A look at the Bitcoin network transaction history

xkcd bitcoin

I have been reading an eye-opening paper by mathematicians Dorit Ron and Ami Shamir, looking at the entire transaction history of the Bitcoin network. For those unfamiliar with Bitcoin, it is a cryptographic currency that was created with the stated intention of generating a decentralised payment system for the Internet era (my earlier take on the idea here). It has proved popular with computer enthusiasts, and specially with libertarians who see it as a perfect antidote against fiat monetary systems. So far Bitcoin has been riddled with security issues, theft, fraud, hacking and other problems that have meant that the scheme has failed to become popular outside of a core following of rabidly faithful enthusiasts.

The paper contains some truly insightful figures that help to assess the Bitcoin experiment fully. There are 1,851,544 different Bitcoin owners. Given that at the time of writing there are a total of 10.1 million BTCs in circulation, this would give an average of 5.5 BTCs per each user, which does not seem to encourage an active economy, and certainly would not justify the current price of $12 USD per BTC.

But some of the real interesting information comes from the data about centrality. From the data collected, it seems clear that there is some large accumulation of the bulk of Bitcoin activity, for example, one single user had 156,722 different addresses. This was identified as Mt.Gox, the largest Bitcoin exchange, which is responsible for an astounding 90% of all of the total BTC transactions recorded. This indicates that BTC is not being used as a payment system, but as a commodity where users exchange bitcoins for cash and vice-versa. This is further evidenced by the fact that the actual number of BTCs in circulation is considerably smaller than it was previously thought, with an almost 78% of the entire BTC reserve (7,019,100 BTC)  being placed in “saving” addresses. So, of all of the Bitcoins ever created, there are only 22% in circulation.

To my mind, this seems to confirm the suspicion that many of us had when Bitcoin started getting popular, and it is that the system encourages hoarding and accumulation, which make it uniquely unsuitable as a currency. This is corroborated when one considers that a large number of transactions seem to consist mostly of operations between the same owner, where the coins are moved from one address to the other.

The most revealing statistic is the astounding amount of wealth accumulation in the BTC network, something that had also been guessed at by critics, but that is confirmed by the data. The study comments:

“Another interesting finding is that the total number of Bitcoins received by most owners is negligible. As can be seen from Table 2, 36% of all owners received fewer than one BTC (currently worth about 12 USD) each throughout their lifetime, 52% received fewer than 10 BTC’s and 88% fewer than 100. At the other end of the distribution there are only four owners who received over 800,000 BTC’s and 80 owners who received over 400,000.”

Furthermore, currently 97% of all owners have an amount of less than 10 BTCs, and during the entire history of the scheme, 88% of all users have never owned more than 10 BTCs. This again hints at a vast wealth of coins being hoarded by a few people, probably developers and early adopters. This is further evidenced when reading the list of BTC owners, where there is a single unidentified user with 2,886,650 coins, or more than a quarter of all BTCs issued so far. The second largest owner is unsurprisingly Mt.Gox.

The paper ends with a discussion of the largest Bitcoin transactions, and it seems like there has been a concerted effort by one party to try to hide wealth accumulation by breaking up a large chunk of Bitcoins and moving it around in thousands of smaller transactions. The authors are very factual about their findings, but to me it seems clear that there is something not right going on.

I encourage people interested in Bitcoin to read the paper. To me it is clear evidence that the current version of a decentralised payment system is not working as intended. Unless you’re a conspiracy theory and believe that Bitcoin is an attempt to get rich quick, in which case it might well be working for some of the minority which hold most Bitcoins.

Interestingly, I think I might still have 0.10 BTC lying around in a wallet somewhere.

Edited to add

Welcome Redditors. There are some comments that claim that I misread the paper, and that there is no evidence of hoarding. I have read it again, and it is clear that there is, at least to me. There are two types of wallets studied, let’s call them savings and current. There are owners that can be identified as managing current accounts, such as exchanges. But the large majority of wallets are savings, namely, they have incoming but not outgoing transactions, which means that they are not being used as currency, but for hoarding. There is one user who controls a large number of such accounts, which makes it different to “current” users such as Mt.Gox. We may argue about the extent of hoarding, it might not be a single 25% user, but this seems consistent with what is otherwise seen in the Bitcoin economy.

