months.

2.2 Spot and term market: Uranium is purchased via the spot and the term market. The spot market is for delivery within a year. The term market is everything else. Most uranium is sold on the term market via 3-15 year contracts. But spot matters as contract prices are often spot price referenced.

2.3 Primary and secondary supply: Primary supply is mined uranium. Secondary supply is everything else: inventories, underfeeding (explained later), down-blending of military warheads, recycling, and tailings.

2.4 Economics of nuclear: Levelized cost of energy (LCOE) is sum of costs over lifetime divided by sum of electricity over lifetime, and is an estimate of the true cost of energy. Capital is at least 60% of nuclear LCOE. Fuel cost is 34% of opex. Uranium is 41% of fuel cost. So uranium is ~5.6% of nuclear LCOE. Hence, demand for uranium tends to be inelastic. As nuclear is capex heavy, opex light, fuel security supersedes price as offline nuclear plants are a huge financial opportunity cost.

 3. Breaking down Uranium demand:

This section will project out annual uranium demand to 2030. Primary demand from reactors, and secondary demand from overfeeding, inventory building and speculation are all increasing.

3.1 Primary demand from reactors

I estimate primary demand for all major regions with 2021 figures from World Nuclear Association (WNA), which has country by country breakdown of reactor count, energy output and uranium consumption. For context, WNA estimates 2021 global reactor demand for uranium to be 162M lbs. Note that these figures are not meant to be down to the lb accurate, but I build in a margin of safety to account for inaccuracies.

(United States) The US has 93 reactors (95.5 GW) consuming 45.7M lbs annually. The US story is a prevention of decline story. Post Fukushima, support for nuclear declined while cheap fossil fuels squeezed nuclear operators. But in 2021, the politics reversed. 700M in subsidies saved Exelon's Byron and Dresden, and a 6B lifeline was extended to save nuclear plants. 88/93 reactors have received 20 year extensions. Fossil fuel prices have risen, making nuclear more competitive. However, an expansion of capacity is unlikely as US nuclear is mired in regulation and high cost.

Conservative estimate: slight decrease to 42M lbs by 2030.

(China) China has 54 reactors (52 GW) consuming ~24.9M lbs annually. It is by far the biggest growth story. China builds reactors cheap and fast, at 5B (<50% cost in west) and 2 years per 1 GW reactor. As fossil fuel prices rise, nuclear is increasingly cost competitive. China is planning 150 new reactors, reaching 120-150 GW by 2030 and ~202 GW of capacity by 2035. For 2030, at 120 GW, annual consumption would be 120/52 x 24.9 = ~57.5M lbs. The initial load of new reactors is 2-3x annual demand, coming to 68/52 x 24.9 x 2.5 = ~81.4M lbs. For 2035, at 202 GW, applying a 10% discount for fuel efficiency improvements, annual consumption would be 150/52 x 24.9 x 0.9 +24.9 = ~89.5M lbs, with initial load 150/52 x 24.9 x 2.5 x 0.9 = ~161.6M lbs. Even if they only build 100 out of 150 reactors, it will be ~68M lbs, with initial load ~107.7M lbs.

Conservative estimate: 57.5M lbs by 2030 with initial load 81.4M lbs, 68M lbs by 2035 with initial load 107.7M lbs.

(EU+: European Union+Switzerland+Ukraine+UK) EU+ has 135 reactors (121.5 GW) consuming 46.1M lbs of uranium annually. EU+ is a growth story. Post 2011, nuclear programs were slowed/halted/reversed. This is reversing due to an energy, geopolitical and climate crisis. There are 16 reactors under construction/planned, and 33 proposed.

Conservative estimate: slight increase to 50M lbs per year by 2030.

(Russia) Russia has 37 reactors (27.7 GW) consuming ~15.4M lbs of uranium annually. Russia is a growth story. Rosatom targets 24 new reactors in Russia by 2045.

Conservative estimate: maintain at 15.4M lbs by 2030 (assume none of the 24 reactors are built by 2030)

(South Korea) South Korea has 25 reactors (24.4 GW) consuming ~11.8M lbs of Uranium annually. It is a growth story. South Korea has 3 1.4 GW reactors under construction. 2 will be completed in 2023, 1 in 2024. South Korea will add 4 more reactors by 2030.

Conservative estimate: slight increase to 13.8M lbs by 2030 (account only for the 3 reactors built by 2024)

(Japan) Japan is a reversal story. Japan had 54 reactors which were all shut down after 2011. 33 reactors (31.5 GW) remain operable, with only 10 restarted by 2021. In a reversal, PM Kishida pledged in 2022 to an emergency restart of all nuclear reactors.

