Unit 7 Lecture Notes

The Bank of International Settlement has recently published a report in which it concludes Bitcoin's incentives model is unsustainable and likely to lead to security failure if Bitcoin were to grow in economic importance. The report is largely based on a recent paper by Chicago School ecnomist Eric Budish, which finds that bitcoin's security model will be vulnerable to attack as the block reward shifts from offering mainly new coins, as is the case now, to consisting mainly of transaction fees, as is expected in the future.

The BIS fundamentally fails to understand that economics is based on a subjectivist conception of value, and not on an objectivist conception of value. This is the starting point of all disagreement in economics, and the underlying difference between correct Austrian economics and the fiat economics taught at universities and popularized by bureaucracies like the BIS.

As elucidated by the father of Austrian economics Carl Menger, all value is subjective and can not exist outside of human consciousness. Objects have no value intrinsic to them, it is only human consciousness that prescribes value to them. Value is not an objective attribute of objects that can be calculated like mass, temperature, or volume. It can not be computed objectively because it is constantly shifting in the human consciousness as time passes and conditions change. Value is determined at the margin, at the specific time and place that the valuing individual is making the decision.

On the contrary, all the main non-Austrian schools of economic thought hold value to be objectively determined. The Marxists think value is determined by labor inputs, while most other mainstream economists think of it as a function of production costs. These schools of thought conflate value with price, and thus assume that both are determined by the cost of production.

From the mainstream perspective, producers produce things at a certain cost, and consumers then need to pay that price to compensate them for these goods. From the Austrian perspective, humans subjectively value things, and producers try to supply them at that price.

Unsurprisingly, the BIS bases its critique on the work of a Chicago school economist. While Chicago economists are generally viewed as pro-free market, their strictly objectivist and positivist methodology has very little in common with the Austrians. The paper makes the classic mistake of putting cost before value. In reality, there is no fixed bitcoin security expenditure that is needed for proof of work to successfully protect the network. It is the very fact that people subjectively value bitcoin that creates demand for holding it and for transacting with it. The bitcoin asset cannot be owned outside of transactions confirmed in bitcoin blocks, which inevitably creates a market for this scarce block space. Bitcoin's difficulty adjustment algorithm ensures the scarcity of this block space (and thus the bitcoin to ken itself) by raising the hash power, and thus the cost, required to produce these blocks. The cost to produce bitcoin blocks is merely a reflection of the market's valuation of bitcoin, which is ultimately the subjective value people place on it when transacting with it on the market for other moneys or goods and services.

If the market places a value on bitcoin block space, an economic incentive will exist for miners to provide this block space securely. The manner in which users will pay for this block space may differ, but the cost is real nonetheless. In all markets, the presence of demand incentivizes entrepreneurs to find the most effective ways to provide the good that people want; the costs and the methods of payment can differ widely, but if the demand exists, the good will be supplied.

Consequently, if there is enough demand for holding bitcoin, then demand will exist for transacting it widely and people will pay the transaction fees necessary to get their transactions into blocks. The notion that block space will go unbid despite their desire to obtain and hold on to their bitcoin is absurd. The BIS emphasizes its deep ignorance of economics and prices when it presents a scenario in which demand for bitcoin is so high as to necessitate massive security expenditure, while demand for block space is nonexistent. In reality, the opposite is always the case. Block space is very scarce and people are constantly finding new ways to use it more resourcefully. This demand is inextricably linked to demand for bitcoin: if demand for bitcoin increases, transaction fees will go up and push scaling solutions onto the second layer, making on chain transactions more valuable settlement transactions which can pay higher transaction fees.

Due to the difficulty adjustment algorithm, the cost of making a bitcoin block is always going to hover around the value of the total reward offered by the block (including the block reward and transaction fees). Given that the average block today is around 1 MB of data and has a total reward of around $50,000, the going rate for a single byte of data on the bitcoin blockchain is around $0.05, making it the most expensive byte of data in the world. By comparison, a byte on a commercially available hard drive is worth around a trillionth of that.

Should demand for bitcoin exist, then demand for bitcoin blocksize must exist because it is the only way in which bitcoin can be owned and transacted. It is perfectly feasible, of course, that demand for bitcoin might one day decline, or even collapse. In such a case, it necessarily follows that bitcoin's value will decline enormously, and the value of its block space will follow. The network could fail due to a collapse in demand, as discussed in the sections below, but that is irrelevant to whether the mining is being rewarded mainly through inflation or transaction fees.

