Description

Description

Argo Blockchain plc (ARBK) is the new kid on the block of US-listed pure-play Bitcoin miners, and I believe it is the cheapest way to gain leverage to the price of Bitcoin.  The company is the most eco-friendly of the public miners, as the first to become “carbon positive”, and it will soon be the lowest cost, with its West Texas expansion increasing its power capacity from 45 MW to 245 MW (mostly 2 cent / kWh power) by the end of 2022.  With mining margins at the top of the peer group, and hash rate capacity ramping up nearly as quickly as its much larger capped peers, I think the market has simply overlooked Argo in favor of the go-to names (MARA, RIOT in particular).  If it narrows the valuation gap with its peer group to 8X CY22E EV/EBITDA (vs 10X for the peer average), the stock will trade at $33 for a nearly 90% total return.  And with its HODLing strategy, I believe that its BTC holdings alone will be worth more than the market cap by the end of 2023.  While Bitcoin is a volatile currency, and this is a leveraged investment on Bitcoin, I think the risk/reward is truly compelling here.

Background

History, Structure

London-based Argo Blockchain plc (ARBK) was founded in 2017 as a mining-as-a-service (MaaS) company, selling operational and service-oriented packages to miners, and listed publicly on the London Stock Exchange in 2018.  The company quickly pivoted its business model from MaaS to mining on the company’s own behalf, with the goal of accumulating (ie HODLing) crypto assets over time.  Argo issued 8.5mm ADRs (ticker ARBK) on September 22, 2021 at $15.00, with one ADR representing 10 ordinary shares.  With 47M fully diluted shares and net cash of $120M, the enterprise value is about $720M.  Argo currently holds about 1,840 BTC and equivalents on their balance sheet (worth about $110M), of which about 1,480 were mined this year.  

Facilities

Argo operates over 21,000 mining machines across 6 facilities in North America (3 in Quebec, 1 in Kentucky, 1 in Georgia, 1 in North Carolina) with a combined hash rate of 1.075 EH/s (exahash per second), which represents 0.7% market share of the current network hash rate of about 160 EH/s.  The company is transitioning from a hosted business model to owning and operating mining sites, with the build-out of a 200 MW campus in Texas (total investment: $17.5 mm) with options for another 600 MW.  This site (their 7th facility) will be powered by > 90% renewable wind energy at a rate of 2-3 cents/kWh.  Argo has deposits down on another 20,000 machines that will be delivered in 2Q22 and 3Q22, and since they operate at a TH/s that is about double the rate of Argo’s current fleet, the new machines will triple the current hash power of Argo’s fleet.  Between that and some previously ordered machines, the company expects to exit 2021 with a mining capacity of 1.7 EH/s and it should ramp to 3.7 EH/s by the end of 2022.  

Other Assets

In addition to Bitcoin, Argo mines a small amount of ZCash, and the company has stated that if the economics make sense, it could expand its mining operations beyond BTC and ZCash, but I am modeling the company only on BTC mining expansion.  Argo’s longer term strategy is to diversify the composition of revenue through other emerging opportunities, and it has invested in several decentralized finance (DeFi) and crypto related startups, including Luxor Technologies, a hash rate management software platform and equihash mining pool operator, and Pluto Digital, a venture capital and technology company focused on DeFi.  These investments are essentially zero cost options.

The ESG-Friendly Miner

Another differentiator is Argo’s commitment to being ESG friendly.  It is the first publicly traded cryptocurrency miner to become climate positive, which means that it has gone beyond carbon neutrality to actually reduce carbon in the atmosphere.  In March 2021, Argo launched the Terra Pool in partnership with DMG Blockchain Solutions, aiming to accelerate the transition to clean power energy sources.  The Terra Pool joined the Crypto Climate Accord and plans to test new open-source software that encourages Bitcoin miners to verify sources of renewable energy.

