Summary Bitcoin transaction fees have failed to live up to expectations, potentially impacting miners' profitability and network utility long term. Marathon Digital recorded a net income of over $200 billion in Q4-23 due to changes in accounting rules, but I calculate Q4 breakeven BTC price at $43.7k. Balance sheet management has been impressive with strong liquidity and a large reduction in debt. From a P/S valuation standpoint, MARA is already pricing in a six figure BTC price. When I last covered Marathon Digital ( MARA ) for Seeking Alpha in December, the Bitcoin ( BTC-USD ) network was experiencing a dramatic surge in transaction fees as a percentage of total block reward. The potential long term stickiness of transaction fees would be hugely beneficial to BTC miners like Marathon because it would minimize the impact of emission halvings - the next of which is roughly two weeks away as of article submission. Miner reward distribution (IntoTheBlock) In the months since that article, transaction fees have failed to live up to the excitement from Q4 2023. While we could theoretically see transaction fees move back up in the future, I suspect the more BTC that flows into the spot ETFs the less likely that outcome may ultimately be. Thus, I'm providing an update. We'll look at the company's forward guidance from its recent report, Q1-24 production data, and assess whether or not MARA is a fundamental buy at a $5 billion valuation. Q4-23 Performance Data by YCharts In the last quarter for 2023, Marathon Digital recorded $156.8 million in quarterly revenue versus $185.3 million in total expenses. What likely jumps off the page is the monstrous positive quarterly net income figure of over $200 million. How could this be with expenses exceeding sales? CFO Salman Khan on the last conference call: Had the company not early adopted the new FASB fair-value accounting rules, our net income attributable to common stockholders for the fourth quarter of 2023 would have been a net loss of $5 million or loss of $0.02 per diluted share. And net income of $33 million or $0.17 per diluted share for the year ended December 31st, 2023. For the full year, MARA produced 12,853 BTC and grew its Bitcoin treasury stack by 2,942 BTC year over year. From a cost of production standpoint, there was a notable move higher in MARA's breakeven price for Q4 based on my own estimates. But trailing twelve months, the company's production costs were actually much better than most other players in the industry. Q1-23 Q2-23 Q3-23 Q4-23 TTM Cost of Revenues $33,400,000 $55,200,000 $59,600,000 $75,100,000 $223,300,000 Adjusted Opex $21,600,000 $42,800,000 $53,900,000 $110,300,000 $228,600,000 BTC Mined 2,195 2,926 3,490 4,242 12,853 Breakeven Price $25,057 $33,493 $32,521 $43,706 $35,159 Sources: Seeking Alpha, Marathon Digital, Author's calculations I do want to mention that these breakeven figures are not meant to be taken as gospel, they are simply an external estimate that can be equally applied to all of the companies in the space to gauge efficiency among peers over time. If we strip out those FASB accounting changes and build a real time breakeven Bitcoin price for MARA that combines cost of revenue, SG&A, and Depreciation/Amortization and then divide that figure by the amount of BTC the company produces each quarter, we get a $43.7k breakeven BTC price for Marathon Digital in Q4 and a $35.1k BTC price for the full year 2023. Current State and Forward Guidance Author's chart (Marathon Digital) As of end of March, Marathon Digital has 17,381 BTC on the balance sheet. At a coin price of $68k, MARA has nearly $1.2 billion in liquidity from BTC alone. The company is in a much better financial position than it was a year ago following the depths of crypto winter. Over the last several quarters, Marathon Digital has done a nice job growing assets while also lowering debt: Data by YCharts Much of this asset growth has come at the expense of shareholder dilution, but with the price of MARA shares up over 400% since the end of 2022, anyone who bought the capitulation in December of that year has substantially outpaced the dilution since that time: Data by YCharts On the conference call, CEO Fred Thiel mentioned just how different the setup is for miners today compared to a year ago when BTC was trading under $25k per coin and the company had more debt than liquidity: We have $1 billion of liquidity on our balance sheet and have reduced debt by over $411 million, while saving our shareholders $100 billion in the process, resulting in net debt of $331 million. What Marathon has pulled off regarding the balance sheet is admittedly impressive. Thiel went on to mention the company plans to take control of about half the of their total hash rate and we're already seeing some of those dominos falling with Marathon recently agreeing to buy Applied Digital's ( APLD ) data center in Garden City, Texas for $87 million. MARA has historically used APLD for Bitcoin mining hosting services. According to Marathon, this deal will reduce mining cost at that site by 20%. Investor Deck, Slide 3 (Marathon Growth) Looking ahead, Marathon Digital is eying substantial exahash growth through the remainder of this year and into 2025. 