What is FMI? More on that later. But first, let’s talk about Bitcoin.
If you aren’t familiar with the Bitcoin block size war and its endless online shouting matches which have evolved into legal and even death threats, then you have probably been a very productive human being and should sell hugs and not wander into a non-stop social media dance off.
Why? Because tens of thousands of man (and woman) hours have collectively been obliterated over a struggle that has illuminated that Bitcoin’s development process is anything but permissionless.
It also illuminates the poor fiduciary care that some VCs have towards their LPs. In this case, more than a handful of VCs do not seem to really care about what a few of their funded companies actually produce, unless of course the quarterly KPIs include “have your new Bitcoin meme retweeted 1,000 times once a week.”
In some documented cases, several dozen executives from VC-backed Bitcoin companies have spent thousands of hours debating this size attribute instead of building and shipping commercializable products. But hey, at least they sell cool hats and built up very large Twitter followings, right?
Fact #1: Satoshi Nakomoto did not ask anyone’s permission to launch, change, or modify the codebase she unilaterally released in 2009.
Fact #2: In 2009, when Satoshi Nakomoto issued and minted a new currency (or commodity or whatever these MLIC are) she did so without asking anyone else’s approval or for their “ack.”
In the approximately seven years since she stopped posting under her pseudonym, influential elements of Bitcoin’s anarchic community have intentionally created a permissioned developer system commonly referred to as the Bitcoin Improvement Proposal (BIP) process. “Bitcoin Core” is the name for the group that self-selected itself to vet BIPs; involvement is empirically permissioned because you can get kicked off the island.1 There are a small handful of decision makers that control access to the code repository.
For example, if you’re a developer that wants to create and launch a new implementation of Bitcoin that includes different block sizes… and you didn’t get it approved through this BIP process, guess what? You are doing permissionlessness wrong because you didn’t get permission from the BIP approval committee to do so.
Oh, but you realize that and still want to launch this new Bitcoin implementation with the help of other elements of the community, such as some miners and exchanges?
According to some vocal members of the current BIP approval committee (Bitcoin Core) and its surrogates, this is an attack on Bitcoin. Obviously this is absurd because there is no de jure or legally defined process for changing or forking Bitcoin, either the chain itself or the code.
There is no terms of service or contract which explicitly states what Bitcoin is and who controls its development process. Or more historically: if Satoshi didn’t need permission from a (non-existent) BIP approval committee to launch a cryptocurrency, then no other Bitcoin developer needs to either.
Fast forward to this current moment in time: if the Bitcoin Cash or Segwit2X forks are an attack on network because either fork did not get ack’ed (approved) by the right people on the BIP approval committee or retweeted by the right “thought leaders” on social media, then transitively every 10 minutes (when a block is generated by a miner) arguably could be an attack on Bitcoin.
Why? At any time a block maker (miner) could use a different software implementation with different consensus rules. They, like Satoshi before them, do not need permission to modify the code.
Oh, but other miners may not build on top of that block and some exchanges may not recognize those blocks as “legitimate” Bitcoin blocks?
That is certainly a risk. In fact, several exchanges are now effectively white listing and black listing — permissioning — Bitcoin-related blocks.
For instance, Bittrex, a large crypto-to-crypto exchange, has said:
The “BTC” ticker will remain the Bitcoin Core chain before the hard fork block. Bittrex will observe the Bitcoin network for a period of 24 to 48 hours to determine if a chain split has occurred and the outcome.
In the event of a chain split, “BTC” will remain the existing Bitcoin chain with 1 MB blocks until the industry and ecosystem demonstrates a clear chain preference for Bitcoin.
The incumbent implementation (based on the existing Bitcoin consensus protocol) will continue to trade as BTC even if the B2X chain has more hashing power.
After heavy public (and private) lobbying by members and surrogates of Bitcoin Core, other exchanges have instituted similar policies favoring the incumbent.2 So what can alternative implementations to do? Bend the knee?
