This past week I gave a new presentation at the 2nd annual Soranomics event (last year I presented on a related topic: pegged coins aka “stablecoins”). It includes a number of illustrations to discuss product market fit and infrastructure market fit.
Below is a copy of the deck as well as the A/V. Note: there are citations and references in the speaker notes. Note: I am to publish a long-form version based on this content.
I’ve previously reviewed at least seven blockchain-specific books in the past number of years. No one has paid me to review them, although I have received a couple copies for free. Unfortunately more than half of the books have been pretty bad… both technically wrong and often very polemical.
Fortunately, a page turner appeared in my inbox about a month ago: “The Billionaire’s Folly” by Faisal Khan. I’ve already posted a couple of short comments on the bird app and an usual for me – do not have a lot more to add. Mostly because it lacked many errors. Sure, it had a couple of typos here and there and a couple of debatable points but overall it was well-written and informative.
It also didn’t try to stray far away from what it aimed to do: discuss Khan’s perspective working at ConsenSys, an Ethereum-focused company, during what turned out to be the heady days of the ICO era. So in some ways, it is closer to Nathaniel Popper’s Digital Gold (which was equally well-written) than most of the other b-word books.
I didn’t mention this in the thread above but a number of anecdotes that Khan shares in the book were either relayed to myself (often through co-workers) or by actually witnessing it first hand. So it is interesting to see some of them independently confirmed.
One that did not (because Khan had yet to join the company) but definitely could have fit right in, involved an event held in the spring of 2016 near Seattle. About 60ish employees of this Fortune 100 tech company hosted a day-long powwow about “blockchains” and only three external companies were allowed to send representatives:
A well-known, large consulting company
ConsenSys, who sent several executives
R3 (my then-employer) sent myself
After presentations were given, the floor was opened for questions and a senior architect in the back questioned the urgency and immediacy that one of the promoters had claimed. And during the ensuing war-of-words, a partner at the consulting firm literally stood on the table at this closed-room event, crooning to everyone that “blockchain was the biggest thing in his career and that it would dramatically impact this tech company.” One of Khan’s future colleagues from the table over made eye contact with me and we just shook our heads. Although in retrospect, he was probably shaking his head for very different reasons than I was. I’m ngmi, right?
Either way, Khan has oodles of stories packed into a book that isn’t polemical. Check it out.
[Note: Below is a guest post discussing a “real world asset” (RWA) tokenization use case that has been proposed and re-proposed for about a decade (first with Colored Coins and Counterparty). Among other articles, the author previously co-authored another thoughtful piece A Quick History of Cryptocurrenices BBTC — Before Bitcoin. Reprinted with permission; and the views are those of the authors alone.]
By Ken Griffith
Since the appearance of Bitcoin in 2008 numerous people have had the idea of issuing gold tokens on a blockchain, or using a blockchain to support a digital gold currency system. This short essay will look at some of the attempts to do this and suggest the reasons why this has not worked, and is unlikely to ever work.
In 2015 Roy Sebag borrowed a phrase from Satoshi and created a company called “Bitgold” which performed a reverse IPO on a Canadian stock exchange. Sebag raised several hundred million dollars to “put gold on the blockchain.” However, instead of creating Bitgold, he proposed a buyout to James Turk at GoldMoney.com. After buying out goldmoney.com, Sebag simply invested the funds into expanding that business which had been founded in 2001. Goldmoney has a ledger, but it does not use blockchain at all.
In the intervening years a dozen or more different gold tokens have been created on Ethereum and other platforms. Of these, many turned out to be scams, and one or two of them were legitimately backed by actual gold. However, none of these have gained any traction in the marketplace.
To understand why gold and blockchain do not mix we need to look at the history of digital gold and cryptocurrencies. E-gold was the first Internet money in 1996, 12 years before the Bitcoin whitepaper was published. By 2001 e-gold had one million users worldwide, and had an annual transaction volume of about US $2 billion per year worth of gold. By 2005 there were six such digital gold issuers. However, the US Treasury began a campaign of prosecuting the digital gold companies using the USA Patriot Act from 2006 to 2009, in which the four USA based issuers were indicted, their gold reserves seized, and officers indicted.
Goldmoney.com was the only digital gold issuer to survive the purge because they were fully licensed in Jersey, UK. However, they never allowed an independent network of exchange agents to provide exchange services to their users. So Goldmoney.com never became a popular means of payment. Goldmoney was primarily used by institutional investors to hold large amounts of physical gold.
The primary obstacle to developing a gold payments network is government regulations in various countries.
Blockchain tokens such as Bitcoin have the advantage of being decentralized with no person or company as the issuer. They are resistant to government regulation and control because there is no central server that can be seized or turned off. There is no person who can be arrested to stop the Bitcoin network from continuing to process transactions. The network continues to operate so long as the people operating servers with this software continue to operate them and keep the same rules.
When you create a gold-backed token you violate the basic social reason that has allowed blockchains to ignore government regulations. An asset-backed token is a promise by some person or company to deliver some asset in exchange for the token. That is implicitly a contract between the issuer and the holders of the token. The token may live on a blockchain, but the person who issues the contract lives in the real world. The physical assets used to back the token also exist in the real world. A government can arrest a person or a company and confiscate the physical assets.
The Crux of the Problem
Once a token is issued on a permissionless blockchain it is impossible for the issuer to control who gains access to that token. It can be transferred to someone in a country where such tokens are illegal or otherwise regulated.
We saw this in the case of EOS, which raised $4 billion in a year-long token sale from 2017 to 2018. The terms of the sale expressly forbade US Citizens from buying the tokens. The sale contract required the buyer to swear that they were not a US Citizen or resident of the United States. The website blocked US IP-addresses so the token sale could not even be seen by web browsers in the United States.
Two years after the token sale was complete, the SEC began an investigation, and a class action lawsuit was filed against EOS for selling tokens to US citizens. This means that some US citizens made a false statement under oath to buy those EOS tokens, and then handed that evidence to US law enforcement agencies. Because the company had raised $4 billion, the large pile of money was very attractive to US law enforcement agencies. EOS ended up settling with the SEC for 24 million dollars.
The value of the assets is the critical factor in determining which companies are prosecuted and which ones are ignored.
We conclude that any tangible asset digitized on a blockchain is vulnerable to lawsuits and criminal prosecutions from countries that regulate or otherwise make such payments illegal.
Law enforcement agencies do not immediately prosecute every company who issues such a token. They bide their time and wait until they see a large pile of value that is vulnerable to prosecution. Only then, do they begin an investigation and prosecution in order to seize those assets using asset forfeiture laws.
How to Avoid this Problem to Issue Real Digital Assets
An Internet ecosystem which allows persons to issue real-world financial contracts for various assets would require a contract-based ledger such as one based on Ricardian Contracts.
Rather than creating one company, such as Goldmoney.com, that offers gold accounts to citizens of many countries, it is better to create a network of lawful institutions in different countries that only offer asset accounts to residents of that country. Thus, each institution need only concern itself with the laws of one nation, rather than the laws of all nations.
A clearing mechanism will allow payments to clear between different institutions in different countries. This short video explains the concept of a clearinghouse in two minutes:
It is better to have many small issuers of gold or other assets than one large centralized issuer. Each of the small issuers should have a license for a money service business in their jurisdiction.
Since corruption is part of human nature, and will always be a problem, it makes more sense to build a system that anticipates the cost-benefit factors that drive corruption. Even law enforcement agencies have a cost to seize financial assets. They have to bring a case in court, which costs money.
