Ethereum just went through its most complex upgrade ever. Justin Drake, researcher at the Ethereum Foundation, discusses what the upgrade, aka the London hard fork, means for ether’s future as what he calls “ultra-sound money.” At the end of the episode, he also drops a few Bitcoin hot takes. Show highlights:
- his background and how he got involved in crypto
- how EIP 1559 will affect Ethereum now that it is live
- why burning ether is good for the Ethereum economy
- what two concepts are crucial to understanding money
- what three ways ether is used as money
- why he believes ether will become “ultra sound money,” when gold and silver are just “sound money”
- what the main drivers of ETH’s value are
- how the merge from proof-of-work to proof-of-stake will happen for Ethereum
- how much more efficient Justin calculates proof-of-stake will be compared to proof-of-work
- why people are saying Ethereum is going through a “triple halvening”
- what factors influence how much Ethereum is issued and burned
- how Ethereum could become deflationary
- why net sell pressure is about to decrease
- what projects are burning the most ETH now that EIP 1559 is live
- why Justin is worried about Bitcoin’s future
- what Bitcoin could do to survive once it only subsists on transaction fees
- whether Justin considers himself a bitcoiner
- how ETH as ultra-sound money is becoming a meme
Thank you to our sponsors!
- June 23 Ethereum Foundation ETH 2.0 Research AMA (command F for Justin Drake responses)
- “If capped-supply gold is sound money decreasing-supply ether is ultra sound money”
- Modeling ETH based on EVM fees
- Calculating ETH supply
- Bankless “Modeling Ultra Sound Money” (April 28)
- Bankless “Ultra Sound Money” (Mar 23)
- Bankless “Bull Case for Cryptography” (Jan 25)
London/EIP 1559/Ethereum 2.0 Resources
- Ethereum’s EIP-1559 Will Solve Some Problems But Big Ones Will Remain
- Ethereum 2.0: What You Need to Know
- Why Ethereum 2.0 Could Fail and How It Could Be Fixed
- Previous resources
- EIP 1559 links (Tim Beiko)
- Messari Research
- Staked ETH address
- DETH of Ethereum
- Ethereum: The Triple Halving
- Coinmonks — talks about Justin Drake a ton!
- The Defiant
- Ultra Sound Money content
- Longer analysis
- ETH 2 Vision (Ethereum.org)
- The Block’s coverage
- Tim Roughgarden EIP 1559 economic analysis
Hi, everyone. Welcome to Unchained, your no-hype resource for all things crypto. I’m your host, Laura Shin, a journalist with over two decades of experience. I started covering crypto over six years ago and, as a senior editor at Forbes, was the first mainstream media reporter to cover cryptocurrency full-time. This is the August 10th, 2021 episode of Unchained.
The Crypto.com App lets you buy, earn and spend crypto, all in one place! Earn up to 8.5% interest on your Bitcoin and 14% interest on your stablecoins – paid weekly! Download the Crypto.com App and get $25 with the code “LAURA” – link is in the description.
Tezos is smart money that is redefining what it means to hold and exchange value in a digitally connected world. Discover how people are reimagining the world around you on Tezos!
Today’s guest is Justin Drake, researcher at the Ethereum Foundation. Welcome Justin.
Thanks for having me, Laura.
So we’re here to discuss ETH as ultra sound money, but before we dive into all that, why don’t we start with you giving your background and telling us what you do at the Ethereum foundation now? Because I think it informs our discussion about the idea of ETH as ultra sound money.
Sure. So I’ve been at Ethereum Foundation officially since December, 2017. So I got into crypto in late 2013. It was the bubble all the way up to a thousand dollars for Bitcoin. And I fell down the rabbit hole pretty quickly. I started the Cambridge Bitcoin meetup group in the UK, and then I started operating a Bitcoin ATM. And then I started a company that was trying to make an Open Bazaar, which is a peer to peer marketplace built on top of Bitcoin, easy to use. And then my startup didn’t do so well. Open bazaar also didn’t do so well. When I left my startup, basically I was looking for open problems in the space, and I was especially interested in Ethereum.
Vitalik gave this presentation on what’s called the data availability problems. So I spent some time thinking about this problem for, for a couple of weeks. And I had some ideas that seemingly no one else had in the space. So I emailed Vitalik with my ideas, and he liked my ideas, and we had this kind of long email thread. Kind of a few weeks later, he hired me and I joined Ethereum Foundation. So that was in December, 2017. So I’ve been at the Ethereum Foundation for a bit over three and a half years. And I guess my role has evolved as Ethereum 2 has evolved because I had been focused on Ethereum 2. So when I joined, there were very, very few people working on Ethereum 2. It was actually quite shocking how few people. I thought, you know, this was a multi-billion dollar project looking to make this huge upgrade — surely they must have like a massive team. It turns out it was like a handful of people.
And so I joined in then and basically working on pretty blue sky, kind of design questions, like how are we going to design sharding? I started focusing on sharding. And then as time went on, you know, we went from research to spec design. So the Ethereum Foundation, part of what we do, is producing the spec for the Ethereum protocol that then gets implemented by various teams around the world. So this is a very unique kind of model where there’s a decoupling from the protocol and the spec versus the implementation. I guess now that a lot of the research has been done, my role has shifted a little bit.
So for example, I’ve helped hiring. So we’ve hired actually some security researchers because one of the things we want to do soon is the merge. Basically securing Ethereum purely with proof of stake, no more proof of work. And so we want to make sure that the proof of stake, at a foundational layer, which is called the beacon chain, is as robust as possible. So we really want to have this very strong before the merge.
Another thing that we’re doing, and this is kind of a general theme, is that as blockchains become more and more sophisticated, they rely more and more on the cryptography. And so we’re building in-house within the EF, a relatively large team of cryptographic experts helping us with all sorts of primitives that it will be part of the Ethereum consensus, the layer one. And I guess my role in the last few months has changed yet again. Now I’m kind of at the mean layer, what we call the layer zero, under layer one like the people navigating this new space for me, which was very interesting through the ultrasound money.
