On the Ownership Concentration of the Bitcoin Network

A key value proposition of cryptocurrencies lies in decentralisation via blockchain technology, which aims to replace the reliance on third parties, such as financial intermediaries and tech platforms, with a large set of network participants. In principle, the absence of a central authority should hamper any attempt of a given “shareholder” to single-handedly alter the growth trajectory of a given protocol and/or crypto asset. To put it differently, decentralisation rhymes with resilience, as far as crypto advocates are concerned. 

However, understanding the degree of actual decentralisation in cryptoasset markets is far from obvious. Perhaps not surprisingly, one key metric of decentralisation is the distribution of a given crypto asset across network participants, namely the number of network addresses. That is, the more the ownership is distributed throughout the network addresses, the lower the possibility of few large owners to manipulate the network itself. Leaving aside the differentiation between Proof-of-Work and Proof-of-Stake protocols, the problem of measuring ownership concentration based on the amount stored in network addresses is twofold: first, not all addresses should be treated equally. For instance, an exchange address holding funds from millions of customers cannot be considered in the same way as a wealthy individual who stores a large amount of assets in self-custody. These participants hold large inventories for different reasons. Second, and perhaps more importantly, an address does not represent an account. One user can control multiple addresses as well as one address can holds the funds of multiple investors (see the example of centralised exchanges above).

For this reason, understanding ownership concentration cannot simply rely on a snapshot consisting of the number of addresses at a given point in time. Two additional things should be considered: first, the evolution of network concentration over time based on wealth in both fiat and native currencies. The logic is that a constantly decreasing concentration could signal increasing decentralisation, in equilibrium. Second, investigating network concentration should be built on the premise that not all accounts are created equal; one needs to consider the distribution across entities of different sizes. For instance, taking into consideration addresses that belong to exchanges and miners as well.

In this article, I try to expand the notion of ownership concentration with a particular emphasis on the Bitcoin network. By looking at some of the existing data on the Bitcoin network as well as borrowing some of the insights in the existing academic literature. In this article I aim to shine some light on the true underlying distribution of Bitcoin across network participants and show that Bitcoin ownership is much less concentrated than often reported, although wealth is still highly skewed and far from being evenly distributed across network participants.

A bird’s-eye view on Bitcoin ownership 

Bitcoin addresses are essentially public identifiers recorded on the blockchain through which an individual or an entity can send and receive Bitcoin. Based on simple heuristic arguments, one can identify how much of the total supply is held by different addresses depending on their size, namely the amount of capital that is stored in a given address. For instance, one can track with good confidence the amount of aggregate Bitcoin supply that is stored by so-called Whales, namely those addresses or entities which hold a relatively large fraction of the total supply. The cut-off to be classified as Whale is not unique, and really depends on the data source and/or individual conventions. For the scope of this article, we define a Whale an individual address which holds more than 1% of the aggregate total supply available in a given time. The left panel of Figure 1 reports a simple breakdown of Bitcoin addresses by size as of May 24th, 2023. The right panel reports the same breakdown over time. Notice that the residual ownership of Bitcoin is clustered in “investors” (those who holds between 0.1% and 1% of the total supply) and “retails” (those who holds less than 0.1% of the total supply). 

Figure 1: A simple breakdown of BTC ownership

                        

Source: Aaro Capital Research
Notes: The left panel shows the relative number of addresses owned by small investors (<0.1% of the Bitcoin supply), medium investors (0.1%-1% of the Bitcoin supply) and whales (>1% of the Bitcoin supply) as of 24th May 2023. The right panel shows the same classification over time from December 2010 to May 2023. 

Perhaps not surprisingly, the vast majority of addresses can be classified as small investors or retail. As of May 2023, retail investors consist of almost 90% of the number of addresses in the Bitcoin network. This is specular to the very few large holders, or whales, which consist of 1.3% of the addresses as of May 2023. The composition based on size has been quite stable over time; in fact, the number of retail addresses relatively to the total went from 70% in 2012 to almost 90% as of May 2023. This suggests a widespread increase in the number of small investors holding BTC. As whales tend to own substantially more wealth compared to retail investors, Figure 1 provides some preliminary evidence on how skewed the Bitcoin ownership towards large investors is. Notably, the asset pricing implications of such ownership concentrations are quite interesting. Clearly, a significant imbalance in the wealth distribution could open the door to market manipulation and opportunistic behaviours, especially within the context of a relatively illiquid – compared to equity, bond and foreign exchange – asset class such as cryptoassets. Perhaps more importantly, an extreme property concentration could have significant implications for market quality and price discovery. This is ultimately relevant for both small and large investors.

Nevertheless, the relatively coarse aggregation in Figure 1 should be taken with a pinch of salt. After all, interpreting Bitcoin data in an oversimplified manner could result in wrong conclusions. In the next section, I aim to provide a more meaningful picture of ownership concentration by breaking down both the distribution of supply by size as well as the number of addresses by size with a more granular classification. 

Delving into the ownership distribution

Since the inception of Bitcoin and cryptoassets more generally, one of the key questions has been to what extent the networks are prone to market manipulation and opportunistic behaviours. The likelihood of both is a direct function of ownership concentration. To put it differently, the question of ownership concentration is not only a matter of curiosity but is of key relevance for both market participants and policy makers. From an individual investor’s perspective, it is important to understand how deep the market is, meaning how large trades by wealthy investors can impact prices. Similarly, from a regulatory perspective it is nonetheless important to understand who, and why, might be more incentive to market manipulation.