Many small piles of gold are much more expensive for a government to seize than one large pile of gold. If there are many issuers, then a government would have to indict each issuer with a separate court case in a separate jurisdiction. This is too expensive and inefficient. Ideally, you would want issuers to hold the amount of gold roughly equal or less in value than the cost of prosecution. So, for example, a typical prosecution of such a company would cost $1 to $5 million USD.
In traditional banking, bank ledgers were protected by financial privacy laws. Unfortunately, the twentieth century saw the steady weaponization of banks against their customers by the State. However, banking privacy worked effectively for centuries. The best long term solution is for states to reform their banking laws to restore and protect financial freedom and privacy.
We find that permissionless blockchains are not environments that are ever likely to work for contract-based financial transactions. They have no native means to record consent to a contract, nor to restrict access to a token to those who have consented to a contract.
By contrast, a Ricardian Contract based ledger would be ideally suited to enable an online financial ecosystem with securities and asset-backed tokens operating within the law. Ricardian Contracts were invented a decade before Bitcoin. Yet, there has been very little community interest in building Ricardian Contract based systems. The reason for this appears to be ideological.
The ideological spectrum of cryptocurrency users leans heavily towards anarchy. Even classical libertarians believe in the necessity of courts of law to enforce contracts. Yet, the majority of cryptocurrency fans seem to reject even the idea of courts of law, as seen in their reaction to the CSW defamation case. Without courts of law, contracts are unenforceable.
The idea that computer code can be a replacement for law and courts is popular but fatally flawed. It is extremely difficult to write and maintain error-free software. It is fine to say that the code is the law, until an error in the code allows an outcome that was not intended. Human courts of law are the ultimate error processing routine. This is unavoidable, and therefore should be embraced in the design of any financial ecosystem.
The dream of creating online financial ecosystems that live on the blockchain, free from courts of law is doomed to perpetual anarchy. Financial institutions, by definition, hold assets in trust for their account holders. Such relationships require contracts and courts of law to enforce. Without contracts and courts of law, all that can be expected is a wild west of online fraud. The long string of failed cryptocurrency exchanges and other projects testifies to the truth of that assertion.
[Note: The views expressed below are solely my own and do not necessarily represent the views of my employer or any organization I advise.]
As we have discussed before, “Web3” is a nebulous term that has been used to market a slew of products and services, often via “chainwashing.”
What is “Web3?”
This past week 25 guys and one gal signed and published a 741-word letter to senior U.S. legislators calling for “Support of Responsible Fintech Policy.” And while many “Web3” promoters do deserve a good chastising, this letter has many technical shortcomings and is a disappointment to those who have been in the trenches for years… before being a “critic” was considered en vogue. Worst, it doesn’t define what “Web3” or even a “blockchain” is or is not.
But let’s start with a comment that I thought was pretty good, the intro:
“Today, we write to you urging you to take a critical, skeptical approach toward industry claims that crypto-assets (sometimes called cryptocurrencies, crypto tokens, or web3) are an innovative technology that is unreservedly good. We urge you to resist pressure from digital asset industry financiers, lobbyists, and boosters to create a regulatory safe haven for these risky, flawed, and unproven digital financial instruments and to instead take an approach that protects the public interest and ensures technology is deployed in genuine service to the needs of ordinary citizens.”
I – along with a number of other independent researchers such as Angela Walch (who they referenced) – have publicly made similar requests in the past. For instance, the original conclusion in my 2018 WSJ op-ed expanded upon the lack of transparency and surveillance sharing for why the SEC has not approved a bitcoin-denominated ETF by stating, “…the retail public wants seductive narratives and fantastical returns. The supply of fraud will therefore grow to meet that demand.”
To reuse a cliché analogy, throughout most of 2021 you could probably throw a baseball at a collection of dapps and hit one that at the very least, played fast and loose with marketing high APR yields.
This was followed with a quizzical take:
“Not all innovation is unqualifiedly good; not everything that we can build should be built. The history of technology is full of dead ends, false starts, and wrong turns. Append-only digital ledgers are not a new innovation. They have been known and used since 1980 for rather limited functions.”
The first sentence probably has a lot of supporters, including myself, as it relates to non-proliferation of weapons of mass destruction. The somber and horrific legacies of the atomic and hydrogen bombs are certainly an example of something that should not have been built.
But the shade thrown at “append-only digital ledgers” is pretty farcical. Why do these authors get to determine what is or is not useful in the spring of 2022?
For instance, if we look at the core moving pieces of the Bitcoin blockchain, all of the main elements (“prerequisites“) had been around for years. And it was by assembling them together that we have arguably the first blockchain.1 The authors are taking a page from the lazy Maximalist playbook, one that does not withstand empirical scrutiny.
In looking at the “tech stack” of Big Tech, Google maintains a project called “Certificate Transparency” (implemented as “Trillian“).2 Certificate Transparency is not a blockchain, but it is a Merkle tree of things which are interconnected and signed and in production today.
From the Trillian team:
The ideas underpinning Certificate Transparency, Revocation Transparency and related efforts are not specific to certificates, but can in fact be used to make almost anything transparent. These technologies are strongly related to the much-hyped blockchain. The reality, of course, is that there isn’t a “the” blockchain, and that decentralisation is not always the answer. We are not making “the” blockchain, and we do not claim to support decentralisation.
As mentioned in a previous post, the problem with the a priori position that anti-coiners (and many maximalists) have is that over time they continually get backed up into a corner. Why? Because over the past decade we continue to see – empirically – how blockchains and blockchain-like elements are incorporated by a spectrum of organizations from Big Tech and Big Finance all the way down to small startups.
As Matthew Green (a cryptographer) explains in a thread on this topic, the granular fine points around “blockchain technology” is mostly bad:
Unfortunately the authors – while seemingly well intentioned – do not clearly state what parts of a blockchain they dislike, what parts of “distributed ledger technology” that they explicitly think is bad.
Furthermore, the idea of a neutrally owned, shared ledger is not a new concept. Several initiatives in the financial industry — such as a Joint Back Office (JBO) — pre-date the euphoria around blockchains but languished in concept mode.3 What is the lure for maintaining a shared ledger between (competing) organizations? Resiliency and reduction of reconciliation often come up as two of the main reasons but the list is long and deserves its own post. Suffice to say, claiming that “append-only digital ledgers” are a plaything of the ’80s is not even wrong.
Another broad sweeping set of statements that lack precision:
As software engineers and technologists with deep expertise in our fields, we dispute the claims made in recent years about the novelty and potential of blockchain technology. Blockchain technology cannot, and will not, have transaction reversal mechanisms because they are antithetical to its base design.
As Green and Byrne (among other responders) have pointed out, there is a missing nuance by the authors in that there are different types of blockchains. For instance, depending on the implementation some permissioned blockchains allow – in theory – certain participants to freeze transactions.4
Likewise on public chains, administrators of USDC, USDT, and other collateral-backed pegged coins, regularly blacklist and freeze transactions. In fact, any chain with smart contract functionality can provide some form of reversibility (or at the very least, freezing of state). We also see this empirically during and after exploits, with developer teams freezing tokens.
This is a strange miss because one of the signatories is Stephen Diehl, who as far back as July 2017 (when I spoke to him in an official meeting) was/is the CTO and director at Adjoint, which is a British private blockchain firm that has previously announced payment-related partnerships.
This statement starts out good:
Similarly, most public blockchain-based financial products are a disaster for financial privacy; the exceptions are a handful of emerging privacy-focused blockchain finance alternatives, and these are a gift to money-launderers. Financial technologies that serve the public must always have mechanisms for fraud mitigation and allow a human-in-the-loop to reverse transactions; blockchain permits neither.