Conservative estimate: ~11.4M lbs per year by 2030 (assume 15 of 23 remaining reactors restarted for a total of 25 operating, no new builds)

(Rest of World) Rest of world currently uses 14.4M lbs annually. Many countries, such as UAE, India and Vietnam are constructing/planning nuclear power plants. Russia and South Korea have confirmed/planned 30 nuclear plants for export.

Conservative estimate: 16M lbs by 2030.

3.2 Increase in secondary demand from overfeeding

What is under/overfeeding? In enrichment, for a given time period, you can produce a given output with more enrichment capacity (termed SWU) or UF6 input. An analogy is juice. To make 5 orange’s worth of juice in a minute, you can use a lot of strength (SWU) on 5 oranges, or less strength on 10 oranges. When SWU is relatively expensive, you use more uranium (overfeeding); when its relatively cheap, you use less (underfeeding) and sell leftovers (secondary supply).

Due to excess SWU post Fukushima, market was underfeeding and generating 20M lbs of secondary supply. However, the Russian-Ukraine war led to self sanctioning from Russia (43% of global SWU). Utilities redirected purchases to western enrichers, leading to SWU shortage and overfeeding. Orano (12% of global SWU) is reporting some facilities reaching 100% capacity. SWU prices, which bottomed at $35 in 2018, are now at $132 (peak was $160 in 2010).

Conservative estimate: switching from underfeeding overfeeding could see 20M lbs of supply turn into 10M lbs of demand (other estimates are much higher, in the 30M lb range).

3.3 Secondary demand from inventory building

Uranium demand is pro-cyclical: when there is perceived threat to security of supply, utilities build excess inventories. In the last cycle, utilities built huge inventories, allowing them to operate despite mines under-producing reactor demand from 2011-2020. Where are inventories now? WNA estimates US inventories at 141.7M lbs (~3x annual demand), a figure backed up by Sachem Cove (which I consider the most credible fund in the space). IAEA estimates ~3.5x annual demand for European utilities. Chinese inventories were estimated at ~400M lbs in 2021. That sounds like a lot, but assuming a linear growth to their 2030 target and summing the arithmetic progression, China would need 493.4M lbs cumulatively to 2030. It has few oil, gas or uranium resources, and will probably want an even bigger inventory for security. Overall, inventories are low, and with an impending shortage and geopolitical risks, utilities are very likely to start building excess inventories.

Conservative estimate: Increasing inventory by 1x 2021 reactor demand (less than previous build ups), delivered over a 10 year contract, will create 16.2M lbs of annual demand to 2030.

3.4 Increase in financial demand from speculators

Commodity funds have started cornering the space. Sprott Physical Uranium Trust and Yellowcake have bought 56.4M lbs of uranium from Jul 2021 to Jun 2022. Asian funds like ANU will also bring in Asian speculators.

Conservative estimate: It’s hard to put a number on this, but likely will go significantly higher.

3.5 Summary

With conservative estimates, we see an increase from 162M lbs reactor demand and 20M lbs underfeeding supply in 2021 to 206.1M lbs reactor demand, 10M lbs overfeeding demand and 16.2M lbs inventory build demand by 2030, not including financial demand and reactor initial loads.

4. Breaking down Uranium supply:

This section will project out annual uranium supply up to 2030. Primary mine supply is depleting while secondary supply is also likely to stay the same or decrease.

4.1 Secondary supply from down-blending Highly Enriched Uranium - HEU

US has estimated 170M lbs U3O8 equivalent of HEU stockpiled. Of that, 146M is for military use (very unlikely to be touched), 24M are for emergency civilian use. Even if all 24M lbs were down-blended for reactors (unlikely), that is not a significant source of supply, although I would not be surprised to see all of it downblended, with the administration already having sold a large portion of the strategic petroleum reserve in the name of "fighting inflation".

4.2 Secondary supply from recycling nuclear waste and re-enriching tails

With self sanction on Russian SWU, enrichment capacity is limited and re-enrichment not practical. New laser re-enrichment technology from Silex, an Australian company, could generate 4-5M lbs of annual supply max, but not enough to derail the thesis.

4.3 Secondary supply from underfeeding

Underfeeding supply will dry up, as discussed previously.

