📣 Access KKR's latest Health Care Growth private equity fund!

Securitize Capital Logo

The Economics of the Ethereum Merge

by Joseph Edwards, Securitize Capital Investment Advisor
Sep 30, 2022

While the environmental narrative clearly dominated in the run-up to the Ethereum merge with respects to its benefits, within the crypto community itself, much more was made of the effects of the merge on the supply dynamic of Ethereum in the longer term.

The supply side of crypto tokens is something that has, historically, tended to be either significantly overstated or significantly understated in any given moment, with not much in between. In the past, this discussion typically centered on Bitcoin, on account of its dominant position in the ecosystem as well as its hard-capped maximum supply being core to its foundational mythos as “digital gold.” As a reminder, Bitcoin supply is set at a maximum of 21 million Bitcoin, with a Zeno's paradox-esque route towards that maximum. One half of all Bitcoin was mined in the first 210,000 blocks, followed by a quarter in the next 210,000, an eighth in the next 210,000, and so on.

The result is a regularly decreasing inflation rate - having been between 8.5% and 12% after that first “halving” in 2012, this dropped to between 3.5% and 4.5% after 2016, and again to around 1.75% presently. For comparison, the supply of physical gold increased 4% year-on-year as of Q1 2022, and has historically increased at somewhere between 1% and 5% annually. The present (and future) rates of Bitcoin inflation are hence seen as a very important part of securing the long-term future of Bitcoin as a store-of-wealth similar to gold and other precious metals.

Most early crypto assets followed Bitcoin in having a maximum supply and tried to find ways to push the envelope rather than break it. Litecoin in 2011 is a classic example, opting for an 84-million maximum supply, four times that of Bitcoin. However, when Dogecoin was launched in 2013, mining issuance was set at a total of 5 billion DOGE a year, but without any long-term cap on total supply. 

As cryptocurrencies grew from being simple currency-equivalents to having the potential to underpin smart contract networks and more complex digital economies, the perceived need for a maximum supply on certain classes of cryptocurrency faded somewhat. While layer-1s like Cardano, Avalanche, and BNB Chain still do have maximum supplies, a number of L1s do not, including Solana, Polkadot, and, most relevantly, Ethereum.

Ethereum started at 72 million total supply in July 2015; by July 2016, it had increased to 82 million, and by July 2017, to 93 million, for an inflation rate of 13% in both years. Block rewards decreased after the Byzantium hard fork in October 2017, and again with the Constantine hard fork in February 2019, while year-on-year growth remained in the region of 5 million ETH, equating to roughly the same inflation rate as Bitcoin between 2016 and 2020 at approximately 5%, but above it going forward. This, coupled with an uncapped supply, was seen as acceptable for the time being, but not a permanent solution (Exhibit 1).

Exhibit 1: Ether Supply Growth Chart

Source: Etherscan.io.

Why do we say that supply-side concerns are either overstated or understated? Bitcoin's aforementioned halving events have tended to attract great focus. Logically, a contraction in new supply means less selling (bearing in mind that even the most bullish of crypto miners have to cover running costs). Simply put, a lower supply coupled with consistent demand should, slowly but surely, drive price up compared to where it was before. Indeed, in all halvings, including the most recent one, there has generally been a significant surge in price in the months following. However, at a current maximum growth of roughly 7x (from $10,000 to just shy of $70,000) this halving has underperformed the estimates of models like the increasingly discredited “stock-to-flow.”

To be dismissive of there being a genuine causal relationship between the two seems unsound; it does appear to matter, and it is worth noting that we have seen smaller swings in the market in the recent past that can be traced back to significant short-term supply gluts (e.g. Q3 and Q4 2019 and PlusToken liquidations). However, a supply-side change will not necessarily represent an immediate turning point for the market. Bitcoin in 2020 was one of the best examples of this (Exhibit 2):

Exhibit 2: BTC Historical Price

Source: Securitize Capital September 23, 2022.

While price did increase in the days following the halving, no substantial break on the local high of $10,000 would be seen until late July, and even that break saw initial rejection before price action started to go truly parabolic from October onwards, five months following the halving. This may seem obvious; however, the notion that the halving would lead to immediate gains was arguably more of an institutional belief than a retail one, with open interest on CME futures (i.e. short-term instruments typically only available to larger investors) increasing dramatically in the run-up to the event.

Over the past couple of months, we have seen the same process play out for Ethereum. There are two parts to understand with respects to Ethereum's new supply situation.

The first is EIP-1559. Instituted in the London hard fork in August 2021, EIP-1559 changed how transaction fees on the network worked. On Bitcoin and most other chains, including Ethereum pre-EIP-1559, transaction fees were redistributed to miners, along with new issuance. In the longer run, transaction fees have almost always been dwarfed by new issuance in terms of providing a mining reward; however, said fees are vital to the long-term viability of any PoW chain with a maximum supply, given that new supply by nature will eventually diminish to practically, or actually, nothing over time. EIP-1559 instead “burns” transaction fees: the Ethereum spent is removed from the total supply permanently. The burning of supply has become a staple of the tokenomics of many chains in recent years. It is, fundamentally, a deflationary maneuver, and hence theoretically increases the value per token of the remaining supply automatically. 