Oh, interesting. Okay. So there are so many things that we’ll dive into based on what you talked about, but just to mention for people, some of the terms that you named there: the merges where Ethereum 1 apps move over to this beacon chain, which is the beginning of this new Ethereum 2.0 — the proof of stake network. And so we’ll talk about that more in a second, but first let’s talk about the most recent news. So we’re recording on Friday, August 6th, and on Thursday, Ethereum underwent a major hard fork, which implemented Ethereum Improvement Proposal 1559. And that essentially changes the monetary policy of Ethereum pretty significantly. So tell us what EIP 1559 does.
Right. So 1559 does a lot. One of the original motivations was around user experience — to improve the user experience. And one of the main things there is that, right now, when you want to use Ethereum, you don’t know how much you need to pay. So there’s basically two options. Option number one is you underpaid in terms of the gas price, and your transaction doesn’t get included on-chain, or it doesn’t get included on-chain when you wanted it — it kind of becomes pending for a period of time that you don’t want. Option number two is that you overpay and then you basically end up wasting some money putting your transactions on-chain. So one of the benefits of EIP 1559 is that users will know 99% of the time, what fee to pay, and if they pay that, they will just get included on-chain.
So it’s a much better user experience. EIP 1559 also improves security. And this is a bit of a subtle thing, but it turns out that transaction fees, so-called MEV, are the fuel for reorgs. So if you have a transaction fee, which goes in the block, another miner could go and take that transaction fee and put it in another, in his block, as opposed to that other miners block. And so one of the things that EIP 1559 does is that it reduces the value that can be extracted by miners or by validators, simply by burning some of the transaction fees that would otherwise potentially be used as fuel for reorgs.
This final aspect, which you mentioned, which is the monetary policy aspect, is that we don’t want to be overpaying for security. So when you look at the fuel for the security engine, if you will, so the consensus engine, which could be proof of work or proof of stake, it’s fueled by, one, issuance and, number two, by transaction fees. And what we’ve done in Ethereum is that we’ve set the issuance to be large enough such that alone they are sufficient to secure the blockchain. And so anything above the issuance is kind of overspending for security. And so the idea here is that instead of overspending for security, let’s destroy the ETH, thereby reducing the ETH supply and basically strengthening the monetary properties of ETH.
And so, in a way, the goals of EIP 1559 were to create more efficiency, both from a user standpoint, in terms of paying fees, and then also from a security standpoint. Why does that matter necessarily for the security of Ethereum, just to make it more efficient? What is the problem with maybe like overpaying for security, or, I don’t know, were there ever times when Ethereum was underpaying for it? Just kind of describe for me how security works prior to EIP 1559.
Right. So what we’re actually doing with EIP 1559 is actually reducing the security in a way. And the reason is that we’re reducing the, the security budget that would go to miners. So the amount of revenue that miners will be receiving will be reduced, and that will most likely reduce the hash rate. In terms of what we’re trying to achieve here is two things. One is guaranteed security. The way that we achieve guaranteed security is using issuance, right? Issuance is the best form of fuel for security, because it’s 100% predictable. It’s non-volatile, you know, you can just set it to be exactly the same amount for every single block, and it cannot be stolen. The issuance in a given block can’t be kind of stolen by the next miner in the same way that transaction fees could.
If there is a transaction, which is worth a million dollars, then that transaction could be included in block a, or it could be included in block B. That really affects the game theory.
So that’s the first thing we want to achieve. We want to achieve guaranteed security with issuance. And now that we have that, we want to achieve economic efficiency in the sense that we don’t want to be wastefully paying for security. If we don’t have to expand the security budget, let’s not do it. It’s just wasting money. So instead, what we do is we capture the value from the transactional utility of Ethereum in terms of transaction fees. And this has been done through the burning of transaction fees. And that, as I mentioned, is going to provide a deflationary pressure on the ETH supply — it is going to be reducing the total amount of ETH in circulation by improving the monetary qualities of ether.
All right. So this leads us to the main topic for today, which is this idea that ETH could be turning into what people in the Ethereum ecosystem are calling ultra sound money. So what were the characteristics of ETH as money before the upgrade, and what would you say they are now, and why would you call that ultra sound money?
One of the things that’s kind of useful is to understand what is money, right? So money, for me, is a money candidate with monetary premium. Okay. So what is the money candidate? And what is a monetary premium? A money candidate is an asset, which has various characteristics, which means that it could potentially be used at money. It’s durable. It’s divisible. It’s fungible. It’s transmissible, blah, blah, blah, blah. So if you look for example, at a cow, a cow that will never be money, right? Because it’s not easily divisible. You know, the head is not equivalent to the tail. It’s not durable, and it’s not transmissible, easily. On the other hand, if you take an asset like salt, for example, that was used as money, or gold or Bitcoin, they have all these basic properties, but it’s not sufficient to be a money candidate.
You can look at all the coins on CoinMarketCap. Most of them are not money. And the reason is that they haven’t really achieved monetary premium. Now, what is monetary premium? It’s this idea that the asset, the money candidate, has value beyond its raw utilitarian value. Maybe a good way to illustrate that as gold. If you look at gold, which has a market cap of $12 trillion, it has a kind of utilitarian value as an industrial metal. So, for example, gold is using every single iPhone, and people have estimated that the utilitarian value of gold is roughly a trillion dollars. And so there’s this extra $11 trillion, which is monetary premium, because central banks will keep gold in vaults and all sorts of reasons around money. And so monetary premium is really kind of this magic meme power, where somehow society agrees that this one asset is special in some way, and we’re going to allow it some special value.