Green (and suzuha) points out that the authors are trying to have their cake and eat it too:
For example, as far back as 2015, banks involved in R3 presented use-cases that required – by law – protection of PII. At the time, any company or organization wanting to engage with regulated financial institutions quickly learned how PII was an unmovable touchstone (see this related presentation). And so from those functional requirements arose different solutions ranging from hardware-based solutions (like SGX) to software-based solutions (like ZK-Snarks). The public chain world was often where these ideas either first originated or at the very least, first tested.5
Over the years I have regularly pointed out how privacy and confidentiality-features could be used for a sundry of illicit activities. But just because it could be used by those types of actors, does not mean it regularly is.
On that point, in 2016 I helped edit a paper on this very topic. It was co-authored by Danny Yang (founder of Blockseer), Zooko Wilcox-O’Hearn, and Jack Gavigan. Wilcox-O’Hearn and Gavigan are executives at the Electric Coin Company, a for-profit company leading the development of Zcash. Worth pointing out that one of the signatories on the letter above amplified false information about myself two months ago, claiming I was not an advisor at Blockseer. Not only is this false, but I still own the equity in DMG Blockchain (which acquired Blockseer four years ago). This calls into question the credibility of the individuals amplifying information they did not fact check. What other false information are they claiming about blockchains?
Scare quotes is not the only thing that harms this section:
By its very design, blockchain technology, specifically so-called “public blockchains”, are poorly suited for just about every purpose currently touted as a present or potential source of public benefit. From its inception, this technology has been a solution in search of a problem and has now latched onto concepts such as financial inclusion and data transparency to justify its existence, despite far better solutions already in use.
The paragraph preceding this one also mentions “public blockchains” but doesn’t use quotes around it. And neither defines or provides nuance to explain the differences between “permissioned” (or private) blockchains compared with “public” (or anarchic) blockchains.
Either way, the authors make a good argument about how pulling on the heart strings of financial inclusion is mostly bupkis and I agree, and others have pointed that this rings hollow too.6 To strengthen this, the authors should have provided a citation or at least an example of “far better solutions already in use.” For example, Raúl Carrillo (who is not one of the listed authors) has pointed to Postal Banking as a possible avenue for (re)banking not just marginalized persons. Blockchains aren’t need for that or arguably for other retail activity.7
The next part of the paragraph is painfully arbitrary:
After more than thirteen years of development, it has severe limitations and design flaws that preclude almost all applications that deal with public customer data and regulated financial transactions and are not an improvement on existing non-blockchain solutions.
First of all, the first web browser (appropriately called the “WorldWideWeb“) was launched in 1990. It wasn’t until 2004 that Google revealed Ajax-based Gmail followed by Google Maps. If the authors are trying to make the claim that anything (everything?) useful should have been invented in 13 years then they should hold other tech initiatives to the same standard.
The lack of nuance in this letter is striking because not every blockchain is based on the purposefully limited architecture of Bitcoin. Between 2009-2015, a typical on-chain user could only access Bitcoin or a Bitcoin-based fork or clone (like Litecoin). Ethereum and other chains with a virtual machine, did not launch until the summer of 2015.8 That is part of the reason why regulated financial institutions (Big Banks) and large technology companies (Big Tech) began deploying resources in this sector in 2015: first with consortia and later setting up their own internal teams of subject matter experts. What a user could do with a blockchain changed over time thus a priori declaring “almost all applications” dead is incredulous.
And again, the authors provide no examples of what “existing non-blockchain solutions” they are referring to. For example, every single major vendor that provides core banking software for banks — such as FIS, Fiserve, and Jack Henry — have integrated tools that enable the software to interact with or hook into a blockchain. Every major Big Cloud vendor provides both tools for blockchain node operators as well as dedicated “Web3” development teams to compete with Alchemy and Infura. Several CSDs and CCPs have invested in a blockchain-focused company (like Digital Asset or Axoni) and have announced blockchain-based pilots. Pretending that this digitization and tokenization trend is not occurring beyond niche NFT art collections is intellectually dishonest.
I agree with most of this statement but it needs nuance:
Finally, blockchain technologies facilitate few, if any, real-economy uses. On the other hand, the underlying crypto-assets have been the vehicle for unsound and highly volatile speculative investment schemes that are being actively promoted to retail investors who may be unable to understand their nature and risk. Other significant externalities include threats to national security through money laundering and ransomware attacks, financial stability risks from high price volatility, speculation and susceptibility to run risk, massive climate emissions from the proof-of-work technology utilized by some of the most widely traded crypto-assets, and investor risk from large scale scams and other criminal financial activity.
The nuance these authors need to include is defining what “blockchain technology” is and is not. Trillian is not a blockchain but shares several common elements. Thus throwing the baby with the bath water flies in the face of the empirical reality.9 As far as criticisms around the negative externalities created by proof-of-work-based blockchains: I 100% agree. I have written on this topic roughly every 18 months. What would strengthen their statement is to provide actual statistics and data regarding each of their points (the data exists from companies like Chainalysis or previously, Blockseer).
Their polemical statement meanders on a bit more but this statement is worth assessing:
The catastrophes and externalities related to blockchain technologies and crypto-asset investments are neither isolated nor are they growing pains of a nascent technology. They are the inevitable outcomes of a technology that is not built for purpose and will remain forever unsuitable as a foundation for large-scale economic activity.
The second sentence falls under Hitchens razor: that which is presented without evidence can be dismissed without evidence. In fact, we do know why Bitcoin was built, Satoshi explained it at length on mailing lists and in the white paper. And Bitcoin was just the first “blockchain,” other chains have arisen later that fulfill other requirements. Onyx from JP Morgan is now being used for trading intraday repos. Maybe Onyx is just a flash in the pan, but it serves as a narrative violation — and there are more than a dozen other examples that the authors are likely unaware of, just read Ledger Insights each week.10
Lastly, in the Financial Times, one of the authors was quoted saying:
“The computational power is equivalent to what you could do in a centralised way with a $100 computer,” said de Icaza. “We’re essentially wasting millions of dollars’ worth of equipment because we’ve decided that we don’t trust the banking system.”
This is true with respect to proof-of-work-based blockchains but not at all relevant to alternate Sybil resistant models like proof-of-stake (P-o-S). Conflating the two is not accurate. Also, de Icaza and others needlessly defend the status quo, both with comments like this as well as the letter itself. Fortunately for retail, “the banking system” is not completely static and changes over time (it is also not a single monolithic entity). Also, not a single author listed works for a financial institution yet opines on it; there are plenty of blockchain “skeptics” within the financial industry why not find one?
Which brings us to the next section.
(Un)intentionally defending the status quo
The only reason to publicly identify themselves is to give weight or credibility to the matters discussed in the letter. Even though this letter was directed at U.S. congressmen and women, more than half of the signees are neither US residents or citizens. Even though more than a handful work at public tech companies or large organizations that rely on donations, let us give them the benefit of the doubt that they were not explicitly defending the status quo.
Yet without offering specifics beyond vague “non-blockchain solutions,” the authors are implicitly defending both systemically important financial institutions (SIFIs) and systemically important cloud providers. Both are bad for society and we should not defend their existence.
It is worth pointing out apart from two or three, most of these authors were not actively critical during the very public 2017-2018 ICO boom.11 What has motivated them to self-deputize and attempt to police what can and cannot be done with a blockchain in 2022 and ignore those who have been pathfinders in prior years? Perhaps there is a good reason, busy solving other worldly problems. I am certainly a fan of more introspection by disinterested parties!
