Banks takeover the blockchain

If 2015 was the year of peak hype around the blockchain, when banks first woke up to its potentially transformative power, then 2016 is the year of busy experimentation with shared-ledger technology and testing of commercial applications that might eventually come into use in 2017 and, more likely, 2018.

The blockchain is an extraordinary breakthrough in computer science that holds out the prospect of market participants owning shared databases of financial transactions, with new entries confirmed by consensus and protected by encryption, and containing immutable golden stores of agreed data and time-stamped transaction history. It holds out the prospect of a new approach to payments, transfer of securities and other assets, identity verification and regulatory reporting that could entirely reconfigure finance.

The global financial system has rested for centuries on interlocking networks of trusted third parties that handle their customers’ money and financial assets for them: banks to manage cash accounts and transfer payments; central banks to back these commercial banks and hold reserves to manage temporary imbalances between them; custodians and clearing counterparties to do the same for equities, bonds and derivatives. 

The development of the blockchain as the rails on which to move bitcoin, a revolutionary new censorship-resistant cryptocurrency, suggests that, rather than trusting third parties to intermediate payments for them, banks’ customers might be able to transact themselves directly with each other on a digital network that is itself trustworthy, even if some of the participants on it might not be.

This presents commercial banks, custodians, depositories and even central banks with a clear threat of eventual disintermediation. Remove the need for trusted third parties and what are banks for?

But it also carries a promise of greater efficiency and reduced cost. Participants in the financial system have always each managed their own separate ledgers, databases of asset ownership and transactions that they each control, that only they can modify and that are password protected. Each bank runs many thousands of them. 

The blockchain holds out the prospect of mutualizing cost through shared databases, secured by encryption that might at the very least remove the hefty annual expense of data duplication, data gaps and of reconciling inconsistencies between separate databases. 

This does not sound anywhere near as exciting as transforming the global payments mechanism, unless you happen to be an investor in bank stock or a bank executive.

If it does nothing else, just by making digital ledgers more efficient, the blockchain could reduce operating costs for the global banking industry substantially, cutting banks’ infrastructure costs attributable just to cross-border payments, securities trading and regulatory compliance by between $15 billion to $20 billion a year, according to one estimate from Santander. That is before billions more in potential savings on collateral management.

Others suggest the annual savings could be tens of billions of dollars more than that because there are indirect expenses and opportunity costs too. The database discrepancies inherent in the current financial architecture in turn require maintenance of costly capital and cash collateral cushions to protect against market breakdowns from disputed trades. If that capital could be freed up and churned more rapidly, banks would boost their returns on equity.

With those returns now so low, banks can hardly sniff at this. Charley Cooper, managing director at R3 CEV, which manages a consortium of 45 of the world’s largest banks working on shared ledger technology, says: “Banks in the consortium funding experimentation at a cost in the single-digit millions of dollars each could each secure operational savings measured in the hundreds of millions of dollars a year each, perhaps higher for the biggest banks. Commercializing shared-ledger technology could be the best investment these banks ever make.”

Blythe Masters, Digital Asset Holdings

Earlier this year Blythe Masters, the former JPMorgan banker who heads Digital Asset Holdings, spelled out at the Morgan Stanley financial services conference how blockchain is the ultimate example of what was once anathema to banks but which many are now desperate to implement – shared infrastructure. It can bring improved security through encryption, real-time regulatory reporting, but most important it can cut costs. 

“And we’re not talking five, 10 or 15% cuts in costs; we’re talking 30%, 40%, 50%,” according to Masters. “There’s only one way to do that and that is to share a mutualized common infrastructure that previously was kept separately and run independently by every market participant.”

The blockchain arrived last year like some mysterious millennial cult. It held out to a global banking system already beset by the cost of regulation, weak core profitability and threats from fintech disrupters the ultimate existential threat of systemic disintermediation. 

But it also held out the hope of new gains if the banks embraced it. Adding to the mystery, no one in the established banking industry really understood it. What have we learned in the last six months as banks have begun testing use cases?

