Bitcoin's Growth In Numbers
A decade's worth of on-chain figures for Bitcoin
Bitcoin’s rising price has often been both a boon and a bane. A boon because it incentivises individuals to connect to the network. A bane because it inevitably forces individuals to look at a single metric when it is far more than “number go up, number go down”. It reduces a complex financial ecosystem to a bubble economy where speculation looks like the most significant use-case. This post isn’t an argument for a better comparison of digital assets against traditional ones based on the growth stage of the asset. It merely looks at the numbers we ignore when we discuss Bitcoin. When Satoshi Nakamoto wrote in 2010 that Wikileaks had kicked the hornet’s nest when they integrated Bitcoin, he may not have foreseen the extent to which it was true. Over the past decade, Bitcoin has enabled inflation hit economies to find an alternative, those in authoritarian regimes to move their wealth and experiment bountifully within fintech.
Long before DeFi was a thing, loans, remittance and trading instruments were built on Bitcoin. Satoshi’s creation has taught (me) over the past decade that code can inspire a revolution. That ideas can offer individuals an alternative where there were none. Even when the entity you are up against is a mighty superpower with military bases around the world and an arsenal of nuclear bombs that could make the gods tremble. When we consistently and repeatedly focus only on price, we ignore the paradigm that existed before Bitcoin. We also ignore the fact that at an age and time where cash is increasingly going digital and authoritarian regimes use increasingly sophisticated machinery to find, separate and target dissidents - Bitcoin offers an alternative. Whether they use it as a store of value or a payments mechanism is up to them, but to me - that is what Satoshi’s work stands for. But more on that in my parting notes.
This post is a look at how Bitcoin has evolved over the past decade in ten different charts. For the sake of brevity, I have broken down this post into three parts
Address Growth - A look at the number of addresses added and the distribution of wealth between them
On-Chain Activity - Observations of the number of transactions and the amount of money transferred over the years
Mining Ecosystem - Numbers on how compute power committed to Bitcoin has surged over time, the miners taking the lead and the rewards they have gained.
Address Growth
Average Active Recipients have spiked 40,676% over the decade
I took average recipients on a monthly scale as a proxy for the number of individuals engaging with the network. As of December 2019, there are ~516,000 unique addresses that engage with the Bitcoin blockchain on an average day as a recipient. This figure peaked on the 14th of December 2017 at around ~1.07 million addresses. However, this does not capture the growth of the network in itself. To observe that, we will have to look at the number of new addresses on Bitcoin’s blockchain.
New Addresses Averaged At 124 Million Since 2017
New addresses are a measure for expansion of the ecosystem. Throw away accounts created contribute to the margin for error here. The average day in 2019 witnessed ~358,000 addresses newly added to the Bitcoin blockchain in comparison to 7100 in 2011. Another interesting point of scale I observed from these numbers was that of the 1 million unique addresses that engaged with the Bitcoin blockchain on 14th of December 2019, over 800,000 addresses were new ones. Proof of the fact that new, retail entrants were likely peaking at the time. One cannot come to this conclusion without studying how wallet holding patterns themselves emerged around the time.
Holding patterns show Resilient Hodling
On studying the average number of addresses that hold more than 0, 1, 10 up to 10,000 bitcoins it becomes quite clear that hold culture is quite serious within the ecosystem. This metric looks solely at the number of addresses and therefore could mean that individuals move tokens from one addresses to another or break them down to multiple wallets. However, the number remains shows a steady growth of 100% if we look at addresses count for those holding more than 10,000 Bitcoin. Dust accounts indeed peaked on December 17th with over 28 million accounts holding more than 0 bitcoins. This figure dropped to 21 million by April 1st - making one wonder if this were multiple addresses owned by the same individual (March 31st is the end of the financial year in many parts of the world). As of December 5th, over 28 million addresses held more than 0 Bitcoin. As of January 2011, only about 34 wallets handled more than 10,000 Bitcoins in them. By 2019 December, this figure lingered at around 130. Those with more than 10 Bitcoin surged from a mere 51,000 to 152,000 by December 2019. An indication that while early adopters and whales have increased in the count - it is still retail adoption that is driving demand for Bitcoin.
The most important take away I had from studying these numbers is that across all hodl demographics, regardless of price - the address count is at the ATH for each segment. For all the negative media about the space, fewer individuals are selling their coins, and more are testing waters with small sums. For a sense of scale, the number of people holding more than 0 bitcoin in Jan 2011 was a mere 70,0000. Today that number is north of 28 million — a 400x growth over the decade.
