Token emissions are closely watched figures, which are often directly computed – and directly compared – as inflation percentages. However, we think this approach fails to account for the important nuances in underlying network structures. For example, the impact of emissions on proof-of-work (PoW) systems tends to be greater than that on proof-of-stake (PoS) systems due to operational costs. Furthermore, the emission formulas within PoS systems can vary greatly, and should not be analyzed in isolation from changes in the staking ratio in our view.
We analyze the emission schedules for three leading cryptocurrencies (bitcoin, ether, and solana) and contrast how incentives have driven differences in the ability of miners/validators to retain their earnings. In particular, we find that Bitcoin’s stalling hash rate growth may indicate that miners may be selling a growing portion of their revenue to cover costs. In comparison, Ethereum’s growth in stake has outpaced that of its net emissions, indicating that staking acts as a net liquidity sink. Meanwhile, Solana appears to have reached a steady staking ratio, suggesting that some validator earnings may be slowly entering the market.
That said, emission schedules are not set in stone and can be changed both by hard forks or onchain votes. For example, Ethereum’s burn mechanism and shift to PoS were only done in 2021 and 2022 respectively. Likewise, Solana’s SIMD-0096 is set to eliminate Solana transaction fee burns from priority fees. Altogether, we think that a simple comparison of headline inflation numbers is insufficient to understand emission based flows, as networks are unique in their incentive structure and maturity.
Bitcoin: Fixed Schedule via Proof-of-Work
Bitcoin’s fixed supply cap and emission schedule is iconic for its simplicity. For each new block generated (on average every 10 minutes), the miner earns a newly minted block reward, which is additive to the token supply. Every 210,000 blocks (approximately four years) this fixed block reward reduces by half in an event known as a "Halving". At its genesis in 2009, Bitcoin’s block reward was ₿50.
Today, that block reward is ₿3.125, and it will continue to taper off until it reaches the smallest BTC denomination possible – one satoshi (or “sat”) equal to ₿0.00000001 – after which the block reward will reduce to zero. The zero block rewards and fixed supply cap will not be reached until 2140, though the inflation rate will continue to decrease with each halving as the block reward is cut in half. Until this supply cap is reached, however, bitcoin remains an inflationary asset. At the current block reward, ₿164K ($10.3B) will be minted each year.
Bitcoin miners additionally earn the sum of any transaction fees included in a block. These transaction fees are recycled from previously minted block rewards, however, and do not contribute to increases in bitcoin supply. That said, we think that transaction fees are an important component of overall bitcoin flows as they also contribute to miner revenues that may be sold to cover operating costs. Furthermore, fees may become a larger share of miner earnings as block rewards shrink.
These miner revenues are a source of consistent bitcoin sell pressure. The cost-intensive nature of bitcoin mining means that a portion of mined BTC needs to be sold in order to cover the upfront financing costs of mining facilities as well as ongoing operational costs (e.g. electricity, taxes, and staffing). This “work” cost associated with PoW systems is also hyper competitive, constantly pressuring miner profit margins. Bitcoin’s total mining capacity (i.e. hash rate) has historically continued to increase even while block rewards decreased, which has reduced the profitability per unit of hash power. See Chart 2.
That said, the network hash rate growth has stalled following the Halving in April 2024. This suggests to us that profitability margins for miners have nearly maximally compressed – it appears to no longer be profitable to expand mining capacity at current bitcoin prices and hardware costs. (This is across the network as a whole, though individual participants may be adding hash power as others are exiting). Indeed, certain bitcoin miners have been diversifying out of the crypto mining space into other energy intensive sectors like artificial intelligence. The compression in profitability implies that a larger portion of mined BTC needs to be sold to fund operational costs, which are largely defined in fiat currency terms. (We covered more specifics on Bitcoin miner economics in a previous report.)
Onchain data reflects these costs. Chart 3 depicts miner holdings in light gray, while the cumulative BTC emitted via block rewards (since January 2020) is depicted in blue. Miner holdings have been on a downtrend since mid-2022, and have overall shrunk since the peak of bitcoin’s 2014 cycle. In fact, it appears that the ₿1.5M mined since 2020 is no longer held by miners. (We think it’s likely they’ve been sold.) The reduction in miner BTC holdings in the past two years suggests that the pace of total miner selling may even have outpaced that of BTC mining. That is, miners may be additionally offloading previously acquired BTC.