Gideon Greenspan, chief executive of MultiChain, a provider to banks of a distributed database for multi-asset financial transactions within a private peer-to-peer network, notes: “To begin with, the first idea that we and many others started with, appears to be wrong. This idea, inspired by bitcoin directly, was that private blockchains or shared ledgers could be used to directly settle high volumes of payment and exchange transactions in the finance sector, using on-chain tokens to represent cash, stocks, bonds and more.”

Banks tell Euromoney that this is perfectly workable on a technical level. So what is the problem? 

“In a word, confidentiality,” says Greenspan. “If multiple institutions are using a shared ledger, then every institution sees every transaction on that ledger, even if they don’t immediately know the real-world identities of the parties involved. This turns out to be a huge issue, both in terms of regulation and the commercial realities of inter-bank competition.”

Gideon Greenspan, MultiChain

Greenspan tells Euromoney: “For now it’s not feasible for public blockchains, which must also work on home computers, to process the volume of transactions banks need to conduct. But more than that, there is just no way that any regulator would ever sanction a system where settlement ultimately depends on some pseudonymous miners in a distant geography.”

It is not much talked about openly, but concerns about state-sponsored cyber terrorism make public blockchains a nonstarter for global regulators and law enforcement agencies. “For the banks, public blockchains are not relevant for the next 10 to 15 years at least,” says Greenspan.

Or, as one banker testing use cases around custody tells Euromoney, when asked if banks are collectively hobbling adoption of the blockchain to protect their vested interests: “Let me get this straight. Some computer scientist sitting in his garage in Palo fucking Alto with $1 million of venture capital tells you that his new distributed ledger is going to put the DTCC out of business in six months, JPMorgan the year after and that the Federal Reserve might not have much longer. And you believed him?”

Huw van Steenis, banks analyst at Morgan Stanley, recently researched the potential for the blockchain to either disintermediate banks or bring big expense reductions through mutualization of market infrastructure utilities. He suggests proof of concepts will continue into 2018, with the first shared infrastructure perhaps just starting to emerge next year and gathering pace up to 2020. Van Steenis says: “We see no impact on bank earnings for at least the next two years. The challenges to make this technology work in a resilient way acceptable to regulators are not exactly trivial.”

Van Steenis tells Euromoney that much investment in the technology is motivated by banks’ instinct for self-preservation. “If you look at post-trade settlement, it is no wonder that the banks that have spent the most and taken the lead for example in the HyperLedger project are the ones that may have the most to lose if something emerges that is truly disruptive.”

He does not see anything much emerging soon. Van Steenis says: “No policymaker will even countenance an unpermissioned blockchain. And so while many of the banks are genuinely worried where this technology might eventually take the industry, the banks may also feel they have a moat protecting them.” He says: “I don’t think it is going to be disruptive in the next three years because the blockchain will emerge into application among closed networks of financial institutions, with a few participants that won’t let new competitors into the system.”

So the question now becomes how much do banks invest in the blockchain at a time when they are not very profitable and the return on that investment is not clear? Van Steenis sees a prosaic challenge in recent weak results. “This is going to be a very tough year for bank earnings. And if all budgets including IT have to be trimmed, then blue-sky investing in change-the-world technology might not be top of the list of projects to be maintained.”

But if the stick of disintermediation appears, for now at least, to be held at bay, the carrot of lower operating costs is strongly tempting. With banking revenues clearly ex-growth, the only way for management teams to improve margins is to invest in cost reduction. 

Jurgen Vroegh, ING

Jurgen Vroegh, global head of payments at ING, has a group of 25 to 30 people working on blockchain use cases, including through the consortium of 45 banks working with R3 CEV and separately on ING’s own initiative. In a business with over 50,000 employees, the efforts of this small group are high profile. “I can tell you that the board of directors of ING is very keen to learn about and is directly involved in what we are doing on blockchain,” Vroegh tells Euromoney. 

Last November, ING together with a small handful of other large Dutch banks conducted tests in exchanging value between them on blockchain. Vroegh says: “We learned first that if we all wanted to do this, we could make it work technically, though of course it would need the participation of central banks because this would be a virtual version of real currency. For our tests, each bank was a node on the network and we also created a central bank node, though the central bank itself did not take part. 