On-Chain Activity
487 Million Transactions - Confirmed...
Throughout its lifetime, Bitcoin has processed over 487 million transactions. A far cry from the mere ~2 million transactions it processed in 2011. The peak for this figure too... you guessed it right - came on the 14th of December with 490k+ transactions occurring on a single day. Judging by the frequency of transactions with price activity in the backdrop, it becomes fairly evident that a larger number of transactions are a result of individual traders moving coins between addresses. The average day in 2019 saw Bitcoin processing over 300,000 transactions in comparison to a mere 5000 in 2011. While 2017 tops the charts for multiple metrics, in terms of a sum of transactions, 2019 leads with over 111 million transactions. Relative to the previous year, Bitcoin has had an uptick in every year since 2010. The only exception to this rule has been 2018 with a -21.7% drop. For a sense of scale, India's unified payment interface processes 500 million transactions in a single quarter (vs Bitcoin's ten years) to reach the same figure. Alilbaba's singles day sees 1.5 billion transactions in a single day. Difference is, one of these cannot be shut down because a handful of executives decided to do so.
To Have Over $7.5 Trillion In Value Moved
Speaking about transaction count without volume in the backdrop would make little sense. To understand this better, I separated Bitcoin volume in dollar amounts and number of bitcoins. The reasoning being that a rising price of Bitcoin could indicate larger volumes being put on display by the same early day whales pushing tokens around. Incidentally, on plotting the two, it became quite obvious (as expected) that price appreciation of the digital asset has changed how individuals move Bitcoin. Of the $7.5 trillion in total moved on Bitcoin's network since inception, ~870 billion transferred in a single month. December 2017. I don't know ... That month seems to be quite important at this point. (wonder why). What I also found interesting was that the sum of Bitcoins moved peaked in September 2012 (at 331 million) and is currently at 17 million as of November 2019. As price appreciates, individuals seem to be holding on to larger number of Bitcoins and allocating smaller amounts for trading.
Note : Each bitcoin moved is counted as one. Since the same coins can be moved back and forth multiple times, this figure will be fairly higher than the supply.
With $1 Billion Paid In Fees
Any mention of fee metric is not full without accounting for fees incurred. Bitcoin transactions have cumulatively cost $1 billion throughout its life. In Bitcoin terms, this number is at 205,000. The highest single-day fee burn was on the 22nd of December (2017) when over $22 million worth of Bitcoin was spent in transaction fees alone. One way to look at this data in light of what we have in terms of transactional volume (in USD) is that for every dollar spent in transactional fee, $75000 has moved on Bitcoin's network. Or roughly about 0.01%. I could buy that for efficiency. The challenge with this notion remains that the figure (0.01%) is dirtied by substantial amounts moved in single transactions. For micro-transactions, this may not be a relevant alternative as the fee has consistently been around $1 and has spiked up to $53 during 2017.
Mining Ecosystem
Hashrate has risen a 1000 times since Q4 2013.
It took 37 months for mining power dedicated to Bitcoin to hit its first TH/s (Jan 2009 to March 2011). Over the next 37 months, that figure would rise to 495,624 TH/s. As of December 2019 mining power committed to Bitcoin was at an average of 95 million TH/s. Part of the reason being the fact that mining has gone from the turf of the hobbyist to those of large-scale, data centre based mining operations. Factors varying from the availability of short term credit to government subsidies and electricity determine winners in the mining industry today.
For a very low-end calculation of how much money is invested in mining - it may help to explore how much money has gone into mining itself. The best I could find in terms of TH/s was Bitfury's Tardis, priced at $3770 for 80 Mh/s. Discounting costs in labour, electricity, logistics, data-centre leasing and other elements, this comes to around $47.125 for each Th/s committed to Bitcoin's network. This number is likely much higher as the Bitfury Tardis released only in late 2018 and earlier miners likely have worse off unit economics. At around 95 million TH/s - it should come around $47.125 * 95 million = $4.475 billion on the low end of the spectrum.
This number can go higher or lower depending on
(i) subsidies
(ii) volume purchases offering discounts and
(iii) older models having a better TH/s.
The point is - at least $5 billion has been invested on the low end of the spectrum into Bitcoin's infrastructure alone. As a percentage of market-cap that is 3.65% of the market-cap. This investment, while seemingly "inefficient" in traditional contexts where databases are switched on and off at the whims and fancies of a sysadmin, is the price this ecosystem has paid for decentralisation.