I have written about it before but if the aim is to (1) influence policy makers and work with (2) regulators, there are at least two ways to achieve their goals:
Set up a not-for-profit lobbying organization modeled after Coin Center… the Anti-Coin Center. Hire former regulators and policy makers and re-use the lobbyist blueprint to engage with decision makers. A couple of years ago I wrote out a general overview to a couple L1 creators, it’s not complicated. You don’t even need a blockchain. But it does require some capital to hire for various roles, so it is not completely lean (e.g., would probably need to hire an actual blockchain engineer instead of relying on IT administrators). Oh and someone who posts frivolous memes all day is a must.
About four months ago, I asked one of the authors to submit their concerns directly with various agencies, such as the SEC and CFTC. This can be done formally through a whistle blower process (I’ve done it!). An ad hoc Hail Mary… is to informally do so through letter writing campaigns coordinated on social media. And as they haven’t stated otherwise, instead of submitting paperwork, some of these authors spend all day engagement farming on social media. If the outcome is “to get regulators to do something” this seems suboptimal because U.S. regulators typically need a paper trail to get the bureaucracy moving.
The blockchain world needs critics and criticism but it also needs criticism that is technically valid. And this letter is not only imprecise but sounds like something incumbent technology firms would write to defend their turf (which probably isn’t how it originated).
Over the past 18 months, the most recent coin bull market brought in a slew of new commentators a few of whom have attempted to co-opt the term “critic.” Clearly no one owns this term, there is no monopoly on it. Heck, I’ve even been labeled a “crypto” or “bitcoin” critic on more than one occasion. Yet we are seeing a cottage industry of professional “skeptics” who have a priori made up their mind irrespective of the evidence presented.
In addition to writing the most widely cited paper on “permissioned” blockchains, I wrote the first long form discussion on potential systemic important cryptocurrency networks in 2018 and think it is a bit absurd that some anti-coin commentary claims that cyber coins currently threaten the entire financial system. Feel free to disagree, but the onus is on the party making the positive claim. The counterfactual occurred the past five months: more than half of the aggregate coin marketcap evaporated. As collateral-backed pegged coins unwound, they did not lead to massive treasury liquidations crushing the traditional financial market.12
This is not defending the way centralized, commercial-bank backed pegged coins arose or currently operate.13 Rather it is a statement of fact: today the cyber coin world is not “too big to fail” and hopefully it never will be. Contagion can be real and should be simulated and stress tested!14 There are plenty of good criticisms to be lobbed at the “Web3” world, none of which requires making up fanciful conspiracies or playing fast and loose with technical verbiage.
If we are going to (rightly) criticize startups, investors, and other interested parties for mis-marketing “Web3” we should provide specific reasons as well as definitions. And while we are at it, let us bring a fine comb and scrutinize other hyped tech verticals that dramatically impact the well being of individuals such as: A.I. and workplace discrimination, privacy rights over data (including identity).15
Crusades can be big tent and incorporate more than just a small echo chamber of folks who (rightly) point out that a lot of cryptocurrency buzz is likely a financial grift with little real utility. Yet it is not a coincidence that perhaps the best critics are actual practioners, engineers, and architects who saw the limitations or drawbacks in certain blockchain designs and decided to build a different way. If there is a second version of this letter, it is highly recommended that input from outsiders be solicited. Including the world’s richest man, Colin Platt!
Or maybe we’ll just have to settle for a Kimberley process for Web3 claims, for both promoters and pundits alike.
Send to Kindle
Depending on how it is defined, a candidate for the “original blockchain” was the Haber and Stornetta timestamping system published in 1990 (and thrice cited in the Bitcoin whitepaper). Therefore archaic blockchains had a useful niche before Bitcoin but were not capable of moving assets without a third party. Note: as they failed to provide a definition of a “blockchain” in their letter, the authors overly broad usage of “not useful” could encompass e-signature providers such as DocuSign and HelloSign. [↩]
One of the authors, Kelsey Hightower, works at Google, and a couple others work for large tech companies partly reliant on adtech revenue [e.g., monetizing personal information and data.] [↩]
At one point Accenture proposed an “edit” feature that does not appear to have been adopted by any chain. Stellar has implemented a feature that allows developers to “burn an asset.” [↩]
The experiments in the “dangerous” public chain space are funding and battle testing some of the new privacy and tech stacks that ‘Big Banks’ were not incentivized to build. Two examples in the U.S.: the FTX clearing proposal might be a better “exchange stack” than existing traditional finance operations and the Silvergate banking API (SEN) quickly confirms transactions based on on-chain data. Both services might not have been built even in the private blockchain world; at least they have not thus far. [↩]
To be fair, a number of financial incumbents and non-blockchain-related fintechs market their products and services as “financial inclusion.” They all attend many of the same events and sit on the same panels too. [↩]
See also the proposed E-Cash Act co-authored by Rohan Grey. [↩]
Technically Mastercoin, Counterparty, and several colored coin projects launched before Ethereum did, but they did not include a virtual machine that can run arbitrary code. [↩]
For balance, traditional financial markets also facilitate the transfer of illicit funds (money laundering) and ill-gotten gains from scams and fraud. The authors would have a stronger argument if they provided actual stats, e.g., what percentage of on-chain transactions involved illicit activities. [↩]
For instance, this coming October, a tokenized pound (‘synthetic CBDC’) on a blockchain platform operated by Fnality International will go-live in the U.K. Uptake may be slow in part because of issues around composability and because initial participants are banks that need to change the way they make payments. AntChain from Alibaba is a production chain used to settle e-commerce payments (connecting their banks to their merchants). Another example would be “perpetuals” which were conceived by Robert Shiller in 1992 and first implemented in 2016 by Bitmex, and now widespread on many major CEXs and a few DEXs. [↩]
The Federal Reserve Board annually conducts stress tests of the U.S. financial system. Similar tests occur in other countries. Researchers at the IMF recently released a paper describing the underlying framework of GST. [↩]
U.S. legislators at the national level have failed at providing a comprehensive digital rights and privacy framework, as well as A.I. auditing guidance. These issues are arguably just as important and impactful as cryptocurrency-related topics. [↩]
[Note: this is part of a standalone document written by Martin Walker in late 2019. It has been edited and condensed as it provides important considerations surrounding the topic of stablecoins. For more context, be sure to read the accompanying Parasitic Stablecoins article.]
In spite of the relative immaturity of “Stablecoins” as both an asset class and as a form of financial sector technology, they has recently attracted a huge degree of attention from regulators, central banks, academia, the media and many parts of the financial sector. This attention has particularly intensified since the announcement by Facebook of its own stablecoin (Libra) on June 18, 2019.
Reportedly prompted by this, a joint committee was formed by central banks from the G7 group of major economies, the International Monetary Fund (IMF) and the Bank for International Settlements (BIS).1 This group reported its own findings, focusing on potential regulatory and economic impact in October 2019.2
Defining stablecoins can be challenging business because there are already a significant number of variations and some of the most discussed stablecoins are still in development. The most basic and broadest definition includes three main characteristics,
They are intended to perform at least two of the main characteristics of money, acting as a means of exchange and as a short-medium term store of value
They use some variant of Distributed Ledger Technology (DLT) to record and transfer ownership in a similar way to cryptocurrencies such as Bitcoin and Ether
They are intended to have a value that is relatively stable compare to major currencies.
While most research on stablecoins focuses on the economic and regulatory implications, the purpose of the this paper is to present an analysis of the practical implications for key processes such as payments and settlement, not to mention the potential impact on systems within financial institutions and overall financial market infrastructure. Stablecoins as both an asset class and to some extent a form of financial sector. Consequently they have challenges to adoption in terms of competing with the current world and interacting with it.