As it currently stands, compensation is incurred in the inflation that will dilute the value of your coins as a percentage of total bitcoins. Even if they don't think of it that way, it is happening. New coins come on the market every day and depress the price of existing coins, effectively devaluing holders' coins. In the future, the majority of the cost will shift toward the transaction fee needed to obtain your coin, and there is no reason to presume that the market participants who desire the block space necessary to own bitcoin would not pay for it using this other method. There is a real cost to bitcoin which holders are happy to incur because bitcoin is still useful even after taking these costs into account.

If users don't pay transaction fees, then miners won't solve the proof of work problems and transactions won't confirm. This will put pressure on coin owners to pay transaction fees so their transactions get confirmed, and fees will rise.

We already have evidence that strongly suggests bitcoin users will be happy to pay transaction fees. In December 2017 during the last bitcoin bull market, fees rose to around $50 per transaction. Yet despite this increase, there was still very high demand for transactions, which suggests that if people want to hold hard money the transaction fee has a lot of room to grow. If one were to look at the exchange fees people usually pay to buy bitcoin around the world, we find that they are usually much larger than the on-chain transaction fees. Bitcoiners still have no problem paying these extra fees, so it is hard to imagine them giving up on bitcoin because on-chain fees have increased. Premiums for buying bitcoin in places where exchanges do not operate are even higher, and it is not uncommon for buyers on local bitcoins to accept a 10 or 15% markup.

If my contention is correct that bitcoin is the hardest form of money ever invented, it is absolutely inconceivable that demand for it will be destroyed by people's realization that they cannot use this technology for free. Every form of money transfer will involve some transaction cost and bitcoin is no different. If people value bitcoin, the economic incentives of the system have proven resilient enough to motivate people to spend the resources needed to keep their network secure. If Bitcoin dies, it will not have died because of misaligned economic incentives (high transaction fees). It will have died because the demand for it has declined.

If demand for bitcoin declines or disappears, then the price will likely crash and Bitcoin will collapse and/or be attacked, regardless of if the miners are being paid in inflation or transaction fees. But if bitcoin continues to appreciate for the next 20 years, even at a rate no more than one tenth of its historical growth rate over the past ten years, it will become a global settlement network with value in the trillions of today's dollars. Would people not be willing to pay for the daily settlement of hundreds of billions of dollars across the world?

The best way to gauge the willingness to pay for these fees is to look at settlement costs across the world today. The only real alternative to a bitcoin payment, as a form of hard cash whose value isn't the liability of a government, is the settlement of gold cash reserves, a hugely expensive process. Bitcoin transaction fees are an inconsequential rounding error compared to gold transaction fees. Given the unique service it provides, there is enormous scope for the growth in transaction fees on top of the bitcoin network, which makes the BIS' concern trolling sound quite misplaced, if understandably motivated.

One counter-argument here is that transaction fees might provide some money to miners, but they will not be sufficient to attract enough mining hash power to protect the network. The mistake here is to assume that a fixed amount of electricity or hashrate is needed to secure the network, when in reality no such stable level can exist because computing is a highly competitive industry where the cost of hashpower is always declining. The network hashpower that successfully protected Bitcoin from attack in 2014 is a tiny fraction of the total network hashrate today, and yet it was sufficient in 2014. To be secure, Bitcoin does not need a fixed sum of electricity or hashrate; instead, it needs to create a liquid market in electricity and hashing power that constantly attracts a serious amount of capital infrastructure to produce mining hardware. By simply providing a highly liquid instrument as a reward for expending electricity and processing power, Bitcoin continues to attract the most efficient producers of electricity and processing power to monetize their resources. As long as this unique market continues to exist and offers valuable rewards, it will make any attack considerably expensive and unlikely to succeed. In particular, Bitcoin's unique impact on the electricity market, as discussed in depth in TBSRB3, means that Bitcoin is an insatiable buyer of any cheap electricity that exists anywhere in the world. Whereas any attacker will need to mobilize enormous amounts of expensive energy in centralized locations to try to attack the network, Bitcoin can draw on the cheapest sources of energy in many locations worldwide by offering rewards for selling electricity that producers would not be able to sell elsewhere.