Adding decimal points does not fix the problem of inherent hoarding built into Bitcoins. I, for one, will not be participating. You may not care that a random blogger is not willing to participate. However, I believe that I am representative of the type of user that needs to be won over for BTC to take off as a real currency: someone just off the mainstream with a little bit of technical knowledge who likes the idea of decentralised systems. If you cannot convince the likes of me, Bitcoin will remain the domain of libertarian geeks.

15 thoughts on “A look at the Bitcoin network transaction history

  1. That study is unfortunately misguided. It confuses "owners" with "addresses", assuming the later equals the former. Saying "97% of all owners have an amount less than 10 BTCs" is flat wrong. The correct answer is that 97% of all addresses have an ammount less than 10 BTC's, but nobody knows who owns which addresses and indeed many addresses belong to one person. Also, many addresses are purposefully used once and never again, so you'd expect over time that most addresses will have zero or marginal amounts of coins.

    Further, even if 3/4ths of the coins are being saved (or "hoarded" if you want to be derogatory about it), why is that a bad thing? What portion of USD is "saved" in bank accounts and bonds? What portion of gold is saved in vaults? Antagonism toward wealth being saved betrays a misunderstanding of how capital is supposed to accumulate, mistakenly believing the only way an economy improves is via spending and consumption.

    • Hi,

      That is an important point, and probably one that will be changed in later versions of the paper. Mt.Gox is a good example of the owner of thousands of addresses that also may not be the owner of some of the coins these addresses contain, as they are probably for transaction pruposes. Still, I was struck by the fact that even though it is clearly the largest exchange out there, it is still only the second possesor of coins. My guess is that owner A in the paper is Satoshi, or an early adopter.

      The question of hoarding is a very important one. The role of a currency is to circulate allowing owners to be able to exchange goods and services. An economy without liquidity is an economy in trouble. You can answer your own question with regards to the US monetary base in this site: http://research.stlouisfed.org

      Very broadly, what you're looking at is the money in circulation at any given time, which at the time of writing is 2,589.477
      Billion USD. The broad idea is that such amount in circulation is a good thing, and that the percentage of wealth stored in other media is not detrimental to the overall economy. As long as there is enough in circulation, the economy is healthy. The problem with BTC is that there is limited supply of money, and if large chunk of that supply is stored, then we are not dealing with a currency.

      For quite a while it is clear that Bitcoin is not a currency, it is at best a speculative vehicle. If that is the intention, then people should be upfront about it.

      • If you add up the 'Accumulated incoming BTCs' for the users labelled A to S in table 7 you get 14Million coins. There are only about 10Million Bitcoins currently in existence – so that should give you a clue that you can't use this table to assume the balance.

        As a further example – I have an address which blockexplorer reports as having received 1200 Bitcoins. That is absurd as it is in a wallet which has never had a balance at any one time of over 40 BTC, and has had a total of around 45BTC that was ever deposited into it. Measures of 'received' BTC seem to be highly inaccurate for wallets which have more than a handful of transactions.

        People need to take great care in drawing conclusions from this paper.

        • The discrepancy is caused by the accounts that are used for transaction purposes, namely Deepbit, Instawallet and Mt.Gox. Some of the accounts are clearly saving accounts given the number of transactions involved. In other cases, as in the November 2010 suspicious account, a number of operations have taken place in order to hide the hoarding.

          I'm not really that surprised about some of the findings, as they tell us things that long-time observers of Bitcoin already knew. Bitcoin is not behaving as a currency, and it is quite clear from the data that this is not the case. Right now it's some sort of virtual commodity or investment scheme.

        • Would it not be more correct to say that Bitcoin is not behaving as inflationary fiat currency that all rational participants convert into hard assets and other investments in order to avoid being ripped of by inflation? Where is the premise that proper currency that is not being inflated at a whim of private central banks must behave in the same way comes from? Logically it should not.

