The Bitcoin Maxi Price: What’s the maximum price of Bitcoin?
A lot has been written about trying to value and find the maximum price for Bitcoin. Whether valuing it as digital gold or the worlds future reserve currency, most models approach valuing Bitcoin based on maximum utility. Here I’d like to explore valuing the maximum price of Bitcoin based on security.
This method of valuation is based answering one question: what is the maximum market cap for the Bitcoin network where the fees paid to miners is enough to keep it secure? To put the question into numbers: if the Bitcoin network is worth $100 trillion (one BTC = ~$4.7m) but miners only get paid $10m in fees to secure the network, there’s a lot of incentive to attack the network. Eventually the market cap of the network has to reflect the security expense of the network. If the security budget isn’t high enough to justify a certain market cap, the market cap should decline to reflect the security budget. Evaluating the future security budget is how we can determine a maximum price for Bitcoin.
Prior work
This question isn’t new. Bitcoin community members have been discussing the fee market and how it affects the maximum price of Bitcoin for quite a while. Dan Held even quotes Satoshi Nakamoto discussing it:
“In a few decades when the reward gets too small, the transaction fee will become the main compensation for [miners].” —
Bitcoin’s security is fine, Dan Held (I couldn’t find the primary source for this Y.Z)
A milestone paper on this topic was released in 2016: ‘On the instability of Bitcoin with block rewards’. In it the researchers present the theory that when block rewards decline and miners are rewarded mainly by transaction fees, even should transaction fees be large enough to justify securing a high market cap network, there will be incentives for miners to create orphan blocks to maximize rewards, hurting network reliability.
While this is a major issue, the much larger one is presented by the very question of how much miner revenue can transaction fees even bring in?
Dan Held discusses this and references much previous work in his article ‘Bitcoin’s security is fine’. He reaches the conclusion that based on a bottoms up calculation of transactions that will come onto the Bitcoin network, such as commerce (with Lightning channels), bank wires, purchasing real estate and banking services. These will create enough volume and thereby pricing power for Bitcoin block space to create enough transaction fees.
What’s the number in Held’s mind where enough transaction fees secure the network? Well here is where things don’t quite add up. Earlier in his article he makes a statement that I agree with:
I hypothesize several hundred billion, in present value USD, would be an adequate security budget since it would be very difficult for a government to justify such a waste of an expense to just 51% attack the tip of the Bitcoin blockchain.
And that makes sense: if you want to secure $10 trillion in assets, spending $200 billion on security makes sense. That’s only 2% of the assets’ market cap, but it’s a high enough spend to make it quite costly and hard to successfully attack the Bitcoin network. It firmly puts an attack on Bitcoin in the nation state rubric.
But Held’s calculations towards the end of his article show a different picture, there he models a security spend of merely 0.365% of the networks’ market cap for a security budget. For comparison, the US spends 11–15% of its budget on security. Other nations who fear attacks spend far more.
If security matters, you have to spend money on it.
Using such a low percentage of the Bitcoin network’s market cap as a security budget marks Bitcoin as a honey pot for nation states and rich financial institutions.
Let’s model this out to illustrate the point of why $36B in annual mining revenue doesn’t defend $10T in market cap. $36B annually, means a daily defense budget of just under $100m. If Bitcoin is a $10T asset, it’s safe to assume that it has a liquid derivative market where smart investors can bet big time, in a highly levered way (using options, CDS or futures) on an extreme drop in Bitcoin’s price. Assuming that, and knowing that there are many multi billion dollar hedge funds out there — this could be the next ‘big short’. They don’t need Bitcoin to be halted permanently to make a high ROI — just to attack the network for a few hours or days. They would pay a multiple of that $100m to rent ASICs, GPUs, TPUs and any other chip that can be used to 51% attack the network. They don’t even need to successfully double spend to shake the price of Bitcoin and cause extreme drops in price.
If this scenario sounds crazy to you, we saw this happen in 2022 with the collapse of UST/Luna. A smart investor chose an opportune time of low liquidity and used $1B to push the UST/Luna system just over the edge enough to cause a collapse in confidence.
This scenario is even more likely with a nation state actor. With only several hundred million dollars, North Korea, China or Russia could halt the payments network of the western world (that’s assuming Bitcoin reaches that status).
So 0.365% of the asset price isn’t enough to defend a $10T network.