Stability and Collateral
The most straightforward step to create a form of digital currency that has a stable value is to peg its value to a financial asset with a stable value. Most stablecoins are pegged in value to a specific currency. Tether is pegged in value to the U.S. dollar on a one-to-one basis. Others are pegged (or proposed to be pegged) to a basket of currencies. Libra was originally proposed to be pegged in value to a basket consisting of the U.S. dollar, euro, yen, British pound and Singapore dollar. Other stablecoins attempt to achieve a higher degree of stability by pegging their value to a basket of assets, including cryptocurrencies, in the belief that diversification alone will achieve a higher degree of stability. Finally there are stablecoins pegged in value to commodities such as gold or oil. Claiming to have a pegged value does not (as is discussed below) mean a stablecoin is fully backed by funds in that currency.
Maintaining a peg is much harder than simply claiming a stablecoin has a value pegged to another asset or basket of assets.3 The degree of stability depends on
The type of reserves
The proportion of reserves relative to the amount of stablecoins issued
The nature of the issuer of the stablecoins
The legal structure including the protection of the reserves from the issuers creditors in the event of the issues default
Real or proposed stablecoins have reserves in one or more of the following types
Deposits in a commercial bank marketed as providing one-to-one back – this is the backing claimed by Tether, the Gemini Dollar, Pax and many others.4
Backed by the balance sheet of the issuer where the issuer is a bank. JPM Coin, at least based on initial news about the proposed stablecoin, would be supported by the balance sheet (i.e. the assets and capital of JPMorgan). From a credit and valuation perspective it should be broadly equivalent to funds deposited in a JPMorgan bank account.
Backed by a basket of bank accounts and other financial assets – According to the Libra whitepaper the stablecoin would be supported by assets held by the Libra foundation consisting of bank deposits and short term debt denominated in a basket currencies, subsequently announced as the U.S. dollar, euro, yen, British pound and Singapore dollar.5 Potentially the set of assets held by the Libra Foundation could include central bank reserves, subject to being allowed to open reserve accounts.
Stablecoins backed by a reserve of cryptocurrencies can be one of the most transparent ways of demonstrating the existence of a reserve. If created correctly holders would be able to check the balances of cryptocurrencies held by addresses relevant to the stablecoin. Unfortunately due the relatively high correlation of all major cryptocurrencies to each other means it is unlikely that the degree of diversification that could be obtained would provide much stability.
Algorithmic stablecoins such as the proposed, “Basis” Coin are intended to be a form of currency that had stable value but which was not fully collateralised. The plan for Basis was for it to be partially collateralised but to use an algorithm to maintain stability by buying or selling the coin in the market. The problem with a “currency” created like this is that it creates the incentive to short the asset, perhaps one of the reasons Basis was abandoned.
It is easy to claim a stablecoin is pegged to the value of an established currency and is backed by reserves is not by itself, it is another matter to maintain a stable value for a stablecoin some of which, such as USDT, experience periods of extreme instability.
Central Banks could potentially issue a form of electronic money that had the same economic characteristics as physical cash or central bank reserves. This is typically referred to as Central Bank Digital Currency (CBDC). CBDC could be issued on some form of DLT (making it a form of stablecoin) or a centralised system. While there have been experiments by central banks with central bank money issued on distributed ledgers, no central bank has announced plans to create a “stablecoin.” The People’s Bank of China has been developing the concept of a form of using digital cash (potentially using DLT) for five years but nothing is in production yet. As of late-2019 the closest thing to a real world CBDC system was Ecuador’s failed attempt, the Dinero Electrónico, which was launched in 2015 and closed in 2018.6
Other relevant issues to maintaining the stability, or even basic credibility of stablecoin relate to legal and operational issues. If the issuer of a stablecoin fails, the assets ideally should be in a legal structure that is “bankruptcy remote” (i.e. the holders of the coins can claim the reserves in preference to other creditors of the issuer). The bankruptcy remoteness of the Libra foundation, or even the general recourse Libra holders would have to the reserves of the Libra foundation are currently unclear. For the stablecoins used in cryptocurrency trading such as Tether and the Gemini Dollar there are varying degrees of bankruptcy remoteness. In the USC model, Fnality funds would be set up in a bankruptcy remote structure. JPM Coin (or almost any commercial bank issued stablecoin) is supported by the overall balance sheet of the bank. Holders of JPM Coins would most likely be treated like any other bank account holder.
For any stablecoin to remain truly stable it would need an issuer willing to buy and sell the stablecoin at par, or a very small spread above and below par. Even existing stablecoins with better controls that Tether such as Pax or the Gemini Dollar shows significant fluctuations in price. Convertibility on demand causes challenges for stablecoins, it would increase the probability in most jurisdictions that issuers would need to treat the owners of their coins as their customers for AML/KYC purposes. It would also cause challenges in terms of liquidity management. According to the Libra white paper, only specified liquidity providers will be able to buy and sell Libra directly with the Libra foundation. Other holders of Libra will not be able to redeem their Libra directly. JPM Coins will simply be transferred to or from client’s existing J.P. Morgan bank accounts.
Auditing of the reserves and the controls that are put in place to ensure the reserves are segregated from the issuers other liabilities is another fundamental feature required for maintaining price stability against the assets pegged against. One of the major reasons for the volatility of Tether was the lack of a recognised audit of their reserves and the worry, subsequently proved to be correct, that the Tether was not fully backed by reserves held as bank balances.7
Payment and Settlement Processes
For stablecoins to be effective as a part of conventional Financial Market Infrastructure as opposed to just being a tool to support cryptocurrency trading, they need to support the following fundamental processes that involve the transfer of money, either one way movement or synchronised with the movement of money or securities in the other direction.
Payments in between two parties in the same jurisdiction in the local currency.
International payments typically involve a foreign exchange transaction as the sender’s home currency is converted in the recipient’s home currency. In many cases such as cross-border payments within the Eurozone there may be no need for a foreign exchange transaction.
Delivery versus Payment is the synchronised exchange of a security for cash. DVP is used in both the settlement of purchase/sale of securities and the temporary exchange of cash in securities in areas such as Repo and Securities Lending. In conventional financial markets. Currently DVP requires the use of a trusted third parties such as a Central Securities Depository/Securities Settlement System e.g. DTCC or Euroclear or a custodian.
Payment versus Payment, is the synchronised exchange of two different currencies. PVP is used for the majority of transactions by volume in the foreign exchange payments using the services of CLS Bank.
Holding a financial asset on behalf of the economic owner of the asset. Custodians provide of a variety of services in addition to basic safe keeping of assets including, lending securities, financing long positions and dealing with corporate actions and events.
Temporarily delivering financial assets to another party to offset credit risk is fundamental part of the operation of most financial markets. Collateral in the form of money, securities or other financial assets may be delivered to the counterparty, a central bank, a tri-party agent or a CCP depending on the nature of the transaction
Novation is the transfer or contractual obligations and rights from one of the original parties to a contract to another party.
Domestic payments between customers within the same bank are always the most technically and operationally simple to process. Most banks should have little difficulty in processing payments in anything less than a few seconds and at minimal cost. Fundamentally all users of a particular stablecoin will essentially have an “account” at the same virtual bank, or in the case of JPM Coin or Signet, the same actual bank. Even if a bank has archaic batch-based or even paper-based solutions for internal transfers, using DLT is just one of many possible approaches to speeding up transfers.
