Pessimists about Bitcoin believe that in the long run it does not have security, that is, the cost of a 51% attack is getting lower and lower.
Their logic is that currently Bitcoin's security spending is mainly paid by the newly issued Bitcoins (block subsidies). As the Bitcoin issuance is halved every four years, in order to maintain the same security spending, there are only three ways:
- Increase the number of transactions Bitcoin can process per second. Allow more people to use Bitcoin to generate more transaction fees to maintain security spending, but Bitcoin's small block size and low TPS mean that this approach is not feasible.
- Double the price of Bitcoin every four years. Although the amount of newly issued Bitcoin halves every four years, the price doubling can maintain security spending. This is also impossible. Because, assuming the current price of Bitcoin is $30,000, Bitcoin has 116 years left before stopping issuance, that is, 29 halving cycles. The price of each Bitcoin will rise to $16,000 billion, which is nonsense.
- Increase Bitcoin's transaction fee rate. Since it is impossible to maintain security spending through increasing the number of transactions and the price of Bitcoin, the transaction fee rate can be increased. Pessimists believe that in order to maintain security spending, the transaction fee rate will rise to an extreme level → driving users away from the Bitcoin network → transaction volume decreases → transaction fees decrease → network security spending decreases → miner income decreases, leaving the network → Bitcoin security decreases → users further leave the Bitcoin network → vicious cycle.
Regarding the pessimists' logic, I agree with the first two points. The Blocksize war in the past determined Bitcoin's strategy to insist on small blocks for decentralization. And the price of Bitcoin cannot rise to astronomical figures like $16 trillion.
However, there is a fatal ambiguity in the third point: In order to maintain security spending, the transaction fee rate will indeed rise, but how high will it rise? Will it rise to extremes that drive users away from the Bitcoin network?
This article will analyze the transaction data of all halving cycles since the establishment of Bitcoin, then make the most conservative hypothesis, and try to answer the following questions:
- Historically, what is the situation of Bitcoin's transaction fee rate?
- With the decrease in issuance, will Bitcoin's transaction fee rate surge to the point of driving users away in order to maintain current security spending?
- Does Bitcoin have long-term security?
What is the Effective Fee Rate#
If you can't measure it, you can't improve it.
--- Peter Drucker
Usually, the Actual Fee Rate actually paid by users for ttransaction is the actual fee divided by the transaction amount, that is:
However, Bitcoin network security spending consists of two parts: the transaction fee paid by users to miners and the block subsidy given by the Bitcoin network to miners. The reason why users can enjoy "low" transaction fees is because the Bitcoin network subsidizes users. As the subsidy decreases, users ultimately need to bear all the costs themselves.
Therefore, the Effective Fee Rate is defined as:
The effective fee rate is the fee rate assuming no block subsidy and users bear all the costs. It can more effectively reflect the fee rate situation of the Bitcoin network.
The Historical Effective Fee Rate of Bitcoin #
Analysis by Year#
This section analyzes transaction data from January 3, 2009 to June 12, 2023, covering 5,266 days (with 5 days of missing data), 793,661 blocks, and 850,615,471 transactions.
From the historical data, it can be seen that due to the issuance of Bitcoin subsidies to miners, the main security spending on the network was borne, and the actual fee rate paid by users was very low. The highest single-day actual fee rate occurred on February 3, 2009, at 0.27%. The highest annual average actual fee rate occurred in 2017 at 0.0108%.
Occasional fluctuations were caused by blockchain congestion, such as on May 7 this year, due to the hype of BRC-20, the actual fee rate around block 788695 rose sharply, even exceeding the block subsidy.
In the early years of the Bitcoin network, the number of transactions contained in many blocks was very small, and each block generated a block subsidy of 50 BTC, resulting in an extremely high effective fee rate for Bitcoin on certain dates. For example, on September 16, 2009, the total transaction amount was 1 BTC, while the block subsidy was 3600 BTC, so the effective fee rate was as high as 360,000%.
However, with the development of the Bitcoin network, such extreme situations gradually disappeared, and Bitcoin's fee rate also became more reasonable. Since 2016, Bitcoin's annual average effective fee rate has ranged between 0.02% and 0.1%.