Historically miners have built on the chain that is both the longest and also has the most accumulated difficulty… and one that has enough profitability to pay for the electricity bills. It just happens that this collective block building activity is never called an “attack” because in general, most participants have been happy enough with the status quo.
Visions of what Bitcoin is and how it should be defined have clearly, empirically shifted over time. But since this network was purposefully designed to be self-sovereign and anarchic — lacking contracts and hooks into any legal system — no one group can claim legitimacy over its evolution or its forks.
As a result, recent war cry’s that Segwit2X is a “51% attack” on Bitcoin are a red herring too because there is no consensus on the definition of what Bitcoin is or why the previous block – in which approximately 51% of the hashrate created a block – is not an attack on Bitcoin. 3
This has now morphed into what the “BTC” ticker on exchanges represents. Is it the longest chain? The chain with the most accumulated difficulty? The chain maintained by Bitcoin Core or now defunct NYA developers? If a group of block makers can build blocks and exchanges are willing to list these coins as “BTC” then that specific chain has just as much legitimacy as any other fork other miners build on top of and exchanges may list.
Furthermore, if the BIP approval committee gets to say what software miners or exchanges should or should not use (e.g., such as increasing or decreasing the block size), that could mean that existing network is a managed and even administered. And this could have legal implications. Recall that in the past, because block making and development were originally separate, FinCEN and other regulators issued guidance stating that decentralized cryptocurrencies were exempt from money transmission laws.
Despite what the trade associations and Bitcoin lobbying groups would like the narrative to be, I recently published an article that went into this very topic in depth and have publicly asked several prominent “crypto lawyers” to provide evidence to the contrary (they have yet to do so). An argument could be made that these dev groups are not just a loose collective of volunteers.
Financial market infrastructure
I’m not defending S2X or XT or Bitcoin Unlimited. In fact, I have no coins of any sort at this time.
But even if you don’t own any bitcoins or cryptocurrencies at all, the block size debate could impact you if you have invested in the formal financial marketplace.
For example, if and when the CME (and similar exchanges) get CFTC approval to list cryptocurrency-related futures products and/or the NYSE (and similar exchanges) get SEC approval to list cryptocurrency-related ETFs, these products will likely result in a flood of institutional money.
Once institutions, regulators, and sophisticated investors enter the picture, they will want to hold people accountable for actions. This could include nebulous “general partnerships” that control GitHub repositories. Recall, in its dressing down of The DAO, the SEC defined the loose collective building and maintaining The DAO as a ‘general partnership.’ Is Bitcoin Core or other identifiable development teams a “general partnership”?
Maybe. In fact, the common refrain Bitcoin Core and its surrogates continually use amounts to arguments in favor of a purported natural monopoly.
For instance, Joi Ito, Director of MIT’s MediaLab, recently stated that:
“We haven’t won the battle yet. [But] I think the thing that is interesting is that Bitcoin Core has substantially more brain fire power than any of the other networks.”
This is problematic for a couple reasons.
First, Joi Ito is not a disinterested party in this debate. Through Digital Garage (which he co-founded) it has invested in Blockstream, a company that employs several influential Bitcoin Core devs.4 Ignoring the potential conflict of interest, Ito’s remarks echo a similar sentiment he also made last year, that Core is basically “The Right Stuff” for NASA: they are the only team capable of sending humans into space.
But this is an empirically poor analogy because it ignores technology transfer and aerospace education… and the fact that multiple countries have independently, safely sent humans, animals, and satellites into space.
It also ignores how competitive verticals typically have more than just one dominant enterprise: aerospace, automobiles, semiconductor manufacturers, consumer electronic manufacturers (smart phones), etc. Each of these has more than one company providing goods and services and even usually more than just one product development team developing those. Intel, for example, has dozens of design teams working on many new chips at any given time of the year. And they are just one of the major semiconductor companies.
Even in the highly regulated markets like financial services there is more than one bank. In fact, most people are unaware of this but banks themselves utilize what is called “Core Banking Software” and there are more than a dozen vendors that build these (see image below).