  2. I don't think you can infer that Bitcoin "encourages hoarding and accumulation" yet: it's still a very new currency.

    But it occurs to me that the existence of hoards is probably a *good* thing. The more expensive Bitcoin becomes, the greater a hoarder's incentive to spend their savings. As they release the stored Bitcoins back onto the market, the increase in supply will bring the cost of the currency back down, making it easier for vendors to continue to shift goods, and new users to buy in.

    So actually, I'd expect hoards of Bitcoins to act as liquidity *reserves*, which will be released *precisely* when the Bitcoin economy most needs it.

    • But the point is that there is no such thing as a Bitcoin economy. The movement we see is mostly from investment through the exchanges and intermediaries. One of the things that struck me is that apparently Silk Road transactions are not even showing up in the system, which is something I have suspected for a while. BTCs are not even being used to buy drugs, at least not in large amounts.

      • (But isn't exchanging, say, US Dollars for Bitcoins *the same action* as exchanging them for books or toiletries? It's all swapping stuff, isn't it? It's an economy, just one limited to very particular types of exchange… so far.)

        I agree that you can't buy much, apart from money and dope, with Bitcoin… yet. At the moment, it's much easier (as in, actually *possible*) to use "official" currencies to purchase pretty much everything. This would explain why currently Bitcoin seems to function somewhat as a poor man's stock market: buy the Bitcoins when they crash, sell them when they peak. Cash them out before buying anything with their stored value.

        But past performance isn't indicative of future performance…

        If Bitcoin stabilizes, more people will want to use it. Even if it doesn't stabilize, it may prove a handy means of avoiding tax on goods. If QE continues, people will become increasingly interested in currencies not subject to inflation. If…

        I think it's just possible that Bitcoin sceptics may find themselves in the same position as those who could never see smartphones, blogs, internet search, or indeed the internet itself, taking off. "Who the hell *needs* it?"

        • "I think it's just possible that Bitcoin sceptics may find themselves in
          the same position as those who could never see smartphones, blogs,
          internet search, or indeed the internet itself, taking off. "Who the
          hell *needs* it?""

          I really doubt it. I like the idea of a decentralised cryptographic currency, but Bitcoin is not the final product for many reasons. The deflationary nature of the scheme, coupled with the disproportionate advantage that early adopters have, mean that BTC will probably not take off as a currency. My guess is that it will continue to be traded as it is now for a while.

          I hope that another similar scheme happens in the future.

  3. The claim that one propensity to hoarding makes something unsuitable as currency is one of the most persistent economic fallacies, and that a lot of economists insist on it too makes it only worse.

    • The reason is that a currency is by definition designed to circulate (hence the word). An economy without liquidity is a stagnant one where nobody is buying or selling, so no goods are made, factories close, businesses close, etc. The confusion comes when people mistake currency with investment, such as bonds, saving accounts, stocks, and similar vehicles. Both are different things.

  4. I will never understand the obsession with, as you put it, "being ripped-off" by inflation. Inflation is practically inevitable in modern economic reality, and this is not problematic whatsoever when it is not runaway inflation followed and/or caused by catastrophic devaluation.

    Inflation serves the purpose of encouraging spending and investment. If you keep all your money in the bank for a long period of time, then you will inevitably lose because interest rates rarely match inflation, so the rational thing to do is to keep some money, and use the rest to invest in anything else that is not currency: stocks, bonds, commodities, property, etc.

    Needless to say, this is good for the overall economy, when people spend money, it creates jobs and keeps money in circulation, which is good. The Bitcoin hoarding deflationary model is not good as a currency, but it might work as an investment.

    I still think that the current bubble will burst when one of the large BTC owners described above cashes out. My guess is that MtGox does not have enough funds to give to people who remove the money.

    • Inflation is bad for the economy as it prevents savings and investment. A compulsive gambler never saves money, his life is a mess. The same with the economy. Wealth is created through investing – factories, r&d, farming, etc – not by putting granite counter tops in your house.

      Inlfation is a result of money printing by central banks. Fed reserve prints $85 billion every month to buy US debt. This is to the benefit of US government, not you.

        • I have a degree in econ from ucsb and I've heard these arguments before. It's mainstream but mainstream is clueless. Look where we are now. I cannot go into further detail or i will waste my years away bashing a keyboard arguing with you. Believe what you want.

Leave a Reply