Not all mined BTC is immediately dumped into the market, however. Miners themselves may speculate on bitcoin performance and stagger selling (and hedging) in line with their views. In fact, some miners have even further financed the purchase of additional bitcoin on top of their ongoing earnings. This is true for only a small subset of miners with access to external capital, however.
Since the April 2024 Halving, miner revenues (i.e. potential sell pressure) has amounted to an average of $218M per week, though it was as high as $489M pre-Halving in April 2024. The potential outflow from miners YTD has been more than offset by the inflows into US spot bitcoin ETFs, with $17.7B net inflows since inception (as of publication). Denominated in BTC, US ETFs have seen inflows of 305K BTC, equivalent to the total new BTC minted between July 2023 and August 2024. However, ETF inflows appear to be slowly tapering off, and have only marginally exceeded net miner revenues since mid July. (US spot ETF holdings grew by ₿17K between July 20 and August 20, while miners earned ₿15K.)
This suggests to us that bitcoin’s inflation rate (which reflects total miner revenues) does represent an important avenue of selling pressure in the market. Furthermore, we think that Bitcoin’s hash rate saturation may have an important implication for any competitively mined PoW token – that the combination of net emissions and transaction fees (i.e. miner revenues) are likely to form a baseline of consistent sell pressure in the long term.
Ethereum: Dynamic Schedule via Proof-of-Stake
Although the headline ether inflation (or deflation) rate is often compared to that of bitcoin, we think that the direct impact of ether’s inflation on flows is extremely different than that of bitcoin. PoS systems generally tend to have very different economics than PoW systems due their smaller operational costs. Ethereum in particular, emphasizes the feasibility of solo staking, which enables especially low upfront hardware costs as well as low continued electrical costs even among PoS systems. Indeed, the estimated total energy expenditure for PoS Ethereum dropped 99.98% following its migration from PoW to PoS in 2022. This largely eliminates the sell pressure arising from funding operational costs, and allows stakers to take more long term speculative views on the asset by retaining more block rewards.
Ethereum has dual supply change mechanisms that operate independently of each other. (This is in contrast to bitcoin, which has a single inflationary component on supply until the supply cap is reached.) First, Ethereum emits gross issuance based on the total value staked. More specifically, the total gross issuance is proportional to the square root of the number of validators following the formula: max issuance per year ≈ 941*sqrt(number of validators). (Note: the leading coefficient is the product of a number of fixed parameters like the BASE_REWARD_FACTOR and annual epoch counts. Changing these would require a hard fork.) At today’s validator count of 1.06M, approximately Ξ18.6K gross issuance occurs per week.
At the same time, a portion of all ETH spent on transaction fees is removed from the circulating supply. This was introduced via EIP-1559 in the London hard fork in August 2021. Each Ethereum transaction consists of a base fee for general block inclusion as well as an optional priority fee to “speed up” transactions. The base fee is burned in its entirety, while the priority fee is awarded to the stakers. This component of ETH transactions is also known as the “burn”, and – in some high transaction fee environments like 2Q23 and 1Q24 – has even temporarily outpaced that of gross issuance. During these periods, ETH enjoyed an “ultrasound money” narrative of being a net deflationary asset.
The wide variability of ETH burn causes constant changes in the headline inflation rate for ETH. Since the Merge in 2021, ETH has been overall net deflationary, since the burn rate has exceeded the issuance rate. In fact, the total supply of ETH decreased by Ξ220K between September 2022 and August 2024. However, Ethereum faced multiple inflationary periods in the interim, and more recent inflation appears to be “stickier” as the network scales. (We discuss this in ETH and the RIse of L2s.) As we go to publish, the annualized net inflation rate for Ethereum is at approximately 0.7%, or Ξ17K ($46M) per week.
That said, we don’t think Ethereum’s inflation translates into sell pressure in the same way that Bitcoin miner revenues do. Prior to “The Merge” in 2022, PoW Ethereum had a similar emission structure to Bitcoin where miners were awarded a fixed number of ETH per block [1]. This means that Ethereum had no supply staked prior to the Beacon chain launch. Since then, the total amount of staked ether has grown to Ξ33M, 28% of the total ETH supply. This staking ratio still is far lower than other PoS chains (e.g. less than half of Solana’s 68% which we discuss below), and continues to grow. In our view, this is a critical distinction, since staking has been a large net supply sink for ETH that has far exceeded any issuance or burn over its existence. Chart 7 highlights this, showing that cumulative gross issuance has only accounted for 7.7% of the staking growth since January 2023.