Various assets have achieved this idea of having monetary premium, one of which has gold and one for example, is Bitcoin. Ether as well arguably has monetary premium. Now in terms of how ether is actually used as money today, it’s mostly used as money as a collateral money. So what does that mean? It means that ETH as a programmable asset can be placed as collateral, for example, in a smart contract and this collateral can be used to fuel DeFi. So right now there’s about 10 million ETH that is placed as collateral in DeFi. And so this ETH is used as money in the context of DeFi. There’s actually a second way that ETH is used as money also as collateral, but here it’s at layer one, not layer two, is with staking.
So there’s about 6.5 million ETH, which is used as collateral money in the context of staking. ETH is also used as money in the sense that it’s used to pay for transaction fees. Now this aspect of money, in a way is relatively weak because it’s doesn’t improve scarcity. When you use ETH today while, sorry, if you used ETH a week ago to pay for transaction fees, then this ETH would go to the miners, and basically you would just be circulating ETH in this kind of closed system. Whereas now whenever you are going to spend ETH, it’s very similar to gasoline, you’re just going to burn it and it’s going to be destroyed and you are kind of improving the scarcity of the thing. And so that’s kind of a a way in which EIP 1559 improves the monetary qualities.
Yeah, and it does so in a way where it reflects like the value of Ethereum contains within it the amount of usage it’s getting. So the more popular Ethereum is then essentially the more ETH that will be burned, which should then make easy asset more valuable. So in that regard, it kind of captures that. So then for this term ultra sound money, how would you say that that description now applies to your, or will eventually apply to ETH, I guess once all these changes are implemented? For now EIP 1559, and then later, eventually the merge.
So I guess before talking about ultra sound money, it’s maybe good to understand what is sound money. So the words sound money actually comes from silver and gold. And the reason is that there was this thing called the ping test, which was kind of a sound-based test to determine if a coin was a real silver or real gold. So what you do is that you put the coin on your finger like this, and then you take another coin and then you tap it and then go “ding.” There was this characteristic “ding” that, you know, silver coins and gold coins had. And you could use that to check if there are real.
Now this is kind of a metaphor for something kind of deeper, which is this idea that sound money is a money that cannot be debased. And here we’re using these metal based monies in comparison, for example, to fiat money. So in fiat money, you have a central entity who could arbitrarily debase the money. And what does debase mean? It means just arbitrarily inflating, the supply and increasing supply. With these metal based monies, you have this idea of a capped supply, and that’s very powerful because you’re protected against debasing.
I guess ultra sound money takes this no debasing idea to kind of the next level, where I guess the opposite of debasing might be rebasing. So it’s not only preventing kind of the debasing, but it’s allowing the rebasing of the money — it’s kind of strengthening over time. It’s this idea that not only do we have a supply cap, but we actually have a decreasing supply over time, because as you said, the transactional utility of Ethereum is being captured by this fee burn. And if the transactional utility is greater than the security budget that we’re expanding, then the supply decreases.
And just to give you an order of magnitude in the first 24 hours since EIP 1559 was launched, we’ve had 4,600 ETH burned, and that’s the equivalent of roughly $7 million. So that’s about $5,000 every single minute being burned. Now, historically if you really zoom out, the fee volume, which is the total amount of fees on the Ethereum network, since genesis, has grown the factor of 10 every single year for the course of six years. So that’s how how old Ethereum is — only six years old. And during the six years we’ve had this exponential growth.
Well, I think you were going to talk about how with the exponential growth that has created a lot of issuance and now finally, there’s an effect here where the supply will be decreasing. Is that where you were leading?
So let me try and finish that. So what we’ve observed is that over the six years of Ethereum, and we’ve had an exponential growth effect of 10X increase in the total fee volume for a Ethereum. Now, today, even today, the fee burn is greater than the proof of stake issuance and by a factor of four. So in the last 24 hours, the proof of stake insurance was 1,125 ETH. Whereas the the burn was over 4,500 ETH. What we say colloquially is that where we’ve reached ultra-mac 4, in the sense that not only have we broken the ultra sounds barrier, but we’ve broken it by a factor of four. And so once we don’t the proof of work issuance, which will happen at the merge, then almost certainly the supply will start decreasing. And as I mentioned, because the fee volume historically has just kept on increasing, and it’s a very nascent system, which is providing relatively little utility for the whole world, we can expect that the the total fee volume will dramatically increase as the utility that Ethereum is able to provide grows.
So before we get to the merge, at which point it could be that the ETH supply does become truly deflationary. Now we do have these three main drivers of ETH. And I know that the Ethereum community likes to describe them in terms that might be appealing to people who like physics because they’re described in terms of features solid, liquid and gas. So can you explain that analogy?
I like to think of money as water kind of flowing in the system. And water has different states of matter, depending on its temperature. And it turns out that the three scarcity engines in Ethereum that corresponds to different states of matter for water.
So on the cold side of the spectrum, on the minus zero, we have this idea that money, ETH, is frozen when it is staked. So when you stake ETH, you take your ETH, you place it as collateral, and you don’t get to touch it while you’re validating. And if you do want to exit the validating process, there is an exit queue so that not everyone can live the queue at the same time — it actually takes several months for this ice cube to melt, as it were.
On the other side of the temperature spectrum, we have the very hot, we have the above a hundred degrees, where basically water becomes gas. And this is the idea of the burn, the fee burn, and where we’re literally vaporizing ETH and we’re having it leave the supply.
And then in the middle, we have this third scarcity engine, which is DeFi. So what is DeFi? It is basically taking ETH as collateral, but it’s not as cold as staking in the sense that you can take it out, you know, more freely. And you can think of it as being liquid water, which is trapped in the pipes of DeFi.
And so these three mechanisms in their own different ways are basically sucking out the liquidity and like literally the liquid money that is kind of on exchanges and things like that —thereby effectively reducing the liquid supply of ETH and improving the scarcity properties of ETH.