The second is the transition to proof-of-stake (PoS). We will avoid going into too much detail on this, but the important part here is that one of the benefits of PoS was that the increase in new supply could be dramatically curtailed - validators were now stakeholders in the network, and therefore were more invested in the long-term token price due to their significant holdings (as opposed to miners, who did have their own sunk costs and therefore stake in the network's longer-term success, but were unlikely to be long-term holders and could easily move their operations to another network if things went badly).

These two parts have coalesced into the “ultrasound money” theory, which conceived as a retort by Ethereum community members to Bitcoin's own “sound money” tagline (relating to the digital gold concept). However, with both EIP-1559 and the PoS transition, the theory took on a more specific meaning - namely, the notion that Ethereum could become a fully deflationary asset. The logic is simple: if new supply is sufficiently decreased and burns are high enough (because the network is being used), then Ethereum becomes deflationary over time. Graphs such as the one below were everywhere, promoting a future vision for Ethereum that would see it massively decrease its supply over ludicrously long time frames.

Exhibit 3: Ether Supply Projection

Source: ultrasound.money.

The big question: does this conceptualization make sense? To understand this, we need to look at both parts of said equation, starting with supply. According to Ethereum researcher Anders Elowsson, there is a set equation for the annual increase in supply under PoS:

I=cF√D

where:
I is new issuance per year in ETH;

D is the amount of ETH currently staked;

F is the base reward multiplier, currently set at 64;

c is a constant of approximately 2.6 allowing for the epochs in a year and conversion from gwei to ETH.

While theoretically new supply remains uncapped, the diminishing gains on more and more staked ETH limits the amount of ETH supply. If all 122 million circulating ETH was staked, new supply would still only be approximately 1.8 million per year, or less than 40% of new supply under PoW as of 2021, for an inflation rate of 1.5% (similar to Bitcoin in its 2020-2024 epoch).

Of course, not all ETH is likely to be staked. However, most predictions on Ethereum staking miss the mark because they use the current staking figure of 13 million ETH, the ETH committed and locked on the beacon chain before the launch of PoS proper. 

Why is this an issue? It seems spectacularly unlikely that Ethereum, as now the biggest PoS asset and staking crypto asset in the world, will stay at what translates to a 12% staking ratio when compared to all other L1s with more than $1 billion staked (Exhibit 4).

Exhibit 4: Staking Ratio for a Sample of L1’s

Source: StakingRewards.com and Securitize Capital.

Due to the loose inverse correlation between real network economic activity and staking ratio, we should not expect Ethereum to ever reach the 70% mark of Solana, Cardano, or Internet Computer. There is simply too much ETH either in constant transit or locked away in riskier yield strategies across the DeFi space, and Ethereum staking is likely to be a higher-friction process than on those other chains.

However, even if we assume an aggressively low equilibrium for staking ratio - say, 45% (around the same as that of Tron's) - it would still represent 55 million ETH staked, equal to issuance of 1.2 million per year. While, at almost 1% exactly, this would see Ethereum keeping pace with Bitcoin's inflation through the next halving in 2028, it is more than double today's figure. With 13 million staked, new supply is just under 600,000 ETH a year, and it is this figure that most actively deflationary models are built on.

The second part is the burn rate. The most optimistic models for Ethereum's deflationary model tend to be built on a long-term average gas fee of some description; the ultrasound.money calculator for instance uses 74 gwei as its default. The exact providence of this specific 74 gwei figure is unclear, but most one-year or two-year averages would land somewhere in this area.

This, however, is somewhat misleading. Gwei is a microdenomination of ETH, equal to one-billionth of one ETH, and is used to denote gas fees because the variable nature of fees on different Ethereum transactions requires such denomination. The value of gwei is naturally tied to the price of ETH as a result. While fees have naturally generally increased in USD terms, in terms of gwei, they peaked in August 2020 and have generally been making lower highs since (Exhibit 5):

Exhibit 5: Ether Average Gas Price

Source: Etherscan.io.

The problem is assessing whether the dramatic recent drop-off in fees - average price was still near 60 gwei as recently as June - is an anomaly or a long-term trend.

We have probably seen close to peak gas prices in USD terms. One observation to note is the peak of Bitcoin fee prices in the 2017-2018 bull market compared to 2020-2021, even in terms of USD fees rather than native units (Exhibit 6):

Exhibit 6: Ether Gas Historical Fees

Source: Coinmetrics.io

Despite dramatically increased price per unit, the peak level of fees in 2020 and 2021 was lower than in the previous cycle. 

