21M Bitcoin and the Promise of Scarcity
In Bitcoin, the promise of a limited supply is great. This was one of the biggest things that initially appealed to me when I heard about Bitcoin. “No more than twenty one million bitcoins will ever exist!” But can this promise be kept in the long run?
Has the Promise Already Been Broken?
Some would say that Bitcoin’s scarcity has already been violated since forking creates multiples of the 21m coin limit, with each chain’s token commanding a different market price. But setting aside this concern over forks, let’s focus on a more modest goal: Can even one bitcoin chain (such as BCH) strictly maintain its issuance schedule over the next hundred years or longer?
The issue is that cryptocurrencies are based on software, and software can always be changed. Some have questioned if Bitcoin’s issuance schedule can maintain security on either Bitcoin or Bitcoin Cash in the coming decades, but we’ll deal with that question later. First, let us examine if it’s likely or not for users to even agree on what is already in place.
Phrases like Bitcoin’s “social contract” or “economic policy” get thrown around a lot, but what do they mean? In a word: change. Any major change to how Bitcoin works is often equated with a modification to the core contract or policy.
Bitcoin Isn’t Set in Stone
Since Bitcoin is based on software, the only real agreement contract, or policy, is a de-facto standard in which everyone agrees to run software with the same rules as everyone else. This purportedly makes Bitcoin “hard to change” and there may be quite a bit of truth to that. However, we know that these social agreements are not set-in-stone revelations sent by the gods. Inevitably, the winds of change blow. Disagreements brew. Fresh actors can enter the ecosystem. Software developers can create new rules.
This leads to a range of possibilities. One possibility is that long-revered tenets of the “social contract” may be cast aside, even if popular. Popularity, in the first place, is heavily influenced by the public discourse, which in turn may be influenced by disinformation and propaganda. We saw this play out during the BCH/BTC fork of Bitcoin.
Forks are the free and fair mechanism by which Bitcoin resolves irreconcilable differences. Regardless of how manipulated the news may be, the market provides an open-ended and continual mechanism that determines the value of the tokens on competing blockchains.
Upon reflection, it seems that “social contracts” and such are more wishy-washy than we would like them to be. Leading up to the BCH/BTC fork, many Bitcoiners simply assumed that, “of course”, the blocksize would have to be raised.
But those were the assumptions of the early community, and opinions change over time. The famous “overton window” is one way of looking at some of the dynamics involving changes of public opinion. An idea can be radically unthinkable at first, but over time see itself become at least debatable, and finally commonly accepted.
Investors often buy coins with the presumption that the network rules are going to be there for the long term, but this isn’t so; the rules of a ledger can change at any time, at least in theory.
In practice, there are checks and balances built into crypto. A change of rules is considered a fork, and forks generally only happen at specific pre-scheduled times. Developers cannot just publish “any code they want” because the miners won’t necessarily run it. And even if they do, if there is any subset of the community which prefers to use the old software (or an alternate software), then the fork becomes contentious, leading to a chain split.
This splitting mechanism protects investors, who automatically get coins on both sides of a split. Although it should also be said that a split isn’t always a net positive outcome for investors. For example, if a split causes a community to lose too much network effect, then the sum value of the coins post-fork may be less than the pre-fork price.
Therefore, we can conclude that there is never a guarantee to the continuation of any blockchain’s rules. The best we can do is rely on rules that seem to have some stability, and enjoy a measure of hope that the rules we cherish the most will continue on in some form.
Even if Everyone Agrees, is Zero Inflation Achievable?
Blockchain governance questions notwithstanding, the next issue that confronts us is: Is it economically feasible to keep the original Bitcoin issuance schedule in place in the far future, even as block rewards tend toward zero?
A corollary question is: are transaction fees alone going to be sufficient to secure the blockchain in the future? (This question can be asked of both Bitcoin BTC and Bitcoin Cash BCH).
One point of conflict in the BTC/BCH split dealt with this question specifically. The BTC Core developers argued that without capping the supply of block space, fees would be too low. By instead limiting the transaction capacity, this creates a fee market and thus sustainable levels of security, assuming users will continue to pay the high fee rates on the BTC chain, rather than use an alternate blockchain.
In practice, this theory has been demonstrated to be true, at least to a degree. We have seen Bitcoin blocks with fee totals comparable to (and in some cases exceeding) that of the block reward subsidies. We have also seen some willingness in the market to keep using BTC regardless of the high fees.
Interestingly, the fixed supply of the block space for transaction plays into this demand-supply equation as demand sits just below supply. You might expect demand to completely collapse, but it looks like the theories that people will pay more for transactions (because they are on the BTC chain) has some truth to it in practice.
Of course, that is only true based on a few years of data and it could change at any point in the future; for example, if crypto users decide they aren’t getting enough for their money. That seems somewhat unlikely at the moment; currently BTC is used primarily for speculation and as an inflation hedge, so its users do not need fast or cheap transactions.
