Market View
The crypto market’s attention remains fixated on flows over fundamentals, as spot bitcoin ETFs in the US garnered their first week of net outflows (US$836M between March 18-21) in two months. There seems to be very little insight into what has driven the surge in outflows from Grayscale Bitcoin Trust (GBTC), which totaled a large $1.83B over four days – averaging $458M per day. To put that into context, daily GBTC outflows in the previous three weeks averaged only around $290M. Moreover, in previous weeks, we saw positive inflows into other funds more than offset the GBTC move, suggesting we were witnessing some capital rotation at that time.
One known source of potential selling pressure that we’ve been anticipating since mid-February has been Genesis Global Holdco LLC’s sale of 35.9M GBTC shares (worth $2.1B). Recall that Genesis received permission on February 14 from the US Bankruptcy Court for the Southern District of NY to sell its GBTC shares. Of those shares, around 31.2M were linked by Gemini Trust Company to the Gemini Earn Program. However, the court ruled that these shares were never properly pledged as collateral by Genesis to borrow from Earn users, so Genesis could sell these shares at market value to pay back creditors – in either cash or bitcoin. The Wall Street Journal reported on March 18 that there’s an outstanding creditor-backed proposal that could return up to 77% of customer holdings in-kind.
Note that this is separate from the 30.9M in GBTC shares that Genesis did pledge as collateral to borrow $1.2B from 232k Earn users back in 3Q22. Gemini recently settled with Genesis to return all of those assets in-kind with 97% to be paid within a few weeks, pending court approval.
It’s not clear whether the recent GBTC outflows are linked to these sales, as there are no direct public filings that immediately announce when they happen. For now, we can only infer that the size and scope of the change in GBTC shares outstanding coincide with recent developments on Genesis’ payment obligations. More importantly, given that the majority of creditor payments will be made in crypto and not cash, the market effect on bitcoin performance should eventually be net neutral, in our view.
So far, there doesn’t appear to be any unusual activity in global BTC spot volumes on centralized exchanges, albeit at around an average of $35B traded on March 19 and 20, this is above the daily (weekday) average of $25B over the prior four weeks. If these GBTC sales have indeed been completed, we think the macro environment remains amenable for more spot bitcoin ETF inflows following the Federal Reserve meeting that concluded on March 20. The board expectedly kept rates on hold, but also:
- opted for no change to its projection of three cuts in 2024 per its dot plot (despite lifting the dots for 2025 and 2026),
- raised its 2024 real GDP growth forecast from 1.4% to 2.1% YoY and
- alluded that tapering of its quantitative tightening program will start “fairly soon.”
In our view, whether the Fed decides to cut rates in May or June is less relevant for markets at this point than the direction of travel established by the board. We expect financial conditions in the US to continue easing, the disinflationary trend to remain intact and markets to be supported when the Fed reduces the amount of bonds rolling off its balance sheet each month.
Meanwhile, ETH’s performance diverged slightly from BTC this week after it surfaced that the Ethereum Foundation received an enquiry from an unidentified “state authority” sometime around February 26, 2024. No further details are available pertaining to the jurisdiction or content of the enquiry. On Twitter, a partner of law firm Willkie Farr & Gallagher said “it’s extremely common for crypto protocol foundations to receive voluntary requests for information.” Public prediction markets have reduced the probability of a US-based spot ETH ETF being approved by March 31 to 17% from 48% a month earlier. Nevertheless, we think ETH seems to be holding up well despite the odds of approval being effectively priced out.
Onchain: Scaling Horizons
Onchain activity has surged across a number of networks following the memecoin frenzy and Ethereum’s Dencun upgrade. The activity pickup was most notable on Solana and certain Ethereum layer-2s (L2s) like Base where the Dencun upgrade markedly reduced fees. In general, the median transaction fees across these networks during low congestion periods has been below $0.01 (see Chart 2).
However, high traffic can impact network performance in different ways. Transaction fees on Base surged to pre-Dencun levels during congestion periods, which left some underpriced transactions in a pending state until the market fees returned to their target rate for execution (there is no transaction expiry unlike Solana). In general, Ethereum and many of its L2s can guarantee transaction landing when outbidding other transactions for blockspace, which results in higher median transaction costs during congestion. On the other hand, Solana has been able to retain a low median fee due to their absence of compute unit based pricing (i.e. fees are not tied directly to computational resource consumption) and lack of effective priority fees. This approach incentivizes spamming the network to increase the probability of landing a transaction, which has resulted in a larger number of both dropped and failed transactions on Solana recently.
Transactions can be dropped on Solana (i.e. never included in a block) during high volume periods if the transaction doesn’t get included in a block before its blockhash expires. Separately, the relatively high number of failed transactions are generally due to onchain smart contract reversions (e.g. poorly set slippage tolerance). Longer confirmation times from an overloaded network greatly increase the likelihood of prices moving outside of predetermined ranges during swap transactions, particularly for volatile and low liquidity assets like memecoins. Precise metrics for dropped transaction counts and realized confirmation times are not readily available since these occur offchain and cannot be rediscovered from the ledger history.
There is still a long scalability roadmap ahead for both blockchains with differing tradeoffs in the routes they are taking. The Base team is considering directly increasing their chain's target capacity in the near term as data fees remain low (Ethereum’s blob count per block still remains below target). The next large bottleneck centers around optimizing execution, particularly in the area of handling state growth. Solving long term state growth problems could take some time, however, leading to continued price surges during congestion periods in the interim.
Solana Labs’s v1.18 client release scheduled for mid-April could also address some of their existing problems by upgrading their scheduler mechanics, though this scheduler upgrade is opt-in only. The Solana Foundation has separately been pushing for more optimal architecture including implementing priority fees (integrating them into dapps), optimizing compute unit usage, and other Sybil resistance mechanisms which can improve overall network performance. That said, we think that Solana’s uptime during these periods of heavy loads is a testament to the progress the network has made over the past several years.
