Bitcoin datalogy (part II): Geology of the lost coins
The first part of the article here.
There are lots of stories about people who lost bitcoins bags, especially at the dawn of the era of cryptocurrencies, when bitcoin almost worthless, and they were forgotten in the old hard drives, flash drives and even on scraps of paper.
Is it possible to calculate how many bitcoins are lost? Blackany after all, constantly monitor your data, and, as we have demonstrated in our previous material, it is possible, for example, to visualize the UTXO-age bitcoins to trace the history of change of their respective owners.
Colored bars show the relative share of existing bitcoins, which participated in the transactions during selected time period. The lower, warmer colors (red, orange) — this is bitcoin, often participating in transactions. Accordingly, more than cool colors (green, blue) — this is bitcoin, a long time did not participate in the transaction. The number of bitcoins in circulation increased for all time with 50 BTC to ~17 million, this figure is reflected on the left vertical axis. Black curve — the exchange rate of bitcoin against the dollar (logarithmic graph, right vertical axis). Chart made by Nelson morrow based on the work of @jratcliff
Many readers of the first part, after reviewing the above chart noted that “probably a lot of old coins lost”. It is a logical assumption. Holders of bitcoin in the last year and had a lot of reasons for the transaction: price rally and the subsequent decline, the flourishing ICO, fork BTC/BCH, new addresses, Segregated Witness, etc. But coins are not used for five or more years are those who most likely lost forever. But can we confirm this guess?
Despite the abundance of data in the blockchain, it is very difficult to estimate what amount of cryptocurrency really is lost, because lost coins are not leave traces in the blockchain. They still hang in UTXO their last transaction, and gradually grow old. The main problem is that a lot of lost bitcoins in the blockchain look exactly the same.
But the distribution of age UTXO can tell you exactly how to measure the loss of BTC. Painted in the cold colors of the strip can be compared with the low-pass filters, which cut off most of the old coins. That is, in this region the changes are not so frequent as in the warm “young” bands.
Age strip UTXO similar to geological formations: evidence for long-term storage of coins in the past lurks under layers of recent transactions. In order to distinguish lost bitcoins from carefully stored, you need to get to the hidden data of the oldest layers, from the depths of the records of the blockchain.
The study of lost bitcoins is the Geology under the guise of analysts.
We believe that the loss of bitcoins took place over two different “cryptograhically” periods.
1. System loss: Satoshi and other miners-explorers mined and lost a lot of bitcoins in the beginning era of cryptocurrency. (Carboniferous period bitcoin)
2. Incremental losses Occurring in different time losses ordinary bitcoin users.
Later in the article we show that the era of system of losses has ended, and will demonstrate that we now live in a period of incremental losses. Finally, we will establish the approximate amounts of lost bitcoins.
Rannie system loss
What was happening in the world of Bitcoin in 2009-m year, when he appeared? The answer is: almost nothing. After several years of work on the concept and code of Satoshi has published a “white paper” in October of 2008. January 3, 2009 mystical Japanese nominal the base block, and 9-th day posted in open access bitcoind software (v. 0.1).
In those days very few people had taken seriously Satoshi and bitcoin. In a wonderful article Guarna Branwen “bitcoin worse is better” are examples of negative reviews of “professional” cryptographers of the time.
The early days Satoshi basically mainil alone. Only from time to time he was joined by other nuts like Hal Finney.
As seen in the graph below, the hash rate was below the limit. Satoshi and the first miners could not reach can see, necessary to activate the algorithm complexity of the network, until the beginning of 2010. The average time between blocks has reached zaplanirovany 10 minutes in February 2010.
The diagram shows hash rate and average time between blocks 2009 and first quarter 2010, inclusive.
Probably throughout 2009, the bitcoin mainily only Satoshi and a few small groups. For the first time this diagram was published in an article by Evan Klitzke.
