The 21 Levers of Bitcoin Price- Part III
#11. Number of miners:
Bitcoin price increases with an increase in the number of miners, because it increases competition, thus increasing aggregate hashrate and therefore network security and trust.
Nakamoto Competition VS Realworld Competition:
In a traditional economic models, when there is a consumer demand for a good or service at a profitable price, more producers enter the market and compete on price until the price of the good quickly reaches equilibrium where marginal costs tend towards marginal revenue. This is how prices drop in free markets as competition drives them to the marginal profitability, unless strong externalities interfere.
In Nakamoto Competition, as more miners compete to find a nonce to be able to construct a block, they directly contribute to the aggregate network hash power, thus increasing the resistance to double spend attacks and increasing network trust and thereby it’s value. As network trust increases, network value and therefore, bitcoin price increases. When mining is perfectly competitive, meaning the marginal revenue from mining approaches marginal cost, the cost of mining a Bitcoin is a constant proportion of the price and only depends on the second derivative of the cost function (how the rate of change of the cost changes over short periods of time), and not on aggregate network hash power. At this competitive equilibrium in the presence of an infinite number of miners, Pagnotta and Buraschi show that the equilibrium price of bitcoin will remain a constant proportion of it’s mining cost.
#12. Block reward:
Block reward is primarily what incentivizes miners to mine and secure bitcoin network until the transaction fees rise high enough. How an increase or decrease in block reward affects bitcoin’s price is purely of academic interest and is not very relevant in the real world because for all intents and purposes, bitcoin’s block reward schedule is set in stone.
If however, block reward were to be changed, bitcoin prices reacts in a bimodal fashion. At a constant price of Bitcoin, all else being equal, as the block reward increases, the number of miners increases drawn in by profitability, increasing security and value of the network, raising bitcoin price. But, if the increases in block reward are large, As block reward increases beyond a certain level, supply inflation will lower the price. Pagnotta and Buraschi derive this intuitive result, showing that hash power increases for small increases in block reward (increasing profits), and decreases for large increases in block reward (Inflation reduces the perceived value/profit).
#13. Mining Profitability:
Mining profitability only moderately impacts Price. This may be because when profitability goes up or down, competition causes hash rate entering the network to get adjusted up or down dynamically, driving profitability to the equilibrium, thus buffering it’s impact on price.
#14. Inflation schedule:
This is a controversial topic. An intuitive but controversial insight from Pagnotta and Buraschi is that, to maximize the “market cap” of Bitcoin, it is more important to have a set monetary policy that is totally predictable, than a monetary policy that dynamically targets an optimum market cap.
As long as the monetary policy is set in stone, it doesn’t matter how the inflation decays or how the block reward halvening schedule is set.
This is likely because people adjust their expectations of inflation based on history (Milton Friedman’s adaptive expectations), and can also learn and change their behavior [Expectations and Neutrality of money].
To change the supply schedule of bitcoin to counteract the perceived threats to security model would defeat the perception of immutability and soundness which users value in Bitcoin.
Immutability of bitcoin’s monetary policy and it’s impact on security:
Given that Bitcoin’s inflation schedule, at least as understood now, is set in stone, what happens when block reward expires around 2140? Who pays for network security? We have two options. The cost of validating transactions will either have to be borne by miners, for zero reward, or the transaction fees have to be high enough to support an adequate level of network security. Obviously, the cost of providing network security cannot be externalized to miners. Even though layer two payment systems like lightning make transactions cheap and fast, they reduce the number of transactions on chain, which negatively impacts the fee market. An answer to this conundrum may be to increase the economic density of transactions on layer 1. In this scenario, high value transactions get settled on chain for high fees, and low value transactions will migrate to layer 2. One way to increase the economic density of transactions is the use of high value smart contracts on the bitcoin blockchain, this capturing adequate fees for the base layer. Pegged sidechains such as Liquid and approaches such as RSK are a move in this direction.
In the absence of increasingly complex and valuable financial contracts that capture economic value and therefore high enough transaction fees, the security of the network is vulnerable. To mitigate this, some ideas put forth include dynamic adjustment of blocksize targeting an optimal fee. This approach factors in the current bitcoin price, mining profitability, average transaction size, and derives a block size limit that dynamically adjusts to keep the hashpower above an “adequate” level for network security.
Whatever the mechanism, if Bitcoin transaction fees cannot provide adequate revenues for securing the network, the only other option will be to re-introduce inflation, which will reduce the value of the network, bitcoin price, and therefore it’s economic security.
It’s clear that there are not free lunches and therefore it remains to be seen what evolutionary path bitcoin takes when block rewards decay to insignificant levels.
#15. Price Spirals:
In Bitcoin, network trust depends on hashrate supply, which is sensitive to miner’s expectations of future bitcoin prices, and therefore, large changes in hash power and price can result from any small changes in user’s expectations, miner’s expectations, or both.