35 EH/s by the end of this year would be 42% growth from the end of 2023. 50 EH/s by 2025 would be roughly double the year end level. Thiel mentioned that it will be possible for this timeline to be accelerated if opportunities to do so manifest; which is likely a reference to post-halving consolidation that figures to happen when weak mining outfits tap out and turn off machines. Q1 Production Author's chart (Marathon Digital) Despite the phenomenal production metrics for most of 2023, Q1 was not particularly great for MARA both from a total BTC mined standpoint and from an efficiency perspective. After mining at least 1,000 BTC in 7 consecutive months between July and January, Marathon has averaged just 864 BTC over February and March. While Q1 production was up year over year, growth has come at the long term expense of BTC per EH/s efficiency: Author's chart (Marathon Digital) Marathon's BTC per EH/s numbers have been broadly trending down for over a year and it speaks to the fundamental problem with everyone in the business chasing share gains from what is a shrinking nominal BTC pie: Data by YCharts To call Bitcoin network global hash parabolic would probably be appropriate. The hope for companies like MARA is that they can grow share of global hash rate against their peers. To this point, the company has been able to accomplish that several quarters running. But even assuming Marathon can hit its scaling objectives, the block reward halving later this month is a major long term headwind. Already Priced For Perfection? If we take MARA's 2024 year end guidance for EH/s and calculate what we could reasonably expect annualized revenue to be in a post-halving environment, we get $1.1 billion in annual revenue from mining at a BTC price of $100k: BTC Price 30 Eh/s 35 Eh/s $60,000 $587,520,000 $685,440,000 $70,000 $685,440,000 $799,680,000 $80,000 $783,360,000 $913,920,000 $90,000 $881,280,000 $1,028,160,000 $100,000 $979,200,000 $1,142,400,000 $150,000 $1,468,800,000 $1,713,600,000 $200,000 $1,958,400,000 $2,284,800,000 $250,000 $2,448,000,000 $2,856,000,000 Source: Author's calculations This calculation makes some assumptions on post-halving efficiency and production. Specifically, I'm using a trailing twelve month weighted average monthly EH/s of 54.4 and cutting it down to 27.2 for my post-halving BTC per EH/s estimate. This obviously doesn't take a change in global hash rate or Marathon's share of global hash into consideration. But at 30 EH/s, MARA would be mining 816 BTC post-halving. At 35 EH/s they're at 952. Like the breakeven price earlier in the article, take this more as a back of the envelope estimate rather than as a forecast. In any case, this methodology has MARA shares already trading at about 3x forward sales at a $150k BTC price - or more than double today's Bitcoin price. The forward P/S sector median for info tech is 2.9. This estimate also assumes Marathon Digital sells production after the halving and history is an indication that the company generally doesn't do that. Other Considerations I've been writing about Bitcoin miners for Seeking Alpha for nearly two years. Over that time, I've been pretty clear that I don't personally view BTC mining companies as great long term investments. I think they're really great for traders and for that reason I still do hold some MARA shares. But if you're buying MARA for the BTC stack, you're likely going to continue getting diluted over time: Author's chart (Seeking Alpha, Marathon Digital) In the chart above, I'm showing nearly a three year trend in quarterly shares outstanding and BTC per MARA share. Over the last five quarters, the BTC per share for MARA shareholders has been nearly cut in half. This isn't meant to say MARA is necessarily a bad BTC proxy bet as much as it's just to point out each investor should be mindful of what their personal objective is. Investor Takeaway If you're trying to optimize swing trades and own companies that have BTC-denominated revenue streams, MARA works very well for that. You just need to be prepared for the volatility. However, if you're simply trying to increase your exposure to BTC specifically, the newly approved spot ETFs are probably a much better long term option. I'm personally still holding MARA shares because I think some of the more long term goals are actually fairly interesting; specifically wasted heat and stranded gas opportunities. But those are highly speculative at this juncture. On the merits of what the underlying business is today, I think spot ETFs are the better way to express a $150k BTC thesis than MARA at $19. If transaction fees have a comeback, that might change things for me. But the more BTC that gets sucked up by ETF managers, the less active addresses I'd expect to see paying fees on chain.