For the liquid version when you talk about it being stuck in the pipes of DeFi, since it’s used as collateral, like in the case of MakerDAO, it’s put up as collateral to mint DAI, then that DAI is used for something, unless that person gets their position liquidated, then that ether there is also not exactly cold the way it would be if it were staked, but it’s kind of temporarily locked up and not not easily burned. So one last question for you. I think that each of these states has a fun sound effect attached to it. Do you want to show it off?
Let me get my things. So we have gas money, which is being brands, which sounds like this: air sound. And so every time you make a transaction on, if you need to hear in your mind a satisfying “pshhh.” And then we have the process of taking liquid ETH and solidifying it when you’re staking, which sounds like this “crack.” So should imagine kind of the water crystals forming as you’re freezing the money. And then you have the DeFi pipes where you have water splashing around. That sounds a bit like this: “glurp.”
Great, great. I love it. All right. So let’s now talk about that big milestone that we keep referring to, which is the next big step to ETH potentially becoming ultra sound money, and that is called the merge. So can you explain a little bit more about what the merge is and why that is going to be such an important step to ETH potentially becoming ultrasound money?
So the merge is a very important step for ETH because it fundamentally changes the way that it is secured. So right now, we have proof of work and proof of stake running in parallel, both at the same time, but all the economic value, all the DeFi and all the dapps and whatnot, are being secured by the proof of work. The proof of stake really right now is in its testing phase, right? We want to make sure that it’s robust and that there’s enough ETH being staked so that this proof of stake beacon chain is solid enough to receive all the economic value that’s being secured by Ethereum. Now at the merge, which is when we will determine that the beacon chain is secure enough, we will basically remove the proof of work and replace it by proof of stake.
Th the contracts don’t have to do anything. It will just keep on working. It’s just that underneath, behind the curtains, the security mechanism will change. So that’s one big change. It’s like from a security standpoint, we’re changing. We’re actually dramatically improving the security of Ethereum. There’s basically two reasons. One is around the concept of economic security. Like how much does it cost to actually attack the beacon chain? And you can look at basically how much ETH is being staked. So there’s roughly actually maybe $18 billion of ETH staked right now. And so if you want to perform kind of this 51% attack, you have to match as an attacker, the 18 billion ETH. So it’s a very expensive attack from a budget standpoint. But we have this other cool trick up our sleeve, which is that if an attack does happen, and it is possible that an attack will happen, then we can penalize the attacker economically with slashing.
So we can take a very large portion of this 18 billion ETH and just destroy it. And this is a process that can happen automatically with automatic smashing. So we basically have a system which is self-healing. So it is possible to attack it, but it’s self-healing. This is very different from Bitcoin, if for some reason or another on Bitcoin, a hacker has more than 50% of the hash rates, then they basically have God mode. They can do whatever they want. They can do arbitrarily long censorship. They can do these reorgs. They really have a God mode without the possibility of the system healing itself. So that’s the security aspect. But the other very exciting thing is the economic efficiency. We will be dramatically improving the economic efficiency. And the reason is that proof of work insurance is too damn high — to use a kind of this common meme.
Just to give you kind of a reason for that, is because the proof of work issuance has to pay for two things: it has to pay for the mining hardware, and it has to pay for the electricity running the mining hardware. These are two expenses that we don’t have in proof of stake. The the main expense for stakers is the cost of money, right? So they’re taking ETH and they’re locking it. And so the opportunity cost of not having this ETH kind of working for them somewhere else, that’s the real cost. And so that’s what we need to kind of account for and compensate in terms of issuance for proof of stake. And to give you kind of an order of magnitude, the cost of money is roughly 3%-5%, in the low single-digit percentages.
On the other hand, the cost of economic security on proof of work systems is more like a hundred percent annualized. So for every $1 of economic security, you’re going to have to spend in terms of issuance, $1 to cover both the hardware and electricity. And so you have this rough factor of 20X improvement in the issuance. For Ethereum specifically, the issuance is actually going to go down roughly by 10 X. And we have this meme, which we call the triple halvening, right? So in Bitcoin, every time you halven, the issuance reduces by a factor of two. But we’re going to be doing at least three issuance, three halvenings in one go because we’ll be reducing by more than 8X the issuance. And so once we remove the proof of work issuance, will live with just a proof of stake issuance. As I mentioned, the proof of stake issuance is smaller than the burn, at least in the first 24 hours, where the burn was four times larger than the proof of stake issuance.
And so if that trend continues, meaning that if the proof of stake issuance continues to be smaller than the burn, then the ETH supply will decrease. The rough projection is that the Ethereum supply will peak at the marriage. And that will be when the supply is roughly 120 million. So 120 million will kind of become the equivalent of Bitcoin’s 21 million. But of course, Bitcoin will inflate for another century until it reaches 21 million. Ethereum will be kind of a century earlier reaching its peak. And well, from that point on the words start decreasing.
Which this part is incredibly fascinating to me. So in a moment, we’re going to talk a little bit more about your projections and how all this will play out, but first, a quick word from the sponsors who make this show possible.
With over 10 Million users, Crypto.com is the easiest place to buy and sell over 90 cryptocurrencies. Download the Crypto.com App now and get $25 with the code: “LAURA.” If you’re a Hodler, Crypto.com Earn pays industry-leading interest rates on over 30 coins, including Bitcoin, at up to 8.5% interest and up to 14% interest on your stablecoins. When it’s time to spend your crypto, nothing beats the Crypto.com Visa Card, which pays you up to 8% back instantly and gives you a 100% rebate for your Netflix, Spotify, and Amazon Prime subscriptions. There are no annual or monthly fees to worry about! Download the Crypto.com App and get $25 when using the code “LAURA” – link is in the description.
Tezos lets you easily exchange smart money throughout our digital world. A self-upgradeable blockchain with a proven track record, Tezos seamlessly adopts tomorrow’s innovations without disruptions today. Because of this adaptability, engineers, conservationists, entrepreneurs, collectors, game developers, and artists from around the world are building, creating, and using Tezos everyday. Discover how people are reimagining the world around you on Tezos.