4.4 Secondary supply from inventories

US, Europe and Chinese inventories have been discussed above. The last inventory seen as potential overhang is Japan. Japan inventory is ~100M lbs, which 6.6x of total operable fleet annual demand, and 5x if you include the 11 reactors constructing/planned/proposed. That is not enormous. Japan is resource poor, and is actually looking to increase inventory for energy security (In 2022, Marubeni and Kazatomprom discussed joint ventures in uranium mining).

4.5 Primary supply from mines

Uranium exploration budgets fell off a cliff after Fukushima, falling from ~1.2B in 2008 to ~100M in 2020, and has not recovered. No new major mines are due. Sachem Cove estimates 33M lbs of mine supply will be lost to depletion by 2029.

4.6 Summary

Mine depletion and loss of underfeeding will reduce supply by ~53M lbs by 2030, with little supply from other sources to make up for it. Conservatively, we can model a drop of 40M lbs.

5. A sustained deficit is the base case

Summarising sections 3 and 4, we can see that under conservative assumptions, uranium demand will increase to ~230M lbs and supply reduce by 40M lbs per year by 2030. There is no imminent solution, and deficit is likely to be sustained. Arrow, the largest known undeveloped deposit, could produce up to 28.8M lbs per year, and is likely at least a decade away from full production. There are no other major discoveries.

6. Market in the near term

The market is already tightening: Spot, mid term and long term U3O8 are between $50-53, up from $32.40 in Jul 2021. Cameco contracted 30M lbs in 2021, but 40M in Jan 2022 alone. I believe contracting will increase and prices are going higher.

The Russia Ukraine war has tightened the nuclear fuel cycle, and the shock will hit the U3O8 market soon. US congress is meeting in July to discuss banning Russian uranium. Western ships are given one year passes to allow transport of Russian fuel. With fuel security threatened, utilities have been buying UF6 and SWU as they are nearer to the end of the fuel cycle and take less time to convert into fuel. Long term SWU is $132 (from $61 in 2021). Spot UF6 is $166 (from 2017 lows of $50). Also, conversion capacity was a bottleneck for U3O8 demand as only so much could be converted at a time. Conversion is now at all time high of $37/kgU. Converdyn (11.3% of global conversion capacity) is restarting early 2023. Orano is ramping conversion from 8600 tons in 2021 to 13000 tons of UF6 in 2022 (increase of 7% of global conversion capacity). U3O8 is the input to conversion and demand will increase soon.

Furthermore, US and EU inventories are ~3 years on average. As the nuclear fuel cycle takes 18-24 months, they have to contract within 12 months or risk shutdowns. As these are averages, some utilities are below 3 years, and have to contract even sooner.

Supplier discipline will keep prices high in short term. Kazatomprom (largest producer) committed to 2/3 capacity from 2021 to 2023. Cameco (2nd largest producer) plans to operate 40% below capacity by 2024.

Long term, prices will rise to the marginal cost to cover all needs. Pre-Covid (before inflation, nuclear turnaround and supply disruptions), Sachem Cove estimated 40% of rector needs from 2020-2030 to be unmet at $50 per lb. Eventually, all current deposits, some profitable only at ~$100 per lb, will be needed.

7. Global Atomic

*$ = USD, C$ = CAD

7.1 Overview

Global Atomic (GA, OTCQX:GLATF, TSX:GLO) is a zinc producer and uranium developer with exploration upside. It has 174.9M shares outstanding, 7.8M warrants at average exercise price C$4.77, 12.5M stock options at average exercise price C$1.085, for fully diluted 195.2M shares. Market cap is $393.6M/C$512.2M. It has C$92.8M in total assets (C$24.2M cash/C$54.5M property, plant and equipment/C$9.4M in Zinc joint venture), and C$4.1M in total liabilities.

7.2 Zinc production

GA owns 49% of a plant in Turkey which recycles steel dust from local steel mills into zinc. Total capacity is 110k tonnes of steel dust for 60M lbs of zinc. In 2021, it processed 70k tonnes for 34.8M lbs of zinc at average realized price of $1.36 per lb for net income C$4.1M. Zinc prices wil go higher as zinc smelters in Europe are shutting down due to the energy crisis. Growing Turkish steel production and protectionist policies will bring the plant to full capacity (110k) in Q2 2022. Conservatively, we model 100/70 x 4.1 = ~C$5.86M net income for coming years.

7.3 Uranium development

GA owns 90% of the Dasa deposit, which has >250M lbs U3O8. They are permitted and are constructing the Phase 1 mine, which will begin production in 2024 and yield 45.4M lbs of U3O8 over 13 years. Phase 1 has $208M capex and $21.93 per lb post tax all in sustaining cost (operating costs+sustaining capex+G&A+exploration costs). Production will peak at 6M lbs U3O8 in 2028.