This strongly suggests to us that the headline inflation number for Ethereum should be interpreted far differently than that of Bitcoin. In fact, we think that the headline inflation rate for Ethereum is a red herring for measuring sustained ETH sell pressure – at least until the growth in staked ETH falls below that of its inflation rate (which would be a heuristic for net selling by stakers in our view). Currently, ETH stakers appear to not only be retaining their staking rewards, but also adding to their staked positions in aggregate.
Solana: Fixed Schedule via Proof-of-Stake
In contrast to Ethereum, Solana has been a PoS network since inception. Furthermore, it has a fixed gross inflation schedule that doesn’t vary based on validator count. That is, even if the total amount of staked SOL increases, its overall inflation rate will remain unchanged. Currently, Solana's inflation rate is 5.1%, and will continue to taper by 15% annually until it reaches a terminal rate of 1.5% in approximately 2031 [2].
Solana also contains a burn fee similar to Ethereum. Currently, 50% of both base and priority fees are burned on Solana. (Ethereum burns 100% of its base fee.) However, Solana’s network fees are much smaller than its current gross issuance rate, only offsetting 6% of issuance YTD with 1.1M SOL burned relative to 18.2M SOL of gross issuance. Thus, the bulk of Solana’s net inflation rate arises from its fixed emission schedule. This may change as the inflation rate tapers, though coming upgrades will impact the transition timeline.
Most notably, the successful passage of SIMD-0096 in May 2024 eliminated SOL priority fee burns (though this will only be implemented in a future upgrade, which has no set timeline for release). Currently, approximately 85% of total Solana’s fees are priority fees, so the upgrade would materially reduce Solana’s burn.
At present, Solana’s net emission corresponds to approximately 528K SOL ($84M) new supply each week. (Compared to $198M for BTC and $46M for ETH). We think that Solana’s higher, and more stable staking ratio indicates that a larger proportion of these staking rewards may be sold on the open market relative to ETH.
Solana’s staking ratio is currently 68% and has been relatively steady since September 2021 when it surpassed the 60% threshold – it hasn’t dropped below that since. Solana’s staking ratio slowly continued to grow since then, reaching a peak of 72% in October 2023. However, the staking ratio proceeded to drop by about 1% per month until March 2024 and has remained largely unchanged since then.
In our view, this suggests some long term holders were taking profit into the price rally between 4Q23 and 1Q24. (The risks of liquid staking tokens and the 2-3 day unlock time for native SOL staking means that staking is generally incompatible with short-term trader positioning.) Additionally, the recent stagnation of the staking ratio further suggests to us that staking profits may be partially sold into the market.
If no new SOL were to be staked or unstaked, the staking ratio should continue to increase since token emissions go strictly to the staked SOL supply. At the current 5.1% inflation rate and 68% staking ratio, there should be a baseline growth of ~0.2% to the staking ratio per month. (This number would decrease as inflation tapers.) Changes to the staking ratio below this threshold indicate a net flow of SOL being unstaked, representing approximately net 5M SOL unstaked between March and August 2024.
Not all of this SOL is necessarily being sold (similar to how not all bitcoin miner movements are immediate sells), though it does indicate that staking is no longer acting as a net liquidity sink for the asset in the same way it is for ETH. In our view, this places more importance on Solana’s headline inflation rate as a possible indicator for flows, albeit not to the extent of Bitcoin.
Conclusion
Ultimately, the means by which token inflation rates translate into flows varies widely. Different costs of issuance (i.e. mining vs staking) as well as issuance variability (i.e. burn rates) may have major impacts on how headline inflation rates translate into flows. In particular, competitively mined PoW chains like Bitcoin are likely to face more net sell pressure from miners to cover costs, while PoS chain stakers may have a better ability to retain a portion of their staked earnings. At the same time, PoS chain inflation rates should take staking ratio changes into consideration. Both metrics are required for a more complete picture on emission-driven flows, which we demonstrate with ETH and SOL – staking is currently an active liquidity sink only for the former.