Intuitively, miners are less sensitive to FUD, because only those with conviction about the longer term prospects and value of Bitcoin are likely to commit capital and enter the mining market. The divergences between the expectations of miner’s and users lead to two interesting phenomena.
a. Negative price spirals:
If there is a sudden change in expectations about future value or size of the network, for example, if users suddenly become pessimistic about the future network size for any reason, it leads to a reduction in price because people could expect the future network size to be lower. The demand for HODling drops and this reduces current price. But, it doesn’t stop there. Because miners see this price change and reduce their hash power supply dynamically, it could lead to a further reduction in price due to a decrease in network trust, which feeds into further reduced value and price, which in turn affects network size expectations and so on, until a new equilibrium price is reached. This presents an interesting opportunity for arbitrage: When aggregate user’s expectations diverge from miner’s expectations of bitcoin prices, it is time to buy/go long or sell/go short bitcoin.
b. Positive price spirals: Similar to the above, if a sudden increase in future network size expectations happens, user demand goes up, and given a high degree of supply inelasticity, price goes up. This invites more miners, hash rate supply goes up, until a new price equilibrium is reached.
Price spirals present an interesting economic opportunity for traders. Rapid, sudden upward moves as have happened in2011, 2013 and 2017 appear to be driven by this positive price spiral effect.
The anatomy of a positive price spiral:
Necessary Conditions:
- Stable demand/price (200 day moving average for example)
- A sudden and unpredictable “trigger”, such as a geopolitical disruption or a high profile endorsement of bitcoin by a leading financial institution, driving new users to fiat gateways for crypto
- HODL demand: Average coin-age is high. HODLers are not selling
- Miners are not selling
This presents the perfect setup for a positive price spiral.
#16. Difficulty adjustment:
Difficulty adjustment acts as a buffer on Bitcoin price volatility. A miner’s probability of finding a nonce and creating a block is dependant on the new difficulty, and therefore, it either goes up or down with each difficulty adjustment. This acts as a buffer, smoothing out sudden and dramatic swings in hashrate supply, and prevents sudden gyrations in network trust, which may accelerate price volatility.
#17. Lag between miners and traders:
It needs to be noted that trader/speculator expectations of Bitcoin value change much faster in real time in response to external shocks, but miners’ expectations of future Bitcoin value change slower. There are three factors at play here.
- Miners have longer time horizons over which they expect their investments to payoff, compared to speculators and traders.
- Most traders do not factor in changing hashpower and mining profitability into their price expectations.
- Lag between fundamentals and trader’s perception: There is a lag between network fundamentals and user’s perception of network security. Mining supply changes dynamically and in real time, based on price, but, a reduction or increase in hashpower does not immediately translate into a change in how traders value the network.
If hashpower continues to increase in the presence of stable or decreasing prices, it may be the beginning of a positive price spiral.
#18. Halvening:
Does reward halving effect price? This is a hotly debated topic. Much has been made of how the halvening schedule coincides with a coming run up in bitcoin prices. However, this makes no sense under the adaptive expectations model of Bitcoin pricing. If traders know that a halvening happens at a predetermined schedule, it should be assumed to be already factored into current price. Then, halvening may or may not contribute to any price changes. Pagnotta and Buraschi demonstrate that depending on whether the price is above or below equilibrium price adjusted for reduced supply, halvening may or may not have an effect on Bitcoin price (thanks to @stephanlivera for providing the counterpoints to this view).
#19. Hashpower Competition:
Any competitor that uses the same SHA 256 algorithm competes with bitcoin for mining hashpower. So, if a new coin using SHA256 is more profitable to mine, it draws hash rate away from Bitcoin. When the aggregate hashpower mining the alternate coin exceeds bitcoin, it suddenly introduces an existential risk to Bitcoin. The miner with more than 51% hashpower on bitcoin network can perform 51% attack on Bitcoin, while simultaneously shorting it, and not risk losing their committed ASICs go to waste because they can switch back to mining the altcoin. This is the risk that currently exists for all altcoins that use the same hashing algorithm as a more dominant coin. Bitcoin is not exempt from this game theoretic risk.
#20. Regulatory shocks:
Regulatory shocks impact Bitcoin prices, but in a complicated way. If a government bans Bitcoin, it could have an immediate effect on the citizens, as the risk averse users sell their coins, dropping demand and driving the price down. However, in the absence of a comparable monetary instrument that is trust disintermediated, a ban might cause an increase in demand for censorship resistance, and cause a long term increase in price.
A ban in one jurisdcition might even increase demand for censorship resistance globally, when users see the threats to financial sovereignty materialize in a highly public manner. The recent ban being considered by Indian government, and how price changed in response provides an interesting case study in this context.
#21. Asymmetric risks:
Asymmetric risks such as code failure (eg: inflation bug ), a successful double spend attack by a state actor, limit failure etc can significantly damage the perception of security and trust among users, miners and other stake holders.
A derived insight from this is that Bitcoin will be much less volatile in the long run than POS coins, as there are more unknown risks to POS systems than simple POW systems like Bitcoin.
Putting it all together: Simulating the Bitcoin Machine
As we’ve seen, even though the closed Bitcoin economy is orders of magnitude simpler than any real economy, the interactions of these simple elements lead to extremely complex and unpredictable behaviors. To be able to successfully model and understand the bitcoin economy, we need methodologies to track all the above variables accurately and consistently across time, and apply machine learning approaches to derive the relationships between those variables and prices, in the long run.
Most of the complexity involved in generating training data for models such as these usually lies in acquiring data, but the Bitcoin blockchain provides a rich public dataset that is immensely valuable in this regard. This presents an unprecedented oppurtunity to leapfrog our understanding of the economics of decentralized networks.
We, at Portal are dedicated to understanding of how the bitcoin economy works and evolves over time. If you would like to collaborate on this effort, email me at chandra@getportal.co.