Analysis by Halving Cycle #
Previously, according to the year, the actual fee rate and effective fee rate of Bitcoin were analyzed. Now analyze by halving cycle. So far, Bitcoin (June 12, 2023) has undergone three halvings, and the block subsidy has decreased from the initial 50 BTC to the current 6.25 BTC:
| Time | Number of Halvings | Block Subsidy (BTC) | Block Height |
|---|---|---|---|
| 2009-01-03~2012-11-28 | 0 | 50 | 209999 |
| 2012-11-28~2016-07-09 | 1 | 25 | 419999 |
| 2016-07-09~2020-05-11 | 2 | 12.5 | 629999 |
| 2020-05-11~2023-06-12 (Current) | 3 | 6.25 | 794111 |
| 2023-06-12~2024-4-26(Prediction) | 3 | 3.25 | 839999 |
The security spending and effective fee rate of the Bitcoin network are shown in the following figure:
Bitcoin's security spending has increased from around $40 million before the first halving to $1.8 billion before the second halving, to $16.2 billion before the third halving, and to $33.3 billion so far. The security spending has been rising year by year, but the rate of increase has gradually decreased. The average effective fee rate within the corresponding halving cycle has also gradually decreased from 0.09067% initially to 0.0386%.
A further analysis shows that during the third halving period (2020-05-11 to 2023-06-12), 93.11% of the 164,113 blocks generated had an average effective fee rate of less than 1%, 69.44% of the blocks were less than 0.2%, 63.99% of the blocks were less than 0.16%, 45.99% of the blocks were less than 0.08%, and 31.44% of the blocks were less than 0.04%.
The Effective Fee Rate Is Not High#
Historical data analysis shows that in the long run, Bitcoin's security spending has been gradually increasing, accompanied by a substantial long-term rise in the price of Bitcoin. The change in the effective fee rate, which measures the actual Bitcoin network fee rate, shows it was 0.0386-0.118% in the last two cycles.
As traditional banks) that provide similar functions, cross-border transfer fees are usually between 0.1% and 3% (although traditional banks will have fee caps).
Further more, the Bitcoin network also provides features that banks do not have, such as permissionless, censorship resistance, decentralization, and unlimited transfer amounts, which are very crucial for large transfers.
Since 2016, single transactions averaged 6-9 BTC within halving cycles. Valued in US dollars, that’s $298,288.9 —indicating large transfers are common on Bitcoin.
This also confirms Lyn Alden's view: Bitcoin's base layer is Tank-like payments, not for daily use.
Kind of like how a tank is designed to get from point A to point B through resistance, but is not well-suited for commuting to work everyday, the base layer of the bitcoin network is designed to make global payments through resistance, but is not well-suited for buying coffee on the way to work.
Imagine, is $120 to $300 really a lot for a $300,000 cross-border transfer?
Predicting Bitcoin's Future Security in the Most Conservative Way#
By introducing the effective fee rate, the first question raised at the beginning of the article has been answered: Historically, what was the fee rate of Bitcoin? The answer is that although the fee rate was not very cheap, it was not high.
Part of the second question has also been answered: As the issuance decreases, will Bitcoin's fee rate soar to the point of driving users away?
Data shows that even if there was no Bitcoin issuance as a block subsidy, the fee rate would not skyrocket, because the effective fee rate already considers block subsidies as costs that users should bear. However, to fully answer this question, we need to analyze how the actual fee rate paid by users will change in the future after the issuance amount decreases.
Conservative Assumptions about the Future #
To reiterate the pessimistic view: If the Bitcoin price does not double every four years and the US dollar-denominated transaction volume remains unchanged, in order to maintain security spending, the fee rate will soar to an extreme level → driving users away from the Bitcoin network → transaction volume in US dollar terms decreases → fees decrease → network security spending decreases → miner income decreases, leaving the network → Bitcoin security decreases → users further leave the Bitcoin network → vicious cycle.
Predicting the future is difficult. To improve the reliability of predictions, I will set the conditions as conservatively as possible. Since pessimists believe that if current security spending is to be maintained, Bitcoin's fee rate will soar, so I assume that until the block subsidy drops to 0:
- Bitcoin network security spending and transaction volume remain unchanged, based on the third halving cycle.
- The Bitcoin price remains unchanged at the average price of the third cycle.
Analyze how the actual fee rate (fee rate excluding block subsidies) of the Bitcoin network will change based on this.
Result and Conclusion#
There are about 319 days left until the third halving (calculated based on the initial draft date of this article on June 12, 2023). Therefore, it is impossible to know the specific transaction data of this cycle accurately.