Note : Moore's law dictates that over time we will be paying lower $ per TH/s. I have gone with the best pricing I could find for TH/s for a simple calculation. It will be apt to look at the best TH/s per $ costing for each year depending on the models available at the time, adding up for an estimate on how long it may have been functional and mapping rent and labour costs in each part of the world. That is an academic exercise I did not want to explore until I head back to university. But if you do it, please e-mail me at joel@decentralised.co
Partly Thanks To Mining Pools Facilitating Growth
This "decentralisation" however comes with its own risk of cartels and unification. In 2014, a mining pool came dangerously close to being able to pull off a 51% attack when it attracted ~42% of mining power committed to Bitcoin. At the time, a "bitcoin developer" named Vitalik Buterin suggested setting up a bounty pool to incentivise the development of peer to peer mining pools. That was 5 years back and one could argue that rising financial incentives keep stakeholders aligned in terms of not attempting something similar again. Since Jan 2009, mining pools have given out $16 billion in rewards to miners committed to the network. What I found interesting was that the launch of Btc.com (2018) and Huobi Pool (2019) found quick traction and integrated rather quickly into the mining ecosystem. Hardware manufacturers such as Bitfury too find it easy to go to market with their pools over time. The percentage of "unknown" pools within the network has reduced over time. For more on mining pool centralisation, I strongly suggest reading Jimmy Song's post on the matter.
Side note : Prayank from IndiaBits was kind enough to link me to this post by Greg Maxwell that looks at the 51% attack angle in-depth. Worth reading for context
Who Cumulatively Paid $16 Billion To Miners
Including transaction fees - miners have been paid out 18,318,525 bitcoin or over $16 billion. Given that in December 2019 there were 18,076,100 bitcoins in circulation, it is safe to suggest that some ~242,525.09 Bitcoins were paid in fees over the past decade. I assume that as mining difficulty increases and halving inches closer - mining will become an increasingly institutionalized play. We will see newer unicorns coming from the facilitation of mining (retail side), development of ASICs, manufacturing and supply chain (e.g.: what Xiaomi did to mobiles) and infrastructure service providers that focus on greener sources of energy. We may also witness government subsidies in real estate and energy being on the rise as Bitcoin's ability to kick-start regional economies through mining enabled interest in the region becomes more evident.
In case you are wondering how those fees were distributed - here's the chart for mining fee distribution in 2019 (excluding fees paid to unknown pools)
What does all of this mean (to me?)
I have always seen Bitcoin as the most significant financial experiment the world has ever known. It was built on 40 years of improvements in cryptography and multiple decade's worths of experimenting with digital currency. When many of look at the ecosystem today - they see loud crowds on twitter, speculative manias, and what may be a chaotic culture. What they ignore is the fact these are by design because Bitcoin was not to compete against existing paradigms in Fintech. If that were the case, Satoshi might have been pitching to VCs in Silicon Valley instead of writing to the cypherpunks mailing list. It is a requirement if Bitcoin is to be a real alternative. The moment we have a "face" or a "headquarter" - these systems could be taken down. Bitcoin is more than the arguments we see on twitter or the angle the media pitches things at. It is change embodied in code.
Writing from a country where currency was demonetised in a matter of hours, I see why alternatives are needed. As with all technology, regulators need to realise that the genie is out of the bag. They can push it to the fringes, malign its name and very well criminalise its ownership. But they cannot tame it (yet). We see it when individuals rush to Bitcoin during CNY devaluation, an inflation crisis in Zimbabwe, a debt crisis in Greece or the demonetisation in India. The point is, people are waking up to the realisation that an alternative exists. As Nic Carter once wrote - Bitcoin is the most peaceful revolution in the history of humanity. That to me, is the allure of Bitcoin. These metrics simply show how significantly the ecosystem around it has grown. Whether these trends hold over the next decade, or price does go to $100k is not a concern (for me). So long as transactions are confirmed - Bitcoin has solved for what it was built for in its own right.
On a parting note, this is what Bitcoin's report card for the decade would look like
Had to go with November since data for December was only partially available when I began work on this piece.
Notes
1. All data mentioned here was sourced from TokenAnalyst.io. I don't get paid to say this but if you are an analyst exploring the space - make sure you ping their team because they have the best data streams in the industry. Also, they are nice guys. Feel free to ping me for any references on them.
2. I cannot be handing over the data in this piece to you since it is a paid product and I did not curate it personally. I also have a job and responding to data queries takes precious time from all the other things I enjoy doing. Like cooking and reading books. Exceptions to this rule are - regulators, academia (professors) and contributors to the ecosystem. I am most definitely not going to export this data for college assignments (don't try).
3. Don't see this as a "buy" recommendation for Bitcoin. I don't know where this thing is headed in terms of price. The objective of this post was to look at the metrics we routinely ignore and track how they have evolved. I did not compare this with traditional fintech platforms because that would be akin to comparing submarines to boats. Same same, but different. As with everything else in life, make an educated choice about ownership of the asset.
4. If you translate this piece - please drop me a link and attribute it to the website. If you are in university - e-mail me from your academic address with a link to the translated work for free coffee.