Despite the apparent stagnation, blocks in 2009 namimili considerably, and in 2011 produced more than 5 million of BTC. More than 23% of bitcoins that will ever exist. Where did they go?
This graph shows the current UTXO zootechnie-age and average balances for BTC UTXO each age group.
The area with UTXO-above the age of 7 years (about 1.9 million BTC), kaminanda until 2011, clearly visible in the right side of the graph, and the average balance is 50 BTC (reward for nameany unit at the time).
[Video link]the Chart above shows that 1.9 million BTC UTXO with the oldest-age is a separate group. Coins, nominingue Satoshi and his followers at the dawn of Bitcoin. They create a “peak” at 50 BTC in the chart, because this was the reward to the miner for the unit (and the fee then was negligible): this is created by assets that have not been spent. (Note the sharp decline in this chart, the incident of 6.5–7 years ago in 2011. We to this date still will return). Carboniferous period bitcoin
Carbon ruled the Earth for 300 million years ago. At that time the planet was already covered with trees, but there was never anyone who could eat. The result is dead trees toppled one on another, and do not decompose, creating layer by layer.
The period 2009-2011 was the Carboniferous period of bitcoin: many kamineni coin was never used by anyone, accumulated in the blockchain, and are gradually forgotten and lost.
It’s funny that millions of years later, the Carboniferous trees turned into charcoal, the main source of energy used for bitcoin mining.
The transition to incremental losses
Something sensational happened with Bitcoin in 2011. The number of the oldest UTXO-age coins about which we spoke above, was sharply reduced. Five years later, in the lane where bitcoin reaches the age more than 5 years, we see a corresponding decrease in level. Here it becomes obvious turning point or twist in the allocation of the band with UTXO-age more than five years in 2016. In this you can see an echo of the dramatic changes which have happened with Bitcoin five years earlier, in 2011.
This graph shows the net rate of change of the number of bitcoins over 5 years in the dynamics (the period of 90 days).
Between 2014 and 2016 (the section in blue) we see the effects of the “Carboniferous period” of Bitcoin, which lasted from 2009 to 2011, when it lost a lot of bitcoins. This period abruptly ends in 2016, echoing the dramatic changes that happened with Bitcoin five years earlier, in 2011. The timeline of the schedule begins in 2014, as this is the first year when bitcoin crossed the milestone of five years of age (base block, recall, was namine in 2009).
[Video link] the Chart above illustrates this transition. Between 2014 and 2016 (painted blue) was a huge amount of bitcoins over the age of 5 years. It meets the demand of those coins of the last transaction which happened in the Carboniferous period bitcoin (2009-2011), when the first miners and Satoshi got and then lost a fair amount now.
Sharp decline in the number of bitcoins over 5 years of age occurred in 2016, symmetrically the end of the Carboniferous period bitcoins in 2011. The geological Carboniferous period ended with the emergence of bacteria able to digest the wood that stopped the endless process of accumulation of dead trees. What led to the end carbon of bitcoin?
From entertainment to profit
In June of 2011 Bitcoin experienced a first impressive price rally. In just a few months the price jumped from less than $1 to $33. Many of the early miners earned it – at least those who did not manage to lose my keys.
Before the sudden jump in price to $33 miners were rather negligent in matters of security of storage of mined bitcoins. This is the time of the first tragic stories about lost drives competitive athletes with incredible digital wealth. As a result, all miners learned two lessons: 1) BTC can be profitable, 2) the price of bitcoin could increase by several orders of magnitude.
These lessons have completely changed the attitude towards mining and the security of bitcoins. At a price of <$1/BTC, total revenues from mining amounted to several thousand dollars per day, reaching probably up to ~ $ 1 million per year — this market is barely enough to support one medium-sized business. And at $33/BTC daily profit from mining reached almost 250 thousand dollars, which gave an annual revenue > $ 80 million.
From trinkets bitcoin has turned into a real product, and mining has evolved from a hobby into an industry.