Lyn Alden also analyzes this issue and asserts similar to Dan Held that:
“If bitcoin reaches a state where the average transaction fee is about $10, it would translate into $1+ billion per year towards miners. If we add a zero, and the average transaction fee gets to about $100, it would translate into $10+ billion per year towards miners. For reference, as of the first couple months of 2021, the typical fee has been up to $20+.”
However Lyn’s analysis on the transaction fee market is based on a bottoms-up calculation. A more appropriate way to calculate this is using the exist market base rate — how much in fees for these transactions are paid today. When calculated that way, her analysis does not hold true. as I’ll model out below. There are simply not enough fees to go around based on current market dynamics.
Other models for valuing Bitcoin network security
Nic Carter of Castle Island Ventures gave a great presentation in 2019 at MIT where he offered three models for modeling security (This presentation seems to anchor both Lyn’s and Dan’s wok).
The three models are:
Threshold: At a given level of security spend, Bitcoin is assumed secure Stock: Security spend should be indexed to the value of Bitcoin itself
Flow: Fees must be large relative to transactional volume
For the threshold model to work, the threshold needs to be very high. As Dan Held says:
I hypothesize several hundred billion, in present value USD, would be an adequate security budget since it would be very difficult for a government to justify such a waste of an expense to just 51% attack the tip of the Bitcoin blockchain.
That’s a statement I can get behind. But is that actually feasible?
Modeling maximum transaction fees
Instead of hypothesizing the bottoms up TAM of what Bitcoin addresses, I think it’s far more accurate to use the current revenue for the very same market that Bitcoin is competing with. If we believe that Lightning will replace credit card transactions and that Bitcoin will replace fiat Fedwire or bank transfers — let’s look at how much revenue those fees generate today.
Below is a high level calculation of these fees, on an annual basis, ex China.
Payment processing revenue, 2021:
Visa: $24.1B
Mastercard: $18.88B
Square: $17.6B
Paypal: $25.3B
Western Union: $5.1B
Total: $91B
Bank revenue, 2021:
Using the top 25 banks in the world and removing Chinese banks brings us to a total: $1,365 trillion in 2021 revenue. Naturally this is just a proxy for global bank revenue: it’s only the top 20 banks. But this is just a thought exercise anyway.
How much of that $1,365T is revenue that comes from fees that Bitcoin competes with (wire transfers, bank accounts)? I don’t know, I’m not a bank analyst. But for the sake of this thought exercise — I’m trying to find the higher upper bound for Bitcoin fees. Let’s take a high guess, and say 30% of bank revenue comes from fees (that sounds abnormally high to me).
That brings us a total of global bank fees that Bitcoin can win over of $409B.
What’s the maximum price for Bitcoin?
Calculating the total Annual fee revenue for Bitcoin
Adding the two fees together we get a global annual fee market of $500B ($409+$91). If Bitcoin captured 100% of that amount, it would be safe to say that we’ve reached the threshold Dan Held put down as an adequate security budget. But that doesn’t seem very likely to me.
How much of this market goes to Bitcoin? Let’s run through some scenarios:
- 5%: $25B annual fee market for Bitcoin ($68m a day)
- 10%: $50B annual fee market for Bitcoin ($136m a day)
- 20%: $100B annual fee market for Bitcoin ($272m a day)
- 50%: $250B annual fee market for Bitcoin ($680m a day)
Only scenario 4 gets us into the numbers where we reach a threshold level for large market caps of Bitcoin. The others are too low which leave us firmly in camp #2 of Nic Carter’s risks: stock. This is the scenario where total market cap is a function of security spend.
Calculating the maximum price for Bitcoin based on security budgets
What’s a reasonable multiple of security budget for the total market cap? Let’s hypothesize some more.
Throughout the past 24 months the Bitcoin price has averaged $36,379.34. Block rewards on an annual basis have been 328,500 * 36,379.34 = 11,950,613,190. Almost $12B. The average market cap over that time period has been $683B. That’s 1.7% of the market cap spent on security, or a 58 multiple security. As in take your security budget, multiply by 58 and that’s what the market is valuing that security worth.
Is the 1.7% ratio of the past 24 months indicative of future rates? Hard to know. Lyn Alden points out that this percentage has been coming down over time.