The Unusual Case of Bitcoin BTC
At the same time, BTC seems to depend more on its network effect for investor dollars, rather than the actual functionality it allows. The price of any coin is based both on its network effect as well as its functionality, so this is not unusual in and of itself. But, BTC is the only coin I am aware of that has intentionally high fees. While this theoretically does help address the fee problem, it is unclear how long BTC can maintain this economic policy without turning off users and investors.
The other approach to fees is the original one: Have tons of transactions and thus keep fees relatively low but make up for it in volume. As Satoshi said “I’m sure that in 20 years there will either be very large transaction volume or no volume”.
Bitcoin Cash is following the original design of Bitcoin in terms of keeping fees low. But how is it working? So far, not great. That’s because there hasn’t been any real growth in the number of BCH transactions. But again, it is really too early to say. Bitcoin Cash may see an explosion of growth at any point in the next several decades.
The idea in Bitcoin Cash is you want to have a large number of transactions. But what if that doesn’t happen quickly enough? One other relevant factor is the price of the coin. If the coin price keeps doubling every four years, then the security level remains unchanged in dollar terms.
Both rising coin prices and rising transaction count can help. Either of these things can offset the diminishing block rewards. But what if neither is sufficient? That seems like it’s the worst-case scenario for Bitcoin playing out, but not everyone believes this is so unlikely. There are some in the crypto community who believe that the fixed supply model is simply not achievable.
Some coins, like Ethereum and Monero avoid this problem with the use of a so-called “tail emission”, which means block rewards will continue forever. While the block rewards found in advanced years are planned to be small, since they go on, the coin cannot advertise itself as having a fixed supply.
What other remedies and solutions exist? One simple idea is just for miners to raise fees on their own. It is certainly possible for miners to do that without a fixed block size. If the developers can enact an economic policy, the miners can do the same.
Of course, this runs into the same issue of fees getting too high for the chain to reliably support the “peer to peer cash” usage, but perhaps there is a market-driven sweet spot where fees can rise for larger transactions while still allowing cheap or even free transactions. In that scenario, obviously the existing pricing-by-kilobyte fee scheme would have to be overhauled.
Going Beyond Proof-of-Work
Beyond the straightforward remedies of higher prices, more transactions, or higher fee rates, increasing security would start to involve looking at more esoteric and radical changes to the technology. Proof-of-work is a powerful tool, but has the limitation that it requires a majority of the network to be honest. This is more of an issue on a minority chain like Bitcoin Cash that shares its hashing algorithm (SHA-256) with BTC.
For years, Bitcoiners have loosely discussed adding some kind of “proof of stake” element to Bitcoin security. In recent years, there have been many interesting developments in crypto technology and consensus mechanisms.
For example, the Avalanche coin (AVAX) brings a new method of achieving distributed consensus. And some have discussed trying to bring elements of this technology over to Bitcoin Cash in an attempt to “bolt it on” to the existing PoW security.
But, it should be noted that all these schemes are ultimately variations of Proof-of-stake, because they require coin holders to participate and provide security based on their holdings.
The ultimate goal of any such scheme would be to ensure that an attacker had to have both 51% of the hashpower AND 51% of the circulating supply. It doesn’t seem like we can do better than that, at least on the surface, if “work” and “stake” are the two fundamentals at play. But this would provide quite an increase in protection.
A less invasive method may be to use coin age as a determinant in qualifying a block for having sufficient work done. This idea was first proposed by Gavin Andressen but hasn’t received much discussion. However, I think it would work well if implemented as overall coindays of the blocks’ transactions. Other schemes involving both coin age plus coin days offer different properties and tradeoffs.
Still more extreme measures can be implemented, such as using time based delay penalties for re-org protection, which I had proposed last year. The idea would be that malicious attack chains wouldn’t be recognized as valid if there was a significant delay between when a node received the block and when it received a competing block of the same height.
But this kind of scheme (and re-org protection schemes overall) is less robust than pure nakamoto consensus. It is possible for an attacker to cause a chain split if alternate blocks are sent to different parts of the network with precise timing. It is also theoretically possible for some internet outtage to create a similar situation. Not to mention that new nodes coming onto the network and syncing do not have knowledge of these time delays. This is sometimes referred to as “weak subjectivity”.
In general, resolving these kinds of potential chain splits caused by impure consensus schemes would involve some centralization — for example manual pool coordination in the event of a chain split attack.
For Now, Business as Usual
I think the bottom line is that nothing is really certain just yet about how the Bitcoin experiment will continue to play out. It has certainly surprised us so far, and may continue to hold many surprises in the future. It’s too early to tell if some or all of the “promises” will come true. We have years or decades to continue researching and observing how things will unfold. The security of the blockchain is something we will keep eye on, but we still have time.