Short Crypto Equities vs Long Spot Thesis
The following thesis was generated and written in a joint effort between Hal Press of North Rock Digital and Quinn Thompson of Lekker Research. You can find the original publishing on Hal’s Medium, linked here.
Disclaimer, NFA, DYOR
As we approach the ETF launch, we have seen an interesting dynamic in crypto equities. The hype among crypto market participants has increased alongside the well telegraphed ETF approval and equity investors have sought crypto exposure, specifically Bitcoin exposure. However, in the absence of a spot ETF it is challenging to obtain this exposure as most investors either do not like futures ETF’s or are not permitted by their mandate to own crypto directly. This has driven demand for proxy assets and resulted in large inflows and extreme upside re-rating across the crypto equity landscape. We believe that once the ETF is approved and available to trade, much of these flows will suffer a one-time reversal as investors rotate away from these less-than-ideal instruments and into the spot Bitcoin exposure they initially desired.
As a result, multiple attractive pair trades now exist, using long spot BTC exposure vs shorting the inflated proxy equities to effectively front run the expected rotation. Our three favorite expressions of this equity trade are BTC holding companies (MSTR), Bitcoin miners (MARA), and exchanges (COIN). Each has its own idiosyncrasies described below, but the high-level thesis for each is the same — investors have been using these equities as Bitcoin proxies and once a true Bitcoin instrument becomes available, we believe the pricing dislocations will correct to the downside.
Due to the strong demand for leverage in the market, there has been an opportunity to generate annualized 10–30% effectively risk-free yields by longing Bitcoin and shorting the futures contract on the CME — the classic basis trade. As mentioned above, given the lack of direct access to Bitcoin, there are reasons to believe some traditional investors have used these proxy assets as the long leg of the trade given the lack of direct BTC access. The initial signs of this may have been seen during the material sell off in these proxy assets on December 28th that started during the fixing window on the CME roll date. If this were to be the case it presents an additional headwind for two reasons. Not only do we expect the CME basis premium to permanently dampen now that a pure form of Bitcoin exposure exists via the ETF, that compression would require a closing of the existing trade that has taken place. That closing would include a purchase of the short future and a sale of the crypto equity used to express long exposure.
Microstrategy (MSTR)
MSTR is commonly used as a Bitcoin proxy and most of the street indicates they trade at a somewhat reasonable 5–10% premium to the underlying book value of the business (operating company + BTC holdings). However, if you make the proper adjustments to their share count it becomes evident that the discount is much more substantial at ~25%, after recently reaching a high of 50–60%. The calculations assume a ~15x EBITDA multiple for their software business, which is generous given the business hasn’t grown in multiple years. The two reasons given by bulls for why this premium should sustain are leverage and lack of a management fee. However, neither of these reasons are valid. Given the December 2025 convert is in the money and should be treated like equity, leverage comprises less than 20% of enterprise value and even including the in-the-money convert as debt the figure is only 27%, with a natural deleveraging effect as BTC trades higher. After considering this premium on the underlying operating company and BTC holdings, there is actually negative leverage in that $1 of MSTR represents less than $1 of BTC. The management fee argument also lacks weight because MSTR issues on average ~150,000–200,000 shares of SBC annually. This effectively serves as a ~130 bp management fee on shareholders, which is significantly greater than the 0–25 bps management fees being charged by spot ETF competitors. In this context, MSTR is extremely vulnerable and there is no reason in our view that it should not trade at a discount to their BTC when you consider the SBC, uncertainty of ever unlocking the BTC and the significantly more attractive spot ETF alternative.
Not to mention that exposure to BTC through MSTR comes with less governance rights than an ETF (fully controlled by CEO Michael Saylor through a dual-class share struct) and a costlier way to access leverage given their elevated cost of capital. On top of this Saylor has recently begun selling stock himself and will be selling over the next several months, adding further downside pressure to the stock. A similar dynamic happened with GBTC where it traded at a significant premium for years while it was being treated as a proxy ETF, only to reverse and trade at a discount when the asset fell out of favor due to better alternatives and worsening sentiment. Visible in the chart below, Grayscale stopped creating new GBTC shares in late January 2021 which kicked off the decline. The lack of an updated ATM shelf offering from MSTR despite a still elevated premium indicates to us that they are cognizant of this problem.
In the face of these realities, the expansion in MSTR premium to NAV is hard to justify. While historically in the 30–50% range, underlying fundamentals would indicate a discount of ~5–10% is likely more appropriate. We estimate ‘par’ to be ~0.0094 MSTR/BTC, with potential downside from there.
It’s fair to expect that Saylor will be reluctant to sell any BTC, but there could be a situation where a constant discount is seen as a detriment to the business. While he controls a majority of voting power, if the discount were to widen enough there could be an opportunity for a more hostile investor to begin accumulating a position and potential fiduciary duty responsibilities. This could be fended off by Saylor, who might personally buy discounted MSTR shares but could mean liquidation of his personal BTC to fund this. While unlikely, there is a non-zero chance that eventually MSTR could be forced to sell some of its BTC holdings, amounting to $8B in total. In a worst case scenario, this could have a reflexively negative impact on crypto markets if MSTR were to attempt to close the discount by selling BTC and buying back MSTR shares.