Domestic payments between parties that bank use different banks is more considerably more complicated than payments within the same bank because of the need for banks to manage intra-day liquidity in order to avoid running out of the funds required to meet their liabilities.
However huge progress has been made in this area over the last two decades. Payments between parties that bank at different banks has been made close to instant in most developed countries through the implementation of low cost and efficient Real-Time Gross Settlement (RGTS) systems and internationally via initiatives such as SWIFT gpi.
Previously settlement of domestic payments was based on systems that used Deferred Net Settlement (DNS), basically settlement of payments was made at the end of the day after all payment instructions had been received and the net amount each bank owed each other was calculated. The existence of RTGS in over 90 countries has demonstrated that making payments instantly and settling in central bank cash does not remove the problems of liquidity or even credit risk. Central Banks have found the need to implement additional measures to avoid problems resulting from the “lumpy” nature of payments flow between banks, stress conditions and banks passively releasing their own payments after receiving payments from other banks.8
To deal with these issues central banks introduced a variety of mechanisms including Liquidity Savings Mechanisms (LSM), which group together payments before releasing to get smoothing payment flow, targets for the proportion of payments released immediately, and lower fees for the releasing payments earlier during the day. Stablecoins, if they reach sufficient scale, would not get rid of any of these problems and it is likely they would have to replicate the same mechanisms. It is worth noting that as part of its experiment with DLT in domestic payments (Project Ubin) the Monetary Authority of Singapore implemented an LSM using DLT.9
While small scale international payments for many countries can take minutes, wholesale payments can still take days, particularly if they involve the settlement of a related foreign exchange transaction. Based on analysis by SWIFT some of the key sources of delays in international payments include, errors within the systems and processes of both the sending and recipient banks, the need to carry out checks for Anti-Money Laundering (AML) and combatting the financing of terrorism (CFT) and in some countries the operations of exchange controls.
Stablecoins do not innately solve any of these issues, particularly where the desired end result of a payment in a deposit in the bank account of the ultimate recipient in the appropriate currency. More retail-focused stablecoins such as Libra may simplify international payments if Libra is used to directly purchase goods and services. However, holders of Libra (assuming Libra is backed by assets in a basket of currencies) will be exposed to the market risk of fluctuations in exchange rates. It is also unclear what the costs will be on converting into and from Libra.
The need for payment-versus-payments is an essential need for wholesale FX trading, to avoid settlement or “Herstatt” risk. This is risk that one party to an FX transaction delivers the currency they have sold but the other party does not deliver the currency they are owed, for example due to bankruptcy.
PVP currently requires a trusted third-party to manage cash flows including the release of funds when both parties have delivered the required currency. The majority of foreign exchange transactions are settled through CLS Bank, which provides multilateral netting and connections to the RTGS systems of 17 central banks. On a typical day CLS settles $5 trillion of transactions. The ability to net settlements on a multilateral basis for over 90 of the world’s largest financial institutions allows CLS to reduce the net amount of funds that have to be transferred by 96%.10
The potential opportunity claimed for some stablecoins is the ability to implement a PVP mechanism without the need for having a third party involved and a shorter (if not instant) settlement cycle. The mechanisms required to support PVP using a stablecoin depend on where and how the two currencies are represented. Excluding cryptocurrency related stablecoins such as Tether or Pax, there are the following combinations.
Scenario 1 – Currency 1 and Currency 2 are stablecoins created by the same issuer
Scenario 2 – Currency 1 and Currency 2 are stablecoins created by different issuers
Scenario 3 – Currency 1 is a stablecoin and Currency 2 is a fiat currency
Scenario 1 – USC is currently planned for up to 5 currencies and also plans to have separate ledgers for each currency. Therefore to achieve PVP they would need to create smart contracts that operate on two ledgers simultaneously. Fnality plans to use an architecture called Ion produced by Clearmatics but this is still a work in progress.11 Ion is also planned to support PVP between different ledger technologies such as Ethereum and Hyperledger Fabric.
A more commonly discussed model for dealing with assets on different ledgers, potentially ledgers implemented using different DLT is the “Atomic Swap” where a smart contract on one ledger will only allow the transfer of funds if funds have been transferred on the other ledger:
“Atomic swaps solve this problem through the use of Hash Timelock Contracts (HTLC). As its name denotes, HTLC is a time-bound smart contract between parties that involves the generation of a cryptographic hash function, which can be verified between them. Atomic swaps require both parties to acknowledge receipt of funds within a specified timeframe using a cryptographic hash function. If one of the involved parties fails to confirm the transaction within the timeframe, then the entire transaction is voided, and funds are not exchanged.”12
All the proposed technical models for achieving PVP for ledger-based assets are in the early stages of development. In some proposed stablecoins the degree of centralisation of the stablecoin would make it easier to use an established technology and process design to achieve PVP.
Achieving PVP between a stablecoin and a conventional currency, without involving an intermediary is considerably more problematic. The nature of conventional forms of money mean they are inherently centralised either as a record at a commercial bank or a central bank. Possible models of interaction with existing payment infrastructure is described in the next section “Interaction with Current Financial Market Infrastructure.”
In terms of shortened settlement cycles, stablecoins used for PVP are likely to come into competition with services such as CLS Now, which allows same day settlement of FX transactions using PVP for Canadian dollars, Euros, Pound Sterling and US dollars.
Interest Charges and Payments
It is very easy in a low interest rate environment to forget stablecoins are likely to need some capacity for the payment and collection of interest on balances. This is a particularly strong requirement even now for stablecoins that are proposed to be based by central bank reserves.
For currencies (at time of writing) where the central bank has negative interest rates on balances in reserve accounts (for example the -0.5 % charged by the European Central Bank), it will be necessary to pass on the charge to the holders of stablecoins otherwise the issuer of stablecoins will rapidly become involvement. The issuer of the stablecoins (who holds the backing funds in a reserve account) will need to carefully track who held what balances for what time periods and charge relevant holder, deducting interest owed from balance in the stablecoin or be able to charge interest directly if there are insufficient balances in their stablecoin wallet to pay interest. This inherently introduces and element of credit risk.
Similarly, where a central bank pays interest on reserve accounts it will be necessary for interest to largely be paid on to the relevant stablecoin holder otherwise there is a major disincentive (even at low positive rates) for firms to hold balances in stablecoins for anything other than the shortest possible duration.
Delivery versus payment is the synchronised exchange of a security for cash. DVP is used in both the settlement of purchase/sale of securities and the temporary exchange of cash in securities in areas such as Repo and Securities Lending. Currently DVP requires the use of a trusted third parties such as a Central Securities Depository (e.g. DTCC or Euroclear or a custodian).
DVP presents many of the same challenges and opportunities as PVP. Three key scenarios would need to be dealt which are similar to the PVP scenarios.
Scenario 1 – Stablecoin and securities are both created by the same issuer that contains the same overall network but data is stored on different ledgers
Scenario 2 – Stablecoin and securities are recorded on different ledgers run by different organisations and potentially using different forms of DLT.
Scenario 3 – Stablecoins would need to be exchanged for securities where ownership is recorded on a central database controlled by a Central Securities Depository or a Share Registrar.
Neither Fnality, JP Morgan, nor Libra have currently announced plans to issue securities on the ledgers they are planning to build to support their stablecoins. This currently leaves only scenarios 2 & 3 as plausible short-term possibilities. Scenario 2 raises the same challenges described for PVP but assumes a significant number of securities would be available as securities that are initially issued as on a distributed ledger or are tokenised versions of conventional securities.