Back to my conversation with Justin Drake of the Ethereum Foundation. All right, so you came up with this pretty interesting spreadsheet that had different projections for how this issuance versus burn and all that would play out. As you mentioned, it’s kind of too early to really say how well your projections have done, because we just have 24 hours worth of data, but why don’t you just describe what those projections look like from the conservative to the optimistic viewpoint…
There are various variables that come into play. For example, if you want to estimate at what ETH supply we will have when we peak at the merge. By the way, there’s this really good tool on ultrasound.money, where you can just play with the assumptions as sliders, and there’s a pretty graph, which will just show you the projected supply. But basically the three key variables are one, how much ETH is being staked. So it turns out that the more ETH being staked, the greater the proof of stake issuance. In the worst possible cases, if all the ETH is being staked, we’re looking at less than a million ETH issuance per year.
Now the second variable is how much are we burning every single day? So in the first 24 hours, we burn about 4,600. So actually closer to 4,700 ETH. We will see how that progresses. My guess, as I mentioned, is that this amount of fee burn will increase, for example, as we provide more utility for scaling, right? Because scaling is an opportunity to provide much more utility to many, many more people. And so, even though the individual transaction fees will go down, the aggregate fee volume will be greater. At least that’s what we’ve observed historically, where the fee volume just keeps on increasing despite the scalability increases.
Just so I understand that, because when there’s congestion on the network, people tend to pay much higher fees. So what you’re saying is even though the individual fee per person will decrease because scaling will enable so many more transactions, that aspect of it will increase the fees for miners? Or the tips we should call them. Or it would be both base fee and tip, I guess?
Yes, correct. One of the things that we’ve observed historically as Ethereum has scaled and it has scaled. So for example, the gas limits when Ethereum was created at Genesis was only 3 million gas per block, and now it’s 15 million gas per block. So we’ve had this 5X increase in scalability. We’ve also had indirect scalability improvements, for example, as smart contracts get optimized for gas. So for example, Uniswap v3 compared to Uniswap v2, and various other DeFi contracts have tried to optimize for gas. And of course, we also have various roll-ups, which are on now live, providing even more scalability. And what we’ve observed historically is that as you increase utility, you also increase the transactional value that’s provided by this utility on an aggregate basis.
What might happen is that in the short term, as you increase the supply, you’ll have a supply shock in the sense that you have too much scalability and then the kind of the fees dip. But over the long-term, as adoption kind of picks up and fills this gap, this demand gap, then the end result is that the final fee volume is much larger than the initial fee volume. Okay. So that’s the fee burn per day.
And then you have the final variable, which is when we’re actually going to merge, right? Because the proof of work issuance is extremely high. It’s 13,500 if every single day. And so every day, so the sooner we merge, the better it is because the sooner we remove this huge fire hose of issuance that we’re drowning under in the context of proof of work.
So I’ve heard you talk about that as net sell pressure. Can you define that? And describe how that used to work in Ethereum? Or not how it used to work, but what the net sell pressure used to be and how it’s changed now? Or how it eventually will by the time of the merge.
When a miner receives income in the form of either issuance or fees, they will sell most of that income. And the reason is that they have to pay their expenses. And there’s basically three forms of expenses. One is hardware, two is electricity, and then there’s a third one, which is a bit more subtle, which is the income tax. So my rough estimate, because we’re in a competitive market, basically the profit margins tend to zero, but let’s say that product margins of 5% — very, very low for the miners. In effect, they’re going to be selling almost all of the ETH that they receive. Now, when you compare this to proof of stake not only is the income much, much lower because… one, the the issuance is roughly 10X smaller, but 2, a large portion of the fees are burned. And so when you compare and contrast the two systems, kind of from a week ago to post-merge, we’re actually in a situation where we’re going to be reducing the sell pressure of a in by 7 million ETH every single year.
So that’s roughly the equivalent of all the ETH being staked right now. Every single year that is not being sold. So you can kind of think of it as buy pressure relative to the past. In the recent past, we’ve just had this huge dampener on effectively the ETH price. And the reason is because there’s lots of ETH that just goes on the market to be sold, to go buy for electricity, to go pay the taxes, and to go buy the hardware. And this 7 million ETH is no longer going to be sold. 7 million ETH at current prices — that’s something like $20 million dollars, right? So that’s the equivalent of roughly $20 million of buy pressure in the future every single year relative to the past.
You started by mentioning income tax pressure. So can you talk about that?
With proof of stake, the main sell pressure, because there’s no expense in terms of hardware, the main sell pressure is actually income taxes. You receive ETH rewards, and you just have to pay some fraction of it to the government. In order to pay it, you have to go ahead and sell it. And you know, income tax in most jurisdictions is, let’s say around 50%. And so the profit margin for stakers is also roughly 50%, a hundred percent minus the 50% that goes towards taxes. So we have reduced sell pressure for kind of two reasons. One is because the profit margins for stakers are much higher. And 2, because the income that comes in in the first place is much, much smaller in this new system with EIP1559 and proof of stake relative to the proof of work system without 1559.
So, one other aspect I wanted to draw out for this period between now and the merge is I was looking at the burn leaderboard on ultrasound.money website. And it’s fascinating because it’s basically a snapshot of what’s hot in Ethereum right now. At least when I checked right before we recorded, it led with OpenSea, which was followed closely by COVID punks, which I actually hadn’t heard of until that moment. Then Uniswap V2, then Axie Infinity, then Tether, then Uniswap V3, MetaMask, and finally USDC. So I was just curious, what does that say to you about how those were the biggest contributors to the burn? How do you think this will play out in terms of what will be driving ETH to get burned in the near future?