It is a bit ironic that Bitcoin Core seeks to have a monopoly on the BIP process yet even banks have more than one vendor to choose from for mission critical software securely managing and processing trillions of dollars in assets each day.5
On the enterprise (non-anarchic) blockchain side of the ecosystem, there are well over a dozen funded teams shipping code, some of which is being used in pilots by regulated institutions that are liable if a system breaks. Note: this is something I discussed in my keynote speech (slides) at the Korea Financial Telecommunications and Clearings Institute last year.
But as one vocal Core supporter in a WeChat room recently said, Bitcoin Core is equivalent to Fedwire or Swift, there is only one of each; so too does it make sense for only one Bitcoin dev team to exist.
Firstly, this conflates at least four different things: a specific codebase, with permissioned dev roles, with acceptance processes, with a formal organization.
It is also not a good analogy because there are many regulatory reasons why these two systems (Swift and Fedwire) exist the way they do, and part of it is because they were either setup by regulators and/or regulated organizations. In effect, they have a bit of a legally ring-fenced marketplace to solve specific industry problems (though this is somewhat debatable because there are some alternatives now).6
If this supporter is equating Core, the codebase, with real financial market infrastructure (FMI), then they should be prepared to be potentially regulated. Bitcoin Core and many other centralized development teams are comprised of self-appointed, vocal developers that are easy to identify (they have setup verified Twitter accounts and attend many public events), so subpoenas and RFI’s can be sent their way.
As I mentioned in my previous article: with great power comes great accountability. Depending on the jurisdiction, Core and other teams could end up with regulatory oversight since they insist on having a monopoly on the main (only) implementation and process by which the implementation is managed.7
Remember that Venn diagram at the very top? The companies and organizations that manage FMI today for central banks (RTGSs), central securities depositories (CSDs), and other intermediaries such as custodians and CCPs, have specific legal and contractual obligations and liabilities.
Following the most recent financial crisis, the G-20 and other counties and organizations established the Financial Stability Board (FSB) to better coordinate and get a handle on systemic risks (among other issues). And while the genesis of the principles for financial market infrastructures (PFMI) had existed prior to the creation of the FSB, how many of the international PFMI standards and principles does Bitcoin Core comply with?
Spoiler alert: essentially none, because Satoshi intentionally wasn’t trying to solve problems for banks. So it is unsurprising that Bitcoin isn’t up to snuff when it comes to meeting the functional and non-functional requirements of a global payments platform for regulated institutions. Fact-check me by reading through the PFMI 101 guide.
When presented with these strong legal accountability and international standards that are part and parcel with running a payment system, there is lots of hand waving excuses and justifications from Core supporters (and surrogates) as to why they are exempt but if Core wants to enforce its monopoly it can’t have it both ways. Depending on the jurisdiction they may or may not be scrutinized as FMI.
But in contrast, in looking at the evolution and development of the enterprise chain ecosystem – as I described in multiple previous articles – there are valuable lessons that can be learned from these vendors as to how they plan to operate a compliant network. I recall one conversation with several managing directors at a large US investment bank over a year ago: maybe the enterprise side should just have CLS run a blockchain system since they have all the right business connections and fulfill the legal and regulatory check boxes.
Note: CLS is a very important FMI operator. Maybe existing FMI operators will do just that. Speaking of which, will Bitcoin Core (or other dev teams) apply to participate with organizations like the FSB that monitor systemically important financial institutions and infrastructure?
Angela Walch has argued (slides) that some coders, especially of anarchic chains, are a type of fiduciary.8 Even if this were not true, many countries have anti-monopoly and anti-trust laws, with some exceptions for specific market segments and verticals. There are also laws against organized efforts involved in racketeering; in the US these are found within the RICO Act.
I haven’t seen a formal argument as to why Core or other development teams could meet the litmus test for being prosecuted under RICO laws (though the networks they build and administer are frequently used for money laundering and other illicit activity). But trying to use the “decentralization” trump card when in fact development is centralized and decisions are made by a few key individuals, might not work.