(Valuing DASA)

According to the feasibility study, at $70 per lb, net present value at 8% discount and internal rate of return of Phase 1 is $885M and 68.7%. To adjust AISC for $70 per lb to account for taxes (assuming flat standard corporate tax of 30% in Niger), we get pre tax cost by solving:

$21.93 = pre tax cost + tax deduction - (1)

tax deduction = (35 - pre tax cost) x 0.3 - (2)

It would yield a pre tax cost of $16.33. Cumulative cash flow and peak cash flow in 2028 would be $1.7B and $225.4M. In the last cycle, without an imminent supply deficit, contracting price peaked near $100, and Japan contracted a large portion of material between $80-120. These numbers are not inflation adjusted. I regard $70 per lb to be conservative, especially since a Dasa begins seriously producing in 2024, where the supply deficit is expected to increase sharply.

Furthermore, Phase 1 only represents 20% of DASA’s total resource of 250M lbs. Concentrations >750ppm are usually considered ore, and Dasa has ~147.9M lbs above 1200 ppm at average 4483 ppm. Management found that grades were consistent across Dasa and NPV of whole project is likely 5x of Phase 1, with total mine life 50 years.

(Financing)

A fear is that significant dilution will be needed to fund DASA’s $208M capex.

Management is aligned (management+board ~15% ownership, CEO 7.6% - highest in industry) and is trying to debt finance the project. GA just formed a financing syndicate of North American financial institutions including Canada’s Export Development Contract which earlier unilaterally signed a letter of interest to provide up to $75M in financing at typical greenfield mining rates. GA just received a letter of intent from a US utility to buy 2.1M lbs over 6 years starting 2025, with revenue potential >$110M. These and upcoming contracts will help underwrite debt financing. Exercise of 2.85M warrants with exercise price <C$3 will bring in C$8.5M with slight dilution. It is likely Phase 1 will be financed by zinc operations, current warrants and debt only.

(Risks)

- Jurisdiction risk: Niger looks dangerous. But in fact, they are a top tier jurisdiction, being the world’s 5th largest uranium producer, with 50 years of uranium mining history. It has infrastructure, excess milling capacity and trained workforce from depleted mines (e.g. Akouta mine). The mine is deep in the desert, surrounded by NATO bases, and uranium can only be sold to utilities, making it an unattractive target. China is looking to acquire uranium mines, and Niger is a prime location vs hostile western jurisdictions. Finally, all jurisdictions are risky (e.g. permitting is much harder in the west due to NIMBY).

- Incorporation of Niger mining company: To start mining, a Niger entity has to be incorporated. The government can purchase (not seize) up to 40% (currently 10%) of the asset. However, in Aug 2021, the Mines minister issued a formal letter stating it would not seek to acquire beyond 10%. Doing so would greatly damage Niger’s 50 year record as a uranium mining jurisdiction.

- Cost inflation: Will impact the whole sector and raise the entire cost curve, and is likely passed on to consumers.

- Currency risk: Operations are completely in USD

- Operational risk: Possible but management is skilled and Niger has expertise and infrastructure.

7.4 Uranium exploration

GA holds six uranium exploration permits in the Republic of Niger covering 729.8km^2. Work carried out by Exploration since 2007 has found resources in 3 areas ex DASA: Tin Negouran (10M lbs), Dajy (17M lbs) and Isakanan (34 M lbs).

7.5 Summary

Take $70 NPV at 8% discount of $885M, subtract (208-24.2) = ~$183.8M of debt financing, we have $701.2M, which is ~1.78x market cap of $393.6M. Shares would only be worth 10% less with full dilution (which would also bring $39.05M proceeds).

This does not include the zinc asset (~$4.5M annual net income), 80% of DASA’s uranium (estimated ~147.9M lbs economic ore above 1200 ppm at average 4483 ppm), further exploration and other discovered resources (61M lbs so far), and pricing upside due to the enormous uranium deficit.

At these valuations, Global Atomic is a strong value play. The worst bear case is a huge acquisition by Niger government.

Catalyst

- Rerating of equities due to increased contracting activity and upward moves in the spot and term market prices

- Conclusion of debt financing to fully fund Phase 1 capex

- Conclusion of Niger entity incorporation with no or moderate increase in stake by Niger government

- Smooth continuation of construction and good progress to production by 2023.

- Entry into production by 2024