There is a complex and a simple way to state the conclusion from this. The complex way is that blockchains tend to undergo a process of economic function discovery and, after a point, fees markets start to reach a lower equilibrium as applications that cannot survive under those fee conditions cease to be built on those blockchains, while those that can are. The simple way is that there are not enough people in the world who are willing to spend $100 per transaction for fees to stay that high permanently.

If we assume that Ethereum price will go up, and gas fees will at worst stay the same, then of course gwei-denominated gas prices should trend lower. Furthermore, it is worth noting that even with Ethereum price being far lower than today, gwei price was consistently below 20 gwei for most of 2018 and 2019. Of course, this was with a much lower level of real economic activity on the network, and prices presently are lower than at any point since May 2020, which makes the comparison significantly more difficult. Still, any estimate of significantly above 30 gwei as a long-term equilibrium feels optimistic at best when average daily burn as of late has been in the region of 10 gwei. And while EIP-1559 was not specifically designed to reduce gas fees, it does appear to have contributed to doing so.

Defining exact burn rate is more difficult than supply, but the best estimates (per ultrasound.money) are that a 1 gwei increase in gas prices equates to an extra burn of 100 ETH per day with the throughput available on the network. Hence, 10 gwei would mean 1,000 ETH per day and 365,000 ETH annually; 30 gwei would mean 3,000 ETH per day and 1,095,000 ETH annually.

In terms of annual supply change, we can compare three scenarios for staking and burn. For staking, we can look at 13 million staked (current numbers), 55 million staked (Tron staking ratio), or 93 million staked (Solana staking ratio). For burn, we can look at 10 gwei (recent average), 30 gwei (general upper limit pre-2020), and 74 gwei (quoted figure). This gives us the following matrix (Exhibit 7):

Exhibit 7: ETH Supply Changes

Source: Securitize Capital.

Green indicates a deflationary state, while red reflects inflationary ones. As we can see, the low current level of staking means that even fairly reasonable projections for gas fees do still get Ethereum to an overall deflationary state; however, this quickly dissipates as staked amount increases. In the most pessimistic case (Solana-like staking ratio and 10 gwei fees), inflation would be more than 1%. 

Of note is that even in the very worst case, the Ethereum merge represents a significant improvement in monetary policy for Ethereum. However, the narrative that it is actively deflationary is likely a step too far, only really functioning if we assume gas fees that would represent a massive increase from recent levels. The 74 gwei modelling would at current prices amount to fees of around $3.5 billion per year, and hence close to $15 billion at levels seen within the last 12 months, which clearly does not pass the test of common sense. The scenario of 45% staking and 30 gwei fees, which isn't entirely unrealistic, would imply an inflation rate of less than 0.1%.

Of course, beyond modelling, ultrasound.money provides real-world data of supply since the merge (Exhibit 8):

Exhibit 8: Ether Supply Since Merge

Source: Ultrasound.money.

An aggressive and clearly orchestrated effort to promote burns did see supply contract at points on the first day, but since then, we have seen supply growth at roughly where we would expect from the 12% staked/10 gwei fee model.

This still represents a significant achievement for Ethereum, and a major improvement in monetary policy. Furthermore, this modelling does not (and cannot) account for the potential of restaking, giving validators further reason not to dispose of inventory and to hence further constrict things on the supply side. By no means are we denying that the supply dynamic changes enacted by the merge (and EIP-1559) represent anything other than a significant improvement in Ethereum's monetary policy. 

The narrative around Ethereum over the past month or so in particular has been overstated with respect to the deflationary scenario. We tend to think that the recent price action with Ethereum surging into the merge (+103% off June's lows by September 11th, having peaked at +131% on August 14th) and then struggling out of it (-32% from September 11th to 21st), owes at least in part to misunderstanding in this regard - partially on the general effect of supply changes in cryptocurrency markets, but partially due to Ethereum's specific situation.

While short-term lows may still not be in, all of this explains somewhat why Ethereum has struggled with improved economics, and what to look for with regards to anticipating any shifts in trends going forward.

© 2022 Securitize Capital
All rights reserved
info.securitizecapital@securitize.io
The information contained herein is for informational purposes only. It is not a solicitation nor an offer to purchase or sell any securities. Securities will only be offered through confidential offering materials delivered to suitable, accredited investors. Private security transactions involve a high degree of risk and are not suitable for all investors. Securitize Capital is a subsidiary of digital asset securities leader Securitize, Inc. Securitize Capital’s investment offerings are available through Securitize Markets, LLC, an SEC-registered broker-dealer, member of FINRA and SIPC, and operator of the company’s alternative trading system. Securities offered on the Securitize Markets ATS have not been registered under the Securities Act of 1933 and may not be offered or sold in the United States absent registration or an applicable exemption from registration requirements. By accessing this site and any pages thereof, you agree to be bound by our Terms of Service and Privacy Policy.