At kind of a macro level, the fact that we have almost 5,000 ETH being burnt in 24 hours, kind of suggests that Ethereum is providing a huge amount of transactional utility. As you say, it is very interesting because now we have more insight into which contract specifically are providing you that utility. You highlighted a few for example, which are NFT and a few which are in DeFi. To roughly ballpark, it’s kind of 50/50 split.
You mentioned this COVIDPunk. Basically what happened here is that there was this really intense period of roughly half an hour or an hour where the base fee kind of spiked to something like 300-400 gwei. And there was a huge amount of burn-in that concentrated amount of time, because they were actually minting a new NFT set. And so people were kind of competing to grab this NFT because there was a limited supply.
And the OpenSea also suggest that people not only are kind of buying these NFTs, but they’re also trading them. And that transactional utility is reflected in the burn for OpenSea. And then there is this other category, which I guess is the DeFi category, with things like like Uniswap. And actually, we do have this other category, which is interesting, which is gaming.
And one of the things that I’ll say is that at the Ethereum Foundation our main focus, what we focus on, is the consensus layer, the layer one.
The layer two, we kind of want to want it to be as organic and as flamboyant and permissionless as possible. I’m just very excited to see — this ecosystem be so vibrant. As you said, there’s these things that I had never heard of before, like this COVIDPunk that kind of just came out of nowhere and just burned 500 ETH in a matter of half an hour, which is a kind of a crazy idea.
To me, when I looked at it, it just reminded me… because when I first started reporting on NFTs, I didn’t realize how kind of gas heavy they were. Cause I didn’t understand from a technical standpoint that it’s not like an ERC 20 token where you just have that one smart contract, like you have to mint each of these individual objects. And it just reminded me like, oh right, NFTs take a lot of gas. And then the Axie Infinity thing, I’ve done some reporting on that, and you can see this game is very popular right now.
I think one thing I want to draw out here is there were people expressing surprise at the fact that gas fees increased after the upgrade, which was counterintuitive since the change was being touted as something that would make them more efficient. So can you just explain why that happened?
Yeah, absolutely. So one of the things that I mentioned is that generally speaking, the free market is extremely volatile. And so it’s difficult to compare it kind of one day to another because there’s a lot of noise. Another thing that I’ll say is that EIP 1559 does not reduce the demand for making transactions and as such, it doesn’t dramatically improve the efficiency. So it does to a small extent, try and avoid overpaying — this idea of overpaying, but it’s not a huge thing. And so one of the things that we’ll be able to do actually in the coming weeks is kind of look at the data and see how much were users actually overpaying the miners to get into blocks.
One thing that I’ll note is this concept of induced demand. Like when you improve something, there’s just more demand for it. I already mentioned induced demand when you have scalability, when you improve scalability, just more people come in. But there’s actually more subtle ways in which you can have induced demand. So one thing that we mentioned already is that EIP 1559 improves UX. It’s just better to be transacting on the Ethereum. Because the user experience is better, people will just transact more — it’s induced demand.
One example, for example, even personally, like, let’s say, I want to make a transaction, I’m going to spend 10 minutes making my transaction or five minutes because I don’t want to overpay for gas. So I set a relatively low gas price, wait a few minutes, and it doesn’t go through. So I kind of give up and then I bump up my gas price and then it’s just a whole faff.
And then I end up taking a whole 10 minutes just to make one transaction. If instead the UX is — I want to make the transaction, and it just happens. Well now, 10 seconds later, I’m ready to make another transaction. So there’s also induced demand in that sense.
Another aspect in which that’s induced demand is what I call eco pain. So right now, or at least before EIP 1559, every time you made a transaction and you paid, for example, $10 in transaction fees, I knew that these $10 would really be bad for the ecosystem in the sense that $5, roughly half, would go to the landfills in the sense that you’re going to buy this mining hardware, which will depreciate over time. And then the other half is going to be used as electricity, literally like as an electric heater, like $5 worth of electric heating somewhere in a data center.
And so I didn’t feel good about making transactions on Ethereum, in that sense. But post EIP 1559, it’s good for two reasons. On the one hand, there isn’t this eco pain aspect. But there’s kind of this monetary pleasure, I guess, in the sense that you know that you’re improving the monetary properties of ETH when you do transact. I think the behavior change of ETH users will affect the demand. So there’ll be this induced demand.
I find that interesting. So one other thing that I wanted to circle back to was we did talk about how Bitcoin is sound money and how now ETH is ultra sound money. And one thing that I do know you have thoughts on is what Bitcoin might look like once it transitions to only transaction fees and there’s no block reward. Can you talk a little bit about the security of that? Because that was something I don’t feel like I had heard before and was really interesting to me.
So when you look at the security of the system, there’s kind of multiple things at play. Like one is, for example, what is the flavor of consensus? So you could have proof of work versus proof of stake. You could look at decentralization of the consensus participants. And then you can look at basically what is the fuel to the consensus engine? What are the incentives for these consensus participants to go do that work? Now, every single blockchain with no exception, as far as I can tell, is secured by issuance — meaning that there’s freshly minted tokens that are given to the consensus participants as incentives to go secure that blockchain, to go provide scarce resources to secure the blockchain.
Now Bitcoin hopes to be secured by transaction fees, which is different from today, where it’s secured by issuance. And so there’s kind of this grand experiment that Bitcoin is embarking itself on from a security standpoint, which is that, can it be secured by transaction fees?
And my guess is that the answer is no. The answer is no for several reasons. One is that you can look at transaction fees from a quantitative standpoint. So transaction fees originate from transactional utility. Now, Bitcoin is not optimized for transactional utility. The utility of Bitcoin is non transactional, it’s in the HODLing, right? So the way that you benefit from the utility of Bitcoin is you go buy a Bitcoin, you put it in a hard wallet, you forget about it for a decade, and then maybe in a decade, you go make another transaction to send it to an exchange and sell it. So you basically make two transactions over a period of a decade. And so the transactional utility is very small. And so the amount of transaction fees that the blockchain could hope to extract a relatively small as well.