A “tokenised” security is one where the original security is “immobilised” i.e. held in trust by a third party such as custodian and economically and legally equivalent representation of the security is recorded on a Distributed Ledger. There is currently only a small number of securities either issued on distributed ledgers or tokenised. Those that have been issued are typically small scale pilots. Interacting with a CSD to achieve DVP is problematic for the same reasons as trying to achieve PVP between a stablecoin and a conventional asset.
In the existing financial world, financial assets are held in the name of a third party for a variety of reasons including security and the desire to gain access to the range of service offered by custodians. Custodians provide a range of services that go beyond simply safe keeping of assets. These include operating lending programmes for securities, lending funds against the security of assets held and the processing of corporate actions on securities.
Keeping cryptocurrencies and other crypto-assets with a third party has grown in popularity because of the inherent vulnerability to theft of that most cryptocurrencies and crypto-assets. Obtaining a private key is all that is necessary to transfer all the assets associated with that key.
It is nearly impossible to cancel ore reverse transactions if assets are stolen or even sent to the wrong party by mistake. This is a feature included in cryptocurrencies such as Bitcoin, by design. Reversing transactions in the event of crime or area depends on either law enforcement seizing the private keys or other parties co-operating to return assets (which may have costs) – This is due to lack of central control. Anyone can attempt to “fork” most blockchain based systems but this technical process which basically comes down to re-writing history and pretending certain events did not happen is dependent on the co-operation of a critical mass of infrastructure providers called “miners.” And the loss of the private key means the assets are essentially gone for ever and impossible to retrieve.
Custody of most crypto-assets means handing over the private keys to a third party and attempting to ensure that private key is not used by the third party or their staff to steal. With some custodians, private keys are printed on paper and kept in physical safes. Private keys are broken up into pieces and distributed across multiple systems. In the worst case this simply increases the risk of losing access to the crypto assets.
The need for this form of custody essentially depends on the extent to which a stablecoin is operated on a decentralised ledger. For Libra the extent of decentralisation is currently unclear. For Fnality and JPM Coin the high degree of central issuance makes it unlikely that cryptocurrency type of custody would be required. It is likely that organisations wishing to hold wholesale forms of stablecoins may wish a third party to hold their balances in order to outsource the processing of stablecoin transactions, including payments, receipts and conversion to or from conventional currencies.
The final area of processing that stablecoins would need to support is the ability to give or receive them as collateral. Collateral is provided either to a counterparty or trusted third party such as tri-party agent or CCP to offset the credit risk arising from other financial transactions such as derivatives trades. In principal there should be no major issues providing stablecoins as long as the recipient has the technical infrastructure to process stablecoin transactions, value stablecoins and the ability to represent them correctly in systems such as their risk, finance, accounting and operational systems.
Inter-Operating and Competing with Existing Infrastructure
Stablecoins that are designed to appeal to a wider range of users than cryptocurrencies have to be capable of integration with existing financial market infrastructure. To be accepted by regulators they also need to comply with the appropriate regulations for each jurisdiction. This section describes the types of market infrastructure that will need to be integrated with and the challenges that creates.
The Challenge of Integration
One of the major and inherent weaknesses in the design of cryptocurrencies is the problems that arise when a new form of financial infrastructure is designed without giving any thought to how to integrate with existing infrastructure, whether in terms of market level infrastructure or internal to financial services firms.
The current cryptocurrency industry did not grow to its existing size by operating as a parallel payments and banking system that provides alternative ways to make payments or store value. It grew by throwing away the basic principles of decentralisation and disintermediation by recreating centralised systems (i.e., intermediaries) that kept a parallel record of cryptocurrency holdings to that stored on the ledgers of the relevant cryptocurrency. The repeated hacks, thefts, and other failings consistently demonstrated that this centralised infrastructure to support decentralised assets was seldom built with any regard to meeting the BIS Principles, or even in some cases local laws.13
Challenges to integration largely arise from the factors present in most forms of DLT:
Lack of central control over the operation of the system
Lack of central control of the deployment of changes to code
General inability to stop transactions
General inability to reverse transactions
Global visibility of all transactions
Owners not identifiable
Dependence on a cryptocurrency to pay for processing of transactions
Many of these features have been abandoned or worked around as the various forms of DLTs have evolved but to vary degree represent challenges both in terms of integration to FMI and the operation of the key processes related to settlements and payments. Sometimes to the point where it is questionable why a form of DLT makes any sense at call compared to conventional Centralised or Distributed Systems.
Forms of inter-operability
The following are the potential conventional forms of infrastructure that the next generation of proposed stablecoins will would potentially need to interact with.
In the following section we focus on the conventional types of financial market infrastructure described below that would be significantly impacted by the more widespread adoption of stablecoins.14
“A set of instruments, procedures, and rules for the transfer of funds between or among participants; the system includes the participants and the entity operating the arrangement.” This includes the various RTGS.
Central Securities Depositories (CSD)
“An entity that provides securities accounts, central safekeeping services, and asset services, which may include the administration of corporate actions and redemptions, and plays an important role in helping to ensure the integrity of securities issues (that is, ensure that securities are not accidentally or fraudulently created or destroyed or their details changed).”
Securities Settlement Systems (SSS)
“An entity that enables securities to be transferred and settled by book entry according to a set of predetermined multilateral rules. Such systems allow transfers of securities either free of payment or against payment.”
Central Counterparties (CCP)
“An entity that interposes itself between counterparties to contracts traded in one or more financial markets, becoming the buyer to every seller and the seller to every buyer and thereby ensuring the performance of open contracts.”
“CLS Bank (CLS) is a limited purpose bank for settling FX, based in New York with its main operations in London. It is owned by 69 financial institutions which are significant players in the FX market. It currently settles trades in 17 currencies. CLS removes principal risk by using PVP – you get paid only if you pay. On settlement day, each counterparty to the trade pays to CLS the currency it is selling – eg by using a correspondent bank, as with the example in the previous box. However, unlike the previous example, CLS pays out the bought currency only if the sold currency is received. In effect, CLS acts as a trusted third party in the settlement process.”16
Internal Financial System Infrastructure
The core internal infrastructure of banks and financial institutions. This includes systems used for risk management, P&L calculation, transaction execution and accounting
Payment Systems (PS)
Stablecoins that are backed (in whole or in part) by bank balances at commercial or central banks will need some degree of integration with a payments system or the payments infrastructure of a given bank. This will be necessary to process the receipt of funds that preceded the issuance of new coins, outgoing payments when there is a redemption and potentially payments or receipts of interest on stablecoin balances. A stablecoin based on central bank reserves would generally need to be connected to some of Real-Time Gross Settlement system to minimise delays in the issuance of new coins.
Such integration is generally straightforward assuming the issuer of stablecoins is allowed to access directly relevant payment systems. A more interesting question is the impact of stablecoins as a competitor to conventional payment systems.
A stablecoin denominated in a single currency needs to demonstrate it has some form of superiority in terms one or more of the following
Reduced Operational Risk
Ability make payment conditional on other parts of a financial transaction (as in the case of PVP, DVP etc.)
While at the same time dealing with the challenges of ensuring there is sufficient liquidity in the stablecoin network for parties to meet their obligations. The experience of introducing RTGS in over 90 central banks since 1980s demonstrated that allowing a pure system of gross settlements in payments, with participants free to release payments at any time can cause liquidity issues that need to be dealt with by technical changes, such as the implementation of Liquidity Saving Mechanisms (LSM).17
To quote the New York Fed:
“Liquidity-saving mechanisms (LSMs) are queuing arrangements for payments that operate alongside traditional real-time gross settlement (RTGS) systems. LSMs allow banks to condition the release of queued payments on the receipt of offsetting or partially offsetting payments;”18
Some central banks also created rules/targets for when payments should be released or financial incentives for early release of payments.