Look no further than the string-pulling Mafia which tried to decentralize its operations only for the top decision makers to ultimately be held liable for the activities of their minions.9 And using sock puppets and pseudonyms might not be full proof once forensic specialists are brought in during the discovery phase.10
Based on observations from how Bitcoin Core evolved and consolidated its power over time (e.g. removing participants who have proposed alternative scaling solutions), the focus on what Bitcoin is called and defined has landed in the hands of exchanges and really just highlights the distance that Bitcoin has walked away from a “peer-to-peer electronic cash” that initially pitched removing intermediaries. To even care about what ticker symbol ‘Bitcoin’ is on an exchange is to acknowledge the need for a centralized entity that establishes what the “price” is and by doing so takes away the bitcoin holder’s “self-sovereignty.”11
While the power struggles between various factions within the Bitcoin development community will likely rage on for years, by permissioning off the development process, Bitcoin Core (and any other identifiable development groups), have likely only begun to face the potential regulatory mine field they have foisted on themselves.12
Historically blockchain-based systems have and still are highly dependent on the input and decision-making by people: somebody has to be in charge or nothing gets done and upgrades are a mess. And the goal of appointing or choosing specific teams on anarchic chains seems to be based around resolving political divisions without disruptive network splits.13
The big questions now are: once these teams are in charge, what will governments expectations be? What legal responsibilities and regulatory oversight will the developers have? Can they be sued for anti-trust and/or RICO violations? With billions of dollars on the line, will they need to submit upgrade and road map proposals for approval?
- Examples of developers who were removed: Alex Waters, Jeff Garzik, Gavin Andresen [↩]
- Thanks to Ciaran Murray for identifying these exchanges. [↩]
- Bitcoin mining is in fact based on an inhomogeneous Poisson process; a participant could theoretically find a block with relatively little hash rate. Although due to the probabilities involved, most miners pool their resources together to reduce the variance in payouts. [↩]
- According to one alleged leak, Digital Garage is testing Confidential Assets, a product of Blockstream. [↩]
- According to a paper from the Federal Reserve: payment, clearing, and settlement systems in the United States “process approximately 600 million transactions per day, valued at over $12.6 trillion.” [↩]
- On AngelList, there are about 3,400 companies categorized as “payments” — most of these live on top of existing FMI, only a handful are trying to build new independent infrastructure. [↩]
- A key difference between Bitcoin and say Ethereum is that with Ethereum there are multiple different usable implementations managed by independent teams and organizations; not so with how Bitcoin has evolved with just one (Bitcoin Core) used by miners. In addition, the Ethereum community early on formally laid out a reference specification of the EVM in its yellow paper; Bitcoin lacks a formal reference specification beyond the Core codebase itself. [↩]
- See also The Bitcoin Blockchain as Financial Market Infrastructure: A Consideration of Operational Risk from Angela Walch [↩]
- Thanks to Stephen Palley for providing this observation. [↩]
- It is unclear why the current Bitcoin Core team is put onto a pedestal. There are many other teams around the world building and shipping blockchain-related system code used by companies and organizations (it is not like there is only just one dev team that can build all databases or operating systems). At the time of this writing Core has not publish any papers in peer-reviewed journals and many of them do not have public resumes or LinkedIn profiles because they have burned business and professional relationships in the past. Irrespective of what their bonafides may or may not be, it is arguably a non sequitur that ‘permissionless’ coordination in open-source code development has to lead to a monopoly on said development. [↩]
- Thanks to Colin Platt for this “appeal to authority” observation. [↩]
- Bitcoin stopped being permissionless when developers, miners, and exchanges needed to obtain permission to make and use different code. And likely there are and will be more other cryptocurrency development teams that follow that same path. [↩]
- For an informed contrarian view on governance and distributed ledger technology, see The blockchain paradox: Why distributed ledger technologies may do little to transform the economy by Vili Lehdonvirta [↩]