There’s other aspects like the fact that there’s very, very little block space in the first place. And so because there’s very little block space, there’s also a very low opportunity to go extract these fees. This is reflected empirically today, when you look at the total fee volume. So if you go to cryptofees.com, I believe, you will be able to see that Ethereum has roughly 20 times the fee volume than Bitcoin and the fee volume for Bitcoin is kind of laughable — it’s half a million dollars versus $10 million. There’s no way that the money of the internet, with a hundred trillion-dollar market cap, can be secured by half a million dollars in fees.
But there’s actually another problem with transaction fees. So let’s just assume that magically transaction fees will be sufficient to secure Bitcoin. For example, let’s imagine that every transaction fee is a thousand dollars. People are willing to pay a thousand dollars per transaction. There’s this other problem, to which I alluded to, which is that transactions can be stolen from one miner to another. So if you have a transaction with a very large transaction fee, let’s say a $1,000 transaction fee or $10,000 or $100,000, it can get included in block a, or it can get included in block b. So let’s imagine that block a included a transaction. Now, block b, the miner for block b needs to ask himself, do I mine on top of block a or do I mind on top of the parent of block a?
Because if I mine on top of the parent of block a, then I can go take these very juicy transactions and include them in my block. So that’s the concept of basically mine extractable value, MEV, which is the fuel for reorgs. Now in a consensus system, you want this idea of convergence, right? You want all the miners to agree and converge very, very quickly on a chain. And this is achieved through issuance. In a way transaction fees are the opposite of convergence — it’s divergence, it encourages forking. And so it would be in a situation where Bitcoin is unstable.
Now, another thing that I mentioned is that transaction fees of volatile and also cyclical. For example, we know that on weekends, there is a dip in transaction fees, and then during the weekdays, a bump.
And so if there’s zero issuance, what will happen when Bitcoin is only secured by transaction fees, is that block times will closely match the incoming transaction fees because that’s the incentives that are coming in. And so on the weekends, you might have 20-30 minute block times. And then on weekdays, you might have five minute block times, because on the weekends, the least profitable miners will have to disconnect the mining hardware because there’s not enough transaction fees to to justify them running the hardware and paying for the electricity. And so very, very quickly you realize that Bitcoin will be a very unstable system and that won’t happen, a hundred years into the future, when the issuance is zero, it will actually happen 20 to 30 years into the future where the issuance will be close enough to zero.
And the reason is that issuance decreases exponentially every four years, it halves. And so even in a 20 to 30 year timescale, the issuance will be very, very small and most likely insufficient to secure the Bitcoin blockchain.
And so one of the things that both Bitcoiners and Ethereans agree upon is this idea of predictability, right? We want to have a system that can be relied upon for decades and centuries, and we’ve taken a very different approach to predictability. Bitcoin has taken a short approach to predictability. They decided to not change since since day one. And basically, the fact that you’re not changing means that you can predict what will happen in the short term — until, of course, the system is unsustainable and then something massive has to happen.
For example, they could remove the 21 million block limit. They could start increasing the issuance again and start reviving the blockchain from a security standpoint. Another scenario that I have in mind is that there is a decoupling of Bitcoin, the blockchain, and BTC, the asset, and BTC the asset, which is super scarce, like gold, can go live somewhere else on another blockchain that will host it for free. And that blockchain, for example, could be Ethereum. So that’s another scenario that I see.
Another scenario is for Bitcoin to start adopting some of the technical innovations that we’ve had on Ethereum, like for example, implementing EIP 1559, or implementing proof of stake. And so this approach to predictability for Bitcoin, which optimizes for short term predictability, is very different from Ethereum, where we actually are happy to trade off short-term predictability.
So if you look at Ethereum over the last six years, it has changed its monetary policy three times. Once it reduced its block reward from five ETH to three ETH, and then it reduced it from three ETH to two ETH. Now it’s reducing again, with EIP 1559, and then it’s going to reduce it again with the merge. So we know we have this hardening of the monetary policy overtime via these various innovations.
And we want to be at a point where we are very happy with the system — where we have guaranteed security and we have this idea of being economically optimal in some sense. Once we’ve reached this optimal point, there’ll be no more reason to change Ethereum because it will already be kind of either optimal or kind of close enough to optimal. So we’ll have a system which will have evolved in the short term and not had the short-term predictability. And then it reaches a point where it’s good enough, and then it can go live for decades and centuries.
This whole thing has been so fascinating. I just wanted to ask when you said that potentially Bitcoin the asset could just live on another blockchain that provided the security. And you said potentially that could be Ethereum, would that be in the form of wrapped ether? There was that project that tried to do trustless BTC before. What do you think that would look like?
Great question. So in order for this migration to happen, I think it needs to happen kind of gradually. The Bitcoin community is very opinionated. It’s difficult to reach consensus. They’ve designed the system and the culture in such a way that the Bitcoin blockchain kind of never changes. And so you kind of need need to find a sly side door and strategy to try and migrate the Bitcoin, the asset, onto Ethereum.
I think the plausible way forward that I have in mind is through these bridges. Now these bridges need to have two properties, which we we don’t have both for (yet). The first one is that the bridge needs to be trustless, right? Because of the bridge is not trustless, then we’re basically trusting this centralized entity to oversee the migration of one of the biggest assets in the world from one blockchain to another — it’s just not viable.
Right. Which means wrapped a wrapped BTC is out because that is managed by the centralized party. So, okay. Keep going..