Bitcoin Mining 101

I’m not going to go down the rabbit hole trying to demystify the complexities of Bitcoin and the blockchain, but I will try to shed light on the basic economics of Bitcoin mining.  Bitcoin miners use mining rigs (powerful computers) to solve complex computational problems to verify transactions on the Bitcoin blockchain.  What is a blockchain?  It is a type of database that stores large amounts of information across a network of computers.  Unlike a typical database, a blockchain is structured to collect information together in groups known as “blocks”.  Each block has a finite storage capacity, and when filled the block is added to the previously filled block, forming a chain of data known as the blockchain.  Each filled block is duplicated and distributed across the network of computers on the blockchain, creating a timeline of data.  The history of information in each block is irreversible.  The only way to change how the system works or the information stored on it is if a majority of the network’s computer power agrees to such a change, which gets increasingly difficult as the network grows and becomes more decentralized.  

Back to the Bitcoin miners: as a reward for solving these computational problems, miners have a chance to receive newly created Bitcoin.  There is a fixed amount of Bitcoins (currently 900 BTC) that miners can receive as a reward each day, so if the number of miners increases, the number of Bitcoins each miner receives decreases accordingly.  The processing power of both miners and the bitcoin network can be measured by what’s called a hashrate, which refers to the amount of computing and processing power being contributed to the network through mining.  It basically measures how many times the Bitcoin network attempts to complete those calculations each second.  At higher hashrates, the miners are generating more guesses.  More powerful machines and/or a greater number of machines on the network will cause hashrates to increase.  If a miner’s hashrate increases, it should be able to mine more bitcoins.  At the same time, the Bitcoin network automatically adjusts its difficulty about every two weeks to ensure that bitcoins enter circulation at a steady and predictable rate, targeting about 10 minutes per block.  As shown in the chart below, the network hash rate has loosely correlated with the Bitcoin price, although since the beginning of 2019, the price of Bitcoin has increased by 20X from $3K to $60K while the network hash rate has increased only 4X from 40 EH/s to 160 EH/s.  China’s ban on crypto mining earlier this year led to a pretty sharp correction in the hash rate, but the rest of the world has quickly picked up the slack and we’re back near highs.  

Correlation to Bitcoin Price

Ultimately this is a bet on Bitcoin prices.  As shown in the chart below, in the period since 12/31/18, Bitcoin miners (equal weighted mix of MARA, RIOT, HIVE and HUT) shown in the black line have correlated strongly with the Bitcoin price (blue line), although with significantly higher volatility.  If we take the period from 12/31/19 to the peak on 2/17/21, this Bitcoin mining index was up 47X while the Bitcoin price was up just 8X.  

Management Team

• Peter Wall – CEO: Peter is a founding member of Argo and has managed the operations of the company since its inception.  His previous experience includes technology entrepreneurship, journalism and filmmaking.  

• Perry Hothi – CTO: Perry has over 7 years of experience in the blockchain space and is responsible for Argo’s technology and mining strategies.  Previously he was CTO for an omnichannel retail company in the US.

• Alex Appleton – CFO: Alex has over 18 years of experience in auditing and corporate, and 8 years of board level experience in finance and operations roles.  His previous work experience includes roles within large multinational organizations and within the cryptocurrency sector.

Valuation

I believe ARBK should trade at 8X EV/EBITDA (CY22E), or $33 per share, for about 90% upside.  This would be a slight discount to comps (peer average is 10X) which I think is justified by its slower growth rate and somewhat less experienced management team, although one could argue that with its first mover status as the green, ESG-friendly miner, and its low cost power (the Texas facility will provide power at 2 cents per kWh), it should trade at least in line with peers.  

Argo is one of the newer, smaller mining operators with a market cap of about $800M vs MARA’s $6B and RIOT’s $3B.  The only other pure play miner with Street estimates is BITF, with a market cap of $1.4B.  HUT ($2.5B mkt cap) and HIVE ($1.8B mkt cap) are two other “majors” but they don’t have reliable Street estimates so I have excluded them from the comp sheet below.  Based on my estimates, which are higher than the Street’s, I have ARBK trading at a EV/EBITDA (CY22E) of 4.3X vs the peer average of 9.9X.  In terms of publicly stated growth plans, its effective hash rate capacity is expected to grow a bit slower than the peer group at 244% (from now to the end of 2022) vs 303%, revenue growth from CY21E to CY22E is similar to the peer group at 127% vs 124% and EBITDA margin (CY22E) is projected to be a bit higher at 77% vs 71%.  