However, the halving cycle has been 80% (319/(365 x 4)=0.8) completed. Assuming the growth rate is the same as before, when estimating the data of this cycle, it is considered that the final data of this cycle has increased by about 25% (0.8 + 0.8 x 0.25 = 1) on the previous basis.
| 20200511-20230612 | 20230612-2024426(Predict) | |
|---|---|---|
| Total Txn Volume(BTC) | 2803368810 | 3504211013 |
| Total Txn Volume(USD) | 9.61121E+13 | 1.2014E+14 |
| Total Txn Counts | 322103525 | 402629406.3 |
| Average Txn Volume(BTC) | 8.703316148 | 8.703316148 |
| Average Txn Volume(USD) | 298388.884 | 298388.884 |
| Total Fee(BTC) | 56893.25863 | 71116.57329 |
| Total Fee(USD) | 1666881364 | 2083601705 |
| Total Subsidy(BTC) | 1025706.25 | 1282132.813 |
| Total Subsidy(USD) | 31687430757 | 39609288446 |
| Total Security Cost(BTC) | 1082599.509 | 1353249.386 |
| Total Security Cost(USD) | 33354312121 | 41692890151 |
| Average BTC Price(USD) | 30893.28037 | 30893.28037 |
| Average Real Fee Rate(%) | 0.00202946 | 0.00202946 |
| Average Efficient Fee Rate(%) | 0.038617805 | 0.038617805 |
Considering the bear market in the past two years, data such as transaction amount, transactions counts, and fees will be slightly lower than the average. At present, the market sentiment is gradually recovering. Using previous data to predict future data has a high probability of underestimation. For example, assuming that the future price of Bitcoin will remain unchanged at $30,893 forever, this is in line with my original intention to make conservative assumptions.
The Bitcoin network will stop issuing new bitcoins around 2140, with 29 more cycles in between. Simulated data shows that if the US dollar-denominated security spending of the Bitcoin network, the total transaction counts, and the average price of Bitcoin remain unchanged after 2024, the actual fee rate paid by users will experience a 28-year increase, rising from an initial 0.002% to 0.0382% in 2052, and then gradually approaching the effective fee rate of 0.0386% in 2024.
By my calculations, though the actual fee rate rose 19-fold, the final rate was under 0.04%. So pessimists aren't convincing: even conservatively, Bitcoin fees remain affordable.
For those with large transfers, censorship resistance, decentralization and security needs or users who cannot access conventional payment services, this fee is negligible.
People living in politically and economically stable countries may not feel the role of Bitcoin. Here are some examples:
- After the Nazis controlled Europe, Jews fleeing were unable to keep their wealth
- When the Soviet Union disintegrated, people were only able to carry belongings worth $100
- In countries like Venezuela, Syria, Iran, Nigeria, eastern Ukraine and China, people cannot take all their wealth because of foreign exchange controls
- Putin banned the bank accounts of domestic opposition
- The income of Afghan women will be confiscated by male relatives because their bank accounts are not in their own hands
- The US Treasury sanctioned the Tornado Cash, prohibiting all US individuals and entities from interacting with Tornado Cash or any Ethereum wallet addresses associated with the protocol
- And so on
So now we can finally answer the question raised at the beginning of the article: Historically, Bitcoin's effective fee rate has not been high. Under the premise that security spending remains unchanged, as issuance decreases, the actual fee rate paid by users is less than 0.04% and not too outrageous.
As long as security spending can be guaranteed, in the long run, the Bitcoin network will most likely remain secure.
Looking at Things with an Open Mind#
Finally, I want to talk about the most important mindset in technology and investment: an open mind.
Satoshi Nakamoto wanted Bitcoin to become electronic cash for everyone, and Vitalik Buterin wanted to build Ethereum into a global computer, but they failed to achieve their original goals due to capacity limitations.
Things evolve and progress. As long as the essence does not change, it is reasonable and right to make adjustments in other aspects to adapt to reality.
The essence of Bitcoin is a decentralized network for transferring value, and electronic cash is just one application built upon it. This world lacks decentralized networks, not electronic cash. This is reflected in the white paper: bypassing centralized financial institutions.
In order to achieve maximum decentralization, the Bitcoin community did not choose on-chain scaling to minimize the threshold for running full nodes. After the hard fork, the original chain gained more recognition, reflecting that people value decentralization more in terms of computing power, market value, and ecosystem development.
Moreover, open protocols can embrace progress more easily. Developers and businesses are building new applications and infrastructure on the Bitcoin network. What attracts them? Certainly not the performance of the Bitcoin network, after all, it is Turing incomplete. It is the brand and network effects of Bitcoin. Ultimately, these are guaranteed by the decentralized attributes.
Network security depends on the market. Though the earlier analysis was based on very conservative assumptions, I’m more optimistic about Bitcoin. I believe its average future price will exceed the $30,893 used in the hypothesis. The Bitcoin ecosystem will continue growing, and infrastructure like scalability solutions will gradually improve.
A thriving ecosystem signals Bitcoin’s wide adoption and the strong demand for it. Miners pursue profits. In such a lucrative market, they won’t leave. They’ll compete and innovate to make the Bitcoin network safer, greener, and healthier.