The echo of the great wave HODL?
A special feature of the above diagram can be considered a sharp jump in the number of bitcoins that crosses the five-year mark, in the last few months. This jump is bitcoins the last time you participated in the transaction in mid-2013, year, 5 years ago.
In the last article we determined the changes in the wave HODL on different categories UTXO-ages.
The recent rise of the price of BTC crossing the mark of five years of age, is the first great rolling waves HODL, beginning in 2013/2014, when bitcoin exploded to the $1000/BTC.
We know that HODL the Great wave broke in 2017 at the peak of $19 000/BTC, so it is unlikely that all the 1.5 million BTC UTXO that in 2016, the year was 3-5 years, lived to the age of >5 years in 2018; many were transferred during the rally, forks, SegWit, &c.
Thus, we tend to think that unlike the Carboniferous period bitcoins, the new growth will be much less and over much earlier. Presumably <500 thousand BTC in the next 18 months will cross the mark of >5 years.
So how many bitcoins are lost?
It is impossible to ascertain. But, rebounding from the above analysis, we can more or less accurately try to guess the approximate framework. It will be easier to do it with the full version of the UTXO distribution of ages.
Full version charts of the distribution of UTXO-ages without an indication of the effects of the affordable BTC. Two previous halving readily apparent as deviations from the normal number of coins produced (as we said, in 2009, the hash rate was unstable, which led to the deviation from the linear dependence). Because of the huge difference in the amount of BTC between 2009 and 2018 the earliest wave HODL difficult to see, but the Great wave HODL 2013/2014 is still clearly visible.
[Video link] a Lower bound estimate of lost bitcoins we will consider cryptocurrency produced from 2009 to 2011 Satoshi and the first miners, not dissipated to this day: 1.9 million BTC. It’s almost 2/3 of the BTC in the current wave of bitcoin age >5 years. Certainly a lot of bitcoins were lost in subsequent years, after 2011, but how count them?
During the rally 2017, the year the band 3-5 years of age is declining dramatically, but the strip of bitcoins over 5 years is almost unchanged. This indicates that many coins at the age of 3-5 years are still under someone’s control, and almost all the bitcoins than five years – lost. Somewhere between the ages of 3 and 5 years we should expect a blend of the previous model to a modern model. This also implies a less conservative lower limit of 3 million lost bitcoins – the whole band coverage over 5 years.
The distribution UTXO-ages “approach” to the band age of 3-5 years, divided into three-month interval covering rally 2017.
[Video link] This “enhanced” version of the distribution of UTXO-ages shows more subtle structural data in the band of the bitcoin age from 3 to 5 years, by breaking them down into three-month segments. The wave HODL noticeable as groups of coins crossing time intervals of 4 months.
The earlier age band (36-39 months) obviously thinner today than it was a year ago, it means that almost all of these coins are under the control of someone who worked with the transaction during a rally in 2017. Older age segments (57 – 60 months) during the rally 2017, almost nothing happens, as in the band of bitcoins over 5 years.
We believe that the boundary between the different behavior lies between 45 and 51 months. If we assume that most of the coins are older than this age lost, it turns out that the top bar level permanently dropped out of circulation of bitcoins will be 3.8 million BTC. Total estimate of the distribution of UTXO-ages, we believe that lost from 3 to 3.8 million BTC.
A more accurate assessment
It’s hard to be accurate when estimating losses using only the UTXO-age. Much more effective would be to mark and track individual UTXO with external metadata to determine the context of transactions: mining, exchanges, etc. This approach is more like archaeology than Geology.
Fortunately, our friends from Chainalysis already conducted research on this topic. In a published last year in the Forbes article they announced your answer: lost from of 2.78 to 3.79 million BTC. We are pleased that our simple approach based on the study of the UTXO-ages, gave results that are consistent with the results of more complex approach Chainalysis.
The first part of the article here.
The original text in English here.