Figure 2: Breaking down ownership concentration

                 

Source: Aaro Capital Research
Notes: The left panel shows the relative number of addresses owned by small investors (<0.1% of the Bitcoin supply), medium investors (0.1%-1% of the Bitcoin supply) and whales (>1% of the Bitcoin supply) as of 24th May 2023. The right panel shows the same classification over time from December 2010 to May 2023.

Figure 2 provides a granular overview of ownership concentration by combining two key pieces of information: first, the distribution of addresses by holdings, and second, the distribution of supply by holdings. Those two data put together can shed some interesting light on the wallets aggregated by holdings expressed in USD. Interestingly, almost 70% of the addresses own less than $100 by the end of the sample, that is by May 2023. On the other hand, only 1% of the addresses own more than $100k over the same period, with a mere 0.1% of addresses holding more than $10 million. 

Notice that, by itself, this figure does not necessarily support the view that the Bitcoin ownership is extremely concentrated. We need another piece of evidence, that is the actual supply held by these addresses. The right panel of Figure 2 provides that information. The figure reports a breakdown over time of the Bitcoin supply held in addresses by their size of holdings. Two interesting facts emerge: first, and perhaps not surprisingly, large addresses account for the majority of Bitcoin supply. For instance, the addresses holding more than $100k hold approximately 90% of the Bitcoin supply by the end of the sample. The percentage decreases to 60% when focusing on addresses with more than $10 million of holdings. Second, the concentration of wealth increased over time, with only 20% of the Bitcoin supply held in the largest addresses (>$10 million) in 2017, compared to 60% in May 2023. This is a threefold increase in six years. The relative increase in the balance held in smaller accounts, between $100k and $10 million did not increase at the same pace.

By coupling the two dynamics in Figure 2 one obvious picture seems to emerge; as of May 2023, almost 95% of the accounts (those with less than $100k holdings as per the left panel) correspond to around 10% of the total supply (as per the right panel). This points towards a moderately high concentration of the aggregate Bitcoin network, at least in terms of wealth distribution.

Remarks and conclusion

A few words of caution are in order. Although instructive, the picture depicted by Figures 2 and 3 are far from definitive, and if anything represent an upper bound on ownership concentration. For starters, one would need to take out exchanges from the calculation of addresses and holdings. In a recent article, Makarov and Schoar (2022) showed that exchanges generate the most transaction volume in the Bitcoin network, with only 3% of activity reducible to illegal activities. The fraction of volume from miners is even smaller. In other words, it is reasonable to assume that a non-negligible, perhaps even considerable, part of the addresses with the largest holdings are indeed exchanges. This would change the interpretation of what the ownership concentration in the Bitcoin network actually means in the first place, as I alluded at the outset of the article. In addition, this simple analysis does not factor in custodians, lost coins, and wrapped Bitcoin. Including these aspects would reasonably push the upper bound even lower. 

Nevertheless, the simple analysis in this article shows that there is a sizable, yet time-varying, ownership concentration in the Bitcoin network. This potentially poses a challenge for developers, investors, policy makers and market participants at large. The resilience of a fully decentralised network critically lies in an evenly distributed property and ownership. Although in the Proof-of-Work protocol underneath Bitcoin wealth does not equal influence (contrary to Proof-of-Stake), the substantial inequality of wealth across network participants could hamper the adoption of Bitcoin, and perhaps of cryptoassets more broadly. Overall, this article poses some questions on the importance of understanding the structure of the Bitcoin network. 

Bibliography

Makarov, Igor, and Antoinette Schoar. Blockchain analysis of the Bitcoin market. No. w29396. National Bureau of Economic Research, 2021.

Haftungsausschluss 

Aaro Capital ist der Firmenname von Aaro Capital Limited („Aaro“), einer Gesellschaft mit beschränkter Haftung, die in England und Wales unter der Nummer 11419585 registriert ist und ihren Sitz im 5. Stock in 10-12 Eastcheap , London, Großbritannien, EC3M 1AJ hat.

Das in diesem Dokument enthaltene Material wird zu allgemeinen Informationszwecken bereitgestellt. Aaro Capital Limited bietet keine Anlageberatung an und gibt auch nicht an, eine solche anzubieten. Auf die in diesem Dokument enthaltenen Informationen sollte sich nicht verlassen werden. Außerdem sollten diese Informationen nicht als Grundlage für eine Anlageentscheidung dienen oder für die Beurteilung der potenziellen Geeignetheit einer bestimmten Anlage herangezogen werden. Die in dieser Präsentation dargestellten Zahlen beziehen sich auf die Vergangenheit oder sind nur als Beispiele aufgeführt. Die Wertentwicklung in der Vergangenheit ist kein zuverlässiger Indikator für zukünftige Ergebnisse.

Dieses Dokument kann Informationen über Kryptoanlagen enthalten. Kryptoanlagen befinden sich im Entwicklungsstadium, und jeder, der Investitionen in diese Art von Vermögenswerten erwägt, sollte vorsichtig sein und sachgemäße Beratung zu den mit diesen Vermögenswerten verbundenen Risiken einholen, einschließlich (aber nicht abschließend) deren Volatilität, des Totalverlustrisikos und der fehlenden Regulierung bestimmter Marktteilnehmer. Obwohl die Geschäftsführer von Aaro Capital Limited angemessene Anstrengungen unternommen haben, um die Richtigkeit der in diesem Dokument enthaltenen Informationen sicherzustellen, können weder die Aaro Capital Limited noch ihre Geschäftsführer eine Gewähr oder Garantie für die Richtigkeit und Vollständigkeit dieser Informationen übernehmen.

Bitte befragen Sie Ihren eigenen, ausreichend qualifizierten Finanzberater, wenn Sie eigene Anlageentscheidungen treffen.