According to the BIS, the limit in bitcoin transaction fees is a result of bitcoin's inability to scale. The BIS divides an incorrect estimate for security costs by the number of transactions that bitcoin can perform to calculate the fixed cost per transaction. Since the bitcoin subsidy is scheduled to run out, they reason that the cost of securing the network will have to be divided by the number of transactions and that only if people pay that transaction fee will Bitcoin survive. This narrowly defined formula itself (let alone the incorrect cost estimate as discussed above) clearly shows that the BIS is unfamiliar with Bitcoin's layered scaling approach. The number of on-chain transactions is not a meaningful limit to how many transactions can be carried out with bitcoin, because as explained in The Bitcoin Standard, Bitcoin's scaling will likely happen on second layer solutions, in a way somewhat similar to how gold banking scaled. Under a gold standard, not all transactions took place through physical gold moving hands. Physical gold was largely stored in banks, and for each movement of physical gold used to settle many transactions between financial institutions, financial instruments backed by that gold would change hands many times more. There is no reason why Bitcoin cannot scale like that, and in that case, each bitcoin transaction cannot be compared to individual consumer payments, but to large settlement payments between financial institutions. If each on-chain bitcoin transaction is settling for many thousands of individual consumer payments, then even infinitely tiny transaction fees on consumer payments could add up to very large fees for individual on-chain settlement payments.

The BIS here is making the mistake than many bitcoin purists often commit, which is to suppose that bitcoin can only succeed and operate if every interaction with it is entirely trustless and decentralized, and if every user is able to make a completely trustless permissionless payment on its main chain. While this sounds nice in principle, in reality the level of security of a bitcoin transaction is absurd overkill for the vast majority of transactions that humans conduct in everyday life, for which less reliable systems are acceptable (even with the occasional security failure). There is no need for a coffee salesman to require on-chain verification of your payment, and the current credit card payment system is much cheaper and faster; even with a regular amount of small fraud, it continues to be a more effective solution for small consumer payments. The value of Bitcoin's decentralization is not in that every consumer purchase is uncensorable and trustless, but rather that it helps the network resist government attack and capture. Some purists seem to think the choice we have is between a world in which everyone is able to trustlessly use Bitcoin's on-chain base layer for all their transactions (no matter how trivial), and a world in which only the base layer of bitcoin is trustless and other layers involve trusted third parties. If that indeed were the choice, any bitcoiner would of course prefer trustlessness for all. However, engineering reality seems to suggest that the choice is actually between Bitcoin being trustless only at the base layer, or a fiat monetary system which is government-controlled at all layers.

If Bitcoin's "only" contribution is to make the world's monetary system's base layer and the money supply free from government control, that would be more than enough. The world of payment processing will vastly improve with a free market in banking and money, but even if nothing improves in it, bitcoin would still be a world-changing success. Trustlessness and immutability are not simple engineering features that can be copied and replicated, and the only proven example of a trustless system we have so far is Bitcoin's on-chain transaction, with a capacity of around half a million transactions per day. The idea that we can scale that level of security is becoming less tenable with time, but that is not really a problem that hinders the core proposition of bitcoin. The level of security bitcoin provides is only really necessary for the most important transactions in the world, while current security arrangements are ok for most coffee purchases.

Beyond the economic incentives for mining bitcoin, the deeper web of economic incentives to run and maintain bitcoin is what makes such a failure unlikely, even if the BIS' economic analysis were correct. If Bitcoin's proof-of-work were to prove compromised after block subsidy diminishes, and if mining hashrate began to decline as the cost to the network of hashrate became more expensive, it would be a clear threat to bitcoin; in such a case, it should not be very difficult to get bitcoiners to agree on a fork that corrects this. Forks are extremely hard to implement with bitcoin for upgrades, but that would likely change in the case of emergencies.

Ultimately, doomsday scenarios in which Bitcoin fails due to a technical design glitch don't take into account the economic incentives to keep the system successfully running. As long as demand for digital hard money exists, many millions of people around the world are motivated to find solutions to continue to make it exist. Bitcoin has a very straightforward technical requirement to operate, and it performs a very simple job that requires very little and has enormous incentives backing it.