But that was during the formative years of Bitcoin. Who’s to tell if 1.7% is where the rate of decline finally changes trend and starts going up or flattens out? In fact, two factors point to a higher percentage in the future:
- Miners are still conceptually long Bitcoin price, and therefore more likely to mine for lower fees than they would in the future when they expect less price appreciation.
- Bitcoin is still a nascent network without a blatant threat vector. As it grows more dominant, we can expect more attacks and therefore will need a higher security budget expressed as a percentage of total market cap.
On the other hand, the market has been quite unique over the past 24 months and in many ways the past two years have been indicative of nothing.
The percentage of market cap, or its inverse, the security multiple is the critical number that anchors the upside for Bitcoin while still maintaining security.
Here’s an example of how they can be used: The gold bull case for Bitcoin often claims that Bitcoin is digital gold and therefore will be worth as much, which is ~$8T. Using our multiple of 58, the fees to secure those assets would be $138B, implying that Bitcoin would capture at least 27% of its fee TAM.
When does this start to matter?
Until the halving cycle in 2028 block rewards should be enough to incentivize the necessary security to protect the network (assuming price appreciation continues as in the past). But afterwards, we start getting into very tricky territory. Markets are forward looking mechanisms, and we can expect the market to start caring long before 2028.
Based on this section and the forecasts for the fee market, I think a total market cap of $5T is a reasonable best case scenario over this decade. Anything much higher would require a security spend probability doesn’t favor (capturing 20%+ of the fee TAM we calculated). With a range of $2 or $3 trillion far more likely. This would put Bitcoin’s maximum price range, while maintaining security in a range of $100,000 — $240,000.
Other factors for consideration when calculating Bitcoin’s maximum price
- It’s important to note that an attack on network security, even a successful one doesn’t mean the network crashes and that its value goes to zero. For that to happen, the cost to attack would be prohibitively high as the attack would have to succeed for multiple days. But any successful network attack will greatly decrease the narrative and faith that people have in the Bitcoin network, and that would be harder to recoup.
- Attacking and defending the network do not have the same cost. While regular miners have all the appropriate hardware and electricity setup (ASICs and cheap electricity), an attacker does not. They would either have to buy ASICs off the market (a costly and not very feasible thing to do in a short amount of time) or use available infrastructure such as existing public clouds (AWS, Azure, GCP). While I expect more specialized chips to be available for rent in future years (and perhaps even ASICs), any attack not using dedicated Bitcoin ASICs has to expend significantly more compute resources.
- There are other use cases for Bitcoin beyond payments and stores of value. Examples include a place to store data in ‘the’ timestamp blockchain. But even if these do play out, which I’m skeptical about, these won’t be relevant in the time frame we’re discussing. They won’t bring a material amount of on chain fees by the time the second halving occurs in 2028.
- Off-chain miner revenue and incentives are a factor that could come into play over the next several years. This would include other forms of revenue (non BTC) that could be paid to miners to incentivize them to mine. Such as service fees by whales, custody services or the electric grid. There could also be nation states who participate in mining for national defense reasons (so as to avoid financial sanctions).
- Currently the fee market is 1.5% of the block rewards. To make up for the lost block rewards we’ll face in 2028, fees need to make up 4.69 BTC per block. This means that fees need to grow to grow by 48x to reach the same amount of security. Bitcoin has strong network effects, but 48x growth over 6 years is a heavy lift.
Conclusion: Bitcoin fees and security
A lot has been said about the security of Bitcoin when block rewards decline. Despite that, as time progresses, it’s worth revisiting again. The landscape of fees in mid 2022 sadly does not reflect the progress earlier works envisioned and hoped for. In fact, we’re back to square one. Despite the rising increased use of the Lightning network.
While there are many ways to forecast the size of Bitcoin’s future fee market from a bottoms up perspective, when we revert to using a base rate as measured by current revenue in the industry today, it’s quite apparent that the fee market is not large enough to justify a Bitcoin market cap much larger than gold. Even under the best case scenario. There’s a price in the Bitcoin Maxi attitude, and that price is less applications that use Bitcoin and therefore a lower fee market. It might just be that the Ethereum community has a better grasp of this dilemma than has been credited to them.
I design tokenomics for crypto protocols and am putting everything I’m learning into a super in depth course on Tokenomics. If you’re interested in my free email course on tokenomics, sign up here!