And the other property that you want is you want it to be collateral efficient. There’s this project called TBTC which is trustless, but the big downside is that in order to create this bridge for every single Bitcoin, you need to have more ether and reserves and value than the Bitcoin. So it’s extremely collateral inefficient. The good news is that there is this moon math technology, which allows Bitcoin to run smart contracts or to verify Snarks. So it’s called indistinguishability obfuscation — it’s technology that we don’t have today. It’s very, very sophisticated, theoretical cryptography, but it’s plausible that in 20 to 30 years, we will have it and that will allow Bitcoin to essentially run the EVM. And so once we have Bitcoin running the EVM, even with the tiny blocks, that’s sufficient to have a two-way trustless bridge, which is fully collateral efficient. So you don’t need any collateral. And so we will be in the position where Bitcoin will slowly start trickling towards Ethereum.
And we’re actually already seeing it. If you go to defipulse.com/btc, you will see, over time, the unavoidable increase of BTC on Ethereum. I believe over 1.5% or something like that, maybe a bit less, of the whole Bitcoin supply, which currently lives on Ethereum.
There will be a breaking point for Bitcoin. A breaking point from a security standpoint. Either the blockchain gets 51% attacked because the issuance is way too low and the security is way too low. Or it could get attacked in another way. For example, it could get attacked in the form of a quantum attack. So quantum computers are coming. If there aren’t suitable preparations which are done to protect the assets on the Bitcoin blockchain, what will happen is that one attacker or maybe a small handful of attackers will be able to basically steal lost coins from the past. And that would be very bad. It would be kind of the equivalent of the DAO for Ethereum. One of the worries of the DAO is that a single entity would have a huge fraction of the supply and basically destroy the monetary value and also the security story around proof of stake, which requires the stake to be, to be decentralized. And so once there’s this event that happens and the Bitcoin blockchain breaks, all the Bitcoin that hasn’t yet migrated over to Ethereum can be deemed lost, and the new canonical home for BTC the assets can be a failure.
Wow, I like this is going to be fighting words amongst the Bitcoin community. This whole thing is very fascinating. Just out of curiosity, cause you started as a Bitcoiner, do you still consider yourself a Bitcoiner?
Absolutely. I hold Bitcoin. I feel very connected to Bitcoin from a philosophical standpoint, right. What Satoshi was able to do was a real breakthrough for humanity. We’ve seen over the period of 13 years, all the innovation that’s come out of it. The way that I see it is basically Ethereum is the accomplishment of Satoshi’s vision. It is Bitcoin. Ethereum is Bitcoin, in that sense.
What aspect of Satoshi’s vision is Ethereum? Like when you say the Ethereum is Satoshi’s vision, what do you mean?
Part of the vision of Satoshi was to provide a decentralized trust layer for the internet. One of the things that Satoshi tried to do in the very early days, and you have commits in the the repo is he tried to build a peer to peer marketplace that built on top of this money layer. The natural progression of decentralized infrastructure is to start with money, and then to start building building blocks that use this money. And you could have, for example, marketplaces, and you can have escrow contracts, and you can have stablecoins, and you can have reputation, and you can have identity, like Ethereum name system, and you can have insurance. Once you have all these building blocks come together, you basically have the substrate for the internet of value. And I think this is the true Satoshi vision, and this is what Ethereum is delivering.
Wow, this whole thing is extremely, extremely fascinating talking to you. All right. Well, one thing I will say is I have said in the past that I wished I could be around when Bitcoin transitioned to only transaction fees to see how that plays out, but if your theory is correct, then it looks like I might get to see what happens in 20 or 30 years, because I will still be around then.
All right, so we’re over time, but I really have to ask you about something super important, which is this idea of ETH as ultra sound money becoming a meme. Can you talk a little bit about how that’s happening? Like what you’re seeing in the community? What are the aspects of this meme? What are you seeing in terms of the development of this?
When the meme was, was introduced, I think it was something like October 2020, it got a reasonable amount of attention on Twitter, but then it kind of got lost, and people forgot about it. Until I was part of a Bankless episode discussing ultra sound money. And this is where kind of the concept really became popularized, and people started talking about it again. And one of the things that happened, which is very interesting, is that David Hoffman, I believe, started wearing the bat signal. Now, what is the bat signal? It’s basically a two emoji, the bat emoji, and the sound emoji. And it was trying to be a play on words where because bats produce ultrasounds, kind of the bat signal is going to represent ultrasound money.
And people just started copying the bat signal all over Twitter. And now we have over 2,500 Twitter accounts, huge accounts, including, for example, the Axie Infinity account, and many, many DeFi founders, and lots of prominent accounts within the Ethereum space, just adopting the bat signal.
We have this meme, which now seems to be spreading across the whole community, and now we have this website and the word ultra sound money was picked up on Bloomberg. It was picked up on CNBC. And so there is kind of a possibility that in the same way kind of bats kind of infected the whole world with genes and with COVID that bats yet again, will infect the whole world with memes. So one is negative, but the other one is as positive.
And actually the it’s interesting because the word meme actually comes from gene. The inventor of the word meme kind of wanted to create a word to describe the propagation of cultural information and kind of the evolution process of these ideas, which follows a very similar pattern to genes. And so I think the analogy is really, really good in a sense that we have these viral ideas that go spread across the world. Right now the meme has mostly kind of infected the Ethereum community, but it is possible that it will kind of go beyond this boundary and infect kind of a wider population.
Well, we’ll have to see yeah, this whole episode has been super, super fascinating. I really appreciate that you came on the show to talk about it. Where can people learn more about you, EIP 1559, and the idea of ETH as ultra sound money?
Sure. So to learn more about me you can follow me on Twitter. I’m @drakefjustin. To learn more about ultra sound money, there’s four hours of content on Bankless — two separate episodes that cover ultra sound money. There’s also the ultrasound.money website. There’s also the ultrasound money, Twitter account. So lots of content for you guys to consume.
Okay. Perfect. Well, thanks so much for coming on Unchained. Thanks so much for joining us today to learn more about Justin and ETH as ultra sound money, check out the show notes for this episode. Unchained is produced by me, Laura Shin, with help from Anthony Yoon, Daniel Nuss, and Mark Murdock.