Bitcoin Miners
The thesis on Bitcoin miners is slightly more complex but similarly negative. Publicly traded Bitcoin miners have been treated as a Bitcoin proxy and no longer will once a spot ETF is available. On top of this, the halving also looms which will bring a 50% decline in revenue overnight, albeit aiding BTC itself by decreasing structural supply. This creates a powerful dual dynamic whereby both the ETF and the halving uniquely benefit BTC while hurting the miners.
The most basic explanation of the challenge the Bitcoin miners face is as follows. Below you will find a LTM chart of hashprice which measures a miner’s expected revenue per unit of hashing power. As a broad generalization, let’s assume the majority of public miners require a ~90+ PH/s hashprice to be ‘good’ businesses. The chart shows that these levels are barely achieved and more never sustained even during recent spikes in transaction activity. All else equal, this number will halve in April, reaching new all-time lows significantly below previous records set in late 2022 which induced carnage for the miners.
The two offsetting factors to this decline will be a reduction in hashrate as unprofitable miners shut in (near-term) and increased transaction activity (long-term). For context, previous halvings have witnessed ~20–30% reductions in hashrate upon the event only to be followed by a quick recovery within 1–3 months thereafter, indicating at best this shut in will provide a temporary ~25–40% boost to profitability. Given this will occur from a new low 40–45 PH/s base, at best this short-term relief equates to a 50–60 PH/s level, just barely bumping up against 2022’s previous lows. While it may be fair to conservatively assume this hashrate reduction could be longer lasting than past given this halving’s outsized negative impact, it will create a race to the bottom and heightened importance for miners’ new machines to be deployed. We wouldn’t count on this as all indications point to tens of thousands of new machines being delivered and brought online throughout the year at these public companies.
The second potential mitigant for miners are the increased fees and transaction activity generated by the rise of ordinals, inscriptions, NFTs and scaling solutions built atop the Bitcoin network. However this transaction activity is highly cyclical, often marking local tops in sentiment and we believe is currently maxed out on an interim basis. The charts below support this view as fees and total network transactions have both rolled over.
Average transaction fees have ranged from $10-$30 in recent months on the BTC L1 network which rivals peak 2021 levels and is already prohibitive, indicating further upside is unlikely to sustain.
The third and final piece of the equation for miners is a longer-term structural issue but important to keep in mind nonetheless. Given there is no way for a miner to differentiate itself on output and production (1 BTC = 1 BTC), the only competitive advantages are on the cost side. As global governments continue to soften monetary and fiscal policy with inflation remaining sticky, energy costs are likely at a local bottom. While there are some producers like RIOT and CIFR with contracted power agreements in place, the majority have variable price exposure, particularly MARA which already has the highest power costs in the industry.
This also feeds into broader interest rates and cost of capital for these businesses. The debt market is fully closed for these companies and most of them couldn’t afford to service leverage even if it was available. This leaves equity financing as the only path forward to extend runway which they inevitably will be forced to. Keep in mind the absolute worst thing a miner can do is shut in machines due to a lack of profitability because if they do, they don’t have a business and most are worth pennies on the dollar in a liquidation. But the second worst thing they can and will do is issue equity leading to extreme dilution of existing shareholders. This happens at every local top and has already begun this time around with Cleanspark’s recent offering. Similar to the July to September 2023 period, there we believe there will be many more of these to come.
There are potential broader market implications as well when it comes to Bitcoin miner balance sheets. The largest publicly traded miners collectively own nearly $2B in balance sheet BTC. In the expected event that the operating environment significantly worsens for these businesses and the public markets re-rate the premium on these proxy assets lower, there will be increased pressure on miners to sell off their BTC to fund operations. During historical periods of stressed profitability, miners almost always increase their BTC sales beginning first with their produced BTC and only stretching into their HODL stacks out of desperation. This balance sheet selling could come sooner this time around given the more dire threat to their businesses post-halving and after all, oil and gas producers don’t HODL the commodities they produce to lever up and speculate on future price appreciation.
Within the mining segment, we favor MARA as the best short. MARA has the highest breakeven operating costs of its peer group and stands to lose against competitors that can operate profitably with less appreciation in BTC prices. We believe MARA’s valuation north of $5 billion is too high in relative and absolute terms.
On a long-term basis the MARA/BTC chart should follow hashprice, not Bitcoin price. Post-halving, we estimate miners will require a $75,000+ BTC price to get back to current profitability levels. Considering this price increase required to restore margins, we believe owning the BTC commodity directly has superior economics and expect the MARA/BTC ratio to revisit or make new lows.
Coinbase (COIN)
The COIN thesis is also simple. The same ETF dynamic that applies to the other two ideas, also applies here. Additionally, there is a narrative that the ETF will benefit COIN as COIN will be used as the custody solution. Further examination reveals a more uncertain equation. While COIN will benefit from ~5–15 bps custody fees, they will potentially lose much higher margin business as some retail volume will inevitably migrate from Coinbase to the ETF. This retail business yields 100+ bps per trade. Therefore, COIN will effectively be replacing 100+ bps per trade business with ~5–15 bps annual business. Furthermore, the low fee ETF’s will put pressure on COIN’s fee structure as a whole. We are already seeing this potential erosion play out as the ETF issuers have begun a race to the bottom on ETF fees. Most significant though, in our view, is the fact that once BTC and ETH spot ETF’s are available, retail users will have less incentive to open Coinbase accounts which may harm user trends. Lastly, COIN now trades at quite a stretched valuation at ~35x forward 12-month EBITDA, which we think is likely to come in below estimates, and looks very vulnerable to any decline in crypto sentiment.