Retail focused stablecoins need to be able compete with faster payments, credit card and debit card networks. Payments mechanisms that can be highly efficient in many economies. In cross-border payments, stablecoins need to be able to demonstrate they are a more efficient mechanism for dealing with the major areas of delays and costs such as compliance with AML rules and in some markets exchange controls.
Central Securities Depositories & Securities Settlement Systems
For infrastructure such as Euroclear, DTCC, ASX’s CHESS system or Takasbank in Turkey to perform delivery versus payment, they need to have access to a security register to update ownership records and a funds belonging to participants, either held directly at the FMI or at a Central Bank. They also need to be able to provide trade capture, matching and netting capabilities.
For a stablecoin to be used in the DVP settlement, the FMI needs to be able to directly access stablecoin balances belonging to the participants in trades, either directly or on behalf of participants by a third party. This would require work by both the CSD/SSS and the stablecoin provider. There is no obvious benefit from this arrangement. Other systems belonging to the CSD/SSS would need to be modified to represent what is effectively a new currency. For countries that have long established infrastructure in this area, it quite likely adding an additional currency would require additional effort.
Central Counterparties (CCP)
Any organisation likely to acquire large balances in high quality stablecoins is likely to want to be able to provide those stablecoins as collateral in bilateral transactions, with central banks and with Central Counterparties. Should CCPs chose to accept stablecoins they would need to make significant changes to their systems to interact with the relevant distributed ledgers and set them up as new currencies or asset classes within their systems.
CLS is such a fundamental part of the global financial market infrastructure that any stablecoin that is used on a very large scale is likely to need some degree of integration. It should be remembered though that the vast majority of currencies (by number not importance) are not supported by CLS. Potentially a stablecoin could be added as another CLS currency allowing the benefits of multilateral netting and integration into the core global FX processing. However it would depend on a very high degree of demand and a many regulatory approvals.
In many ways stablecoins compete directly with existing CLS services so it is also questionable the extent to which CLS may support their adoption.
Interoperability with Financial Sector Internal Systems
There are two main areas where the internal systems of financial sector firms would require modification. Their outward facing interfaces that would need to interact with a range of distributed ledgers (unless they outsource this interaction to third parties – essentially creating a new class of correspondent bank) and modifications to inward facing systems such as those belonging to the risk, finance, trading, operations and treasury departments. Perhaps the closet analogy was the creation of an offshore version of the Chinese Yuan, commonly known as “CNH.” Though no wholescale re-engineering was required, it did commonly require changes to be made across a great many systems to recognise the difference between CNH and the on-shore version of the Yuan, “CNY.” This had a particularly large impact for those banks offering services in CNH.
Having two versions (or more) of essentially the same currency creates a great deal of scope for confusion in trading, treasury and support processes. Subtle differences in liquidity and conversion costs also mean that the different versions of the same currency have to be treated differently in many different ways including charges, interest rates and the curves used in pricing positions.
Interoperability with Distributed Ledger Based Infrastructure
Interoperability with emerging infrastructure based on DLT is also likely to create a number of challenges.
Some forms of Market Infrastructure in-progress (or beta) such as ASX’s CHESS system (for securities settlement) and the HQLA-X system for exchange of High-Quality Liquid Assets for lower grade assets are essentially centralised systems that use elements of DLT as part of the overall system design. Interfaces would need to be like any other form of FMI. Those interfaces would need to take into consideration security, privacy and the need for agreed data standards. There would also be the complications of adding what is effectively a new currency.
One of the proposed methods of allowing interaction between different types of ledger or even different instances of the same DLT but recording different assets or used by different parties, is the Atomic Swap. Using this method, funds on the two different ledgers are only released when both parties acknowledge that assets have been transferred. If the two acknowledgements are not received within the agreed time, the assets will be transferred back to the original addresses.
Atomic Swaps are still an emerging technology that have been widely tested in cryptocurrencies. However on a theoretical level they raise governance issues. If assets are on ledgers ultimately controlled by two different parties, whose has governance over the transaction? It also provides an element of optionality to each party to change their mind about whether to go ahead with the transaction. They could simply not deliver and have their asset returned to them. There are similar problems in the current world. Some counterparties have high rates of settlement failure on securities related trades because of issues in their operational processes or systems. Others at times have financial incentives to allow trades to fail, which had created significant problems in the operation of the Repo market.19 This has resulted in stricter rules and fines in many jurisdictions.
Creating stablecoins as forms of either financial market infrastructure (i.e. used by multiple financial bodies in the case of USC or as essentially internal systems, as is in the case of JPM Coin, Wells Fargo Digital Cash or Signet) clearly does not require the use of any form of Distributed Ledger Technology. Most of the use cases ultimately involve some form of book transfer of funds within essentially the same systems. Allowing customers of the same bank to transfer funds between each other in real-time 24*7 at little to no cost is a service provided by many banks today. The only bottleneck to allowing this in other banks is either a lack of willingness to provide the service or the use of antiquated systems that rely on batch processing.
Liquidity issues out of hours
At the market level, real-time payments within a currency bloc, that settle in central bank money have been implemented using Real-Time Gross Settlement Systems in over 90 countries to date. Some of those payments systems such as the Eurosystem’s TARGET2 have been extended to support securities settlement (T2S) and smaller scale instant payments (TIPS). For the cross-border market CLS connects together the RTGS of 17 currencies to allow PVP settlement against central bank reserves.
The challenges faced in creating creditable stablecoins that can grow beyond simply supporting speculation in cryptocurrency trading are large. Stablecoins backed by Central Bank reserves require the explicit backing of the relevant central banks. Stablecoins such as Libra have attracted extreme scrutiny if not outright opposition from Central Banks and politicians both because of concerns over the stability of the financial system and a lack of trust by some politicians in Facebook as an organisation. Any stablecoin that is regarded as a key part of Financial Market Infrastructure is likely to be required to meet strict regulatory controls, reflecting the principles laid out by the BIS in “Principles for financial market infrastructures.”
Creating interoperability between the infrastructure on which stablecoins operate and existing infrastructure, not to mention potential future infrastructure that runs on different versions of DLT is a non-trivial task, not made any easier by the use of DLT. Such interoperability will be vital if stablecoins ever hope to be anything more than parallel RTGS systems.
Finally the obstacles that have been encountered by RTGS in managing liquidity are unlike to be removed by the use of DLT. In many countries the introduction of RTGS, identified the need to create mechanisms to ensure firms did not hold back payments, creating intra-day funding needs, intra-day credit risk and general systemic risk. If is very likely that if stablecoins were used in a significant volume of transactions there would be a need to introduce many of those measure described that had to introduced for RTGS such as Liquidity Savings Mechanisms.
Then there are the challenges with DLT. None of the various forms of DLT have proven themselves at scale and in a regulated environment and it is questionable whether they a better form of technology, even for implementing stablecoins that existing technologies.
Stablecoins may succeed in the long-run if they can demonstrate an ability to support better ways to manage liquidity including broader, if not continuous, settlement cycles for both money and securities. Finally one of the key concepts between more advanced forms of DLT such as Ethereum or Fabric was to allow parties to agree bilaterally or in groups to deploy agreed business logic in the form of “Smart Contracts” that can be executed when transactions are processed. This type of flexibility could be a potential path to the a higher degree of standardisation in processing financial transactions without the need to have a central, and inherently slow moving body, setting standards for a whole area of business or jurisdiction.