“The second thing we learned is that each test throws up more questions in turn: around resiliency, security, scalability, processing speed and governance. There are more questions than answers right now, and we continue to experiment. A key question for the regulators is identity. We recently ran a pilot test with the tax authorities on using bank credentials to verify identities for tax returns.” 

He says: “It may be that our experiments on payments eventually show the need for different kinds of oversight and regulation than we have right now.”

In almost every discussion on blockchain there are certain clichés that are constantly repeated. One is the quote from Bill Gates that in assessing the impact of new technology, markets tend to overestimate what progress should be made in two years and underestimate what can be achieved in 10.

“If we are now in the blockchain’s equivalent year to 1993 for the internet, there’s no point complaining that no one has yet created the Netflix of money,” says Peter Kirby, chief executive of Factom, a company building blockchain-based immutable, distributed ledgers for banks and other customers including governments. “Let’s just get on and create the email of money.” 

Another phrase often heard is that while blockchain is almost certainly the answer, the banks have not yet defined quite what the question is.

As banks continue to experiment on their own and in collaboration with each other, it seems that many are reaching the same conclusion about transparency – and how unhelpful it can be on the blockchain – that Greenspan identifies. 

Last year, banks stripped the blockchain away from bitcoin and decided to work on it separately from the cryptocurrency for which it was first built. Now they are decomposing the blockchain itself and searching for ways to keep the benefits of a shared ledger, protected and uncontested, while removing the ability of all participants to see every transaction even if the identities of parties are hidden.

Instead of waiting for the blockchain to reshape the banking industry, the banks are reshaping the blockchain.

The three most important projects underway testing the potential application of blockchain to wholesale financial services are: the collaborative efforts run by R3 CEV and 45 member banks to build a base layer-distributed ledger platform and then test commercial applications across multiple use cases to run on it; the project announced this January by the Australian Stock Exchange in conjunction with Digital Asset Holdings to explore whether shared ledger technology might replace the ageing Chess system for post-trade clearing and settlement of Australian cash equities; and the so-called HyperLedger project, also supported by Digital Asset Holdings, in which leading custodians such as JPMorgan, BNY Mellon, State Street plus central clearing counterparties and various exchanges aim to advance blockchain technology for recording and verifying transactions.

Cliff Richards, ASX

Cliff Richards, general manager of equity post-trade services at ASX (the Australian Stock Exchange), tells Euromoney the exchange is not an evangelist for distributed ledger technology. “ASX is being pragmatic in its assessment of distributed ledger technology (DLT). A key to this is asking ourselves if the problems we’re seeking to solve for customers and the market generally can be better solved with existing technology or, if not, whether DLT is most appropriate. 

“Once we decided to work with Digital Asset Holdings we set out to examine three broad aspects of the technology by building a meaningful prototype that will: first, address non-functional requirements, including through-put capacity, scalability, resiliency, privacy and security; second, test whether there is enough sophistication in the core business logic of distributed ledger technology to, at a bare minimum, replace what Chess already does and be robust and extensible; and third, explore blue-sky possibilities about what new opportunities the technology may open up – for example, capturing in encrypted form more granular data on the beneficial owners behind legal entity names in the settlement system.” 

In the blockchain world everyone wants an update from the ASX. On the timetable for making its choice, Richards says: “We are still in assessment phase and have not made any decision about whether to go forward with distributed ledger technology yet. However, we will inform the market by 2017 of our decision and how stakeholders will be engaged through any subsequent phases.” 

He tells Euromoney: “We are cautiously optimistic about the potential application of distributed ledger technology. We have enough time to continue to experiment and we have learned a lot already. We want to involve all interested stakeholders in the process.” 

And what has the exchange learned in the experiments it has undertaken since January?

Richards says: “There is a spectrum of configurations you can apply to distributed ledger technology. There is much discussion around a public blockchain verified with proof of work, versus private permissioned blockchains. The Australian equity market has a capitalization of A$1.6 trillion ($1.16 trillion). We need a resilient, scalable system that can process hundreds if not thousands of transactions per second, and we believe a public blockchain based on proof of work simply cannot handle that at present. In any case, operating in highly regulated markets, ASX could only ever countenance a private network, permissioned only to parties that already meet our and