While I am bullish on the price of Bitcoin and think the Stock to Flow models with price targets as high as $250K in the next few years are not outlandish, I’m using more conservative assumptions in the model below, assuming prices rise to $72K in 2022 and $84K in 2023.  I’ve extrapolated the network hash rate based on its relationship with bitcoin prices historically, with some catchup in 2022 (up 62% y/y) as the miners outside of China pick up the slack left by the ban in China.  This in turn drives the volume of Bitcoin equivalents mined.  The direct mining cost is built from the ground up and is largely based on the electricity cost of owned and leased facilities.  Similarly, capex is built based on estimated prices per unit of machines ordered.  I have assumed no further orders beyond what has been publicly announced, so 2023 estimates are likely conservative.  I have recurring EBITDA ramping from $79M in 2021 to $186M in 2022 and $309M in 2023.  Free cash flow breaks even in 2022 and ramps to $281M in 2023, with capex down to just data center builds.  Net cash builds from $80M at the end of 2021 to $606M at the end of 2023 and the value of the Bitcoin holdings grows from $118M to $855M (ie greater than the market cap).

Note that I have ignored BTC related valuation changes from an accounting perspective except in the calculation of taxes as that is a real obligation, my feeling is that investors will only give credit for recurring cash flows.

Risks

• Bitcoin is a volatile currency.  Argo’s revenues and earnings are sensitive to the value of BTC and the network hash rate, which are outside of its control.  Bitcoin has had wild swings, losing 80% of its value in corrections.  Argo does not hedge this exposure (nor would we want them to as investors), and its earnings will therefore be quite volatile.  Not also that in 2024, the next halving event, block rewards will be cut in half, a direct hit to revenue.  I believe that BTC prices will rise as with each of the previous halving events, to offset this revenue hit, but there’s not certainty that the next halving event will be similar to the last.  

• Regulatory risk. The future adoption of BTC as a currency or store of value is uncertain, and BTC may lose public favor if there is regulatory intervention.  Central banks and governments around the world have raised concerns around the security related to Bitcoin transactions, for instance.  The use of BTC to fund illegal activities is one concern, as it is hard to trace the identity of individuals behind crypto transactions.  There are also ESG-related concerns, but as mentioned previously, we believe Argo is a good actor here, and they have been a pioneer in reducing the carbon footprint of mining BTC.

• Proof of Work vs Proof of Stake.  The Bitcoin network uses a proof-of-work consensus mechanism, relying on mining infrastructure, to secure its distributed ledger.  If, like Ethereum, it were to shift to a proof-of-stake mechanism, there would be no need for Bitcoin rewards, and the mining business model would turn on its head.  We believe that as the total mining hash rate increases, it becomes harder to reverse a historical transaction (as it entails solving a new proof-of-work for each precedent block of transactions), and the security of the Bitcoin network is unrivaled.  It would increasingly require a prohibitively massive sum of capital to acquire the hardware necessary to corrupt the Bitcoin blockchain (or to achieve the voting power to shift the methology to proof-of-stake).  

Catalyst

• Completion of West Texas Expansion.  By the end of 2022, Argo’s power capacity will have increased from 45 MW to 245 MW.  The first 100 MW will come online in 1H22 with the second 200 MW in 2H22.  The company placed an order for 20,000 S19J Pro miners from Bitmain as part of this expansion.  The machines will be installed in the Texas facility in monthly batches from 2Q22 through 3Q22.  Argo is expected to end 2021 with a hash rate of 1.7 EH/s and the incremental Bitcoin machines will add 2 EH/s so that by the end of 3Q22 the total capacity will be 3.7 EH/s.

• Institutional Adoption of Bitcoin.   The market cap of Bitcoin, at about $1.1T, is about 10% of that of gold.  There is over $100T of assets under management by institutional managers worldwide.  A shift of just 1% to Bitcoin would nearly double its market cap.  Some investors may be waiting for a more well-defined regulatory framework to make the plunge, and on that end, I’m encouraged that SEC Chairman Gary Gensler is not only knowledgeable about Bitcoin, he has taught classes at MIT with a focus on digital currencies!  We have seen the approval and launch of the first BTC futures-based ETFs and may soon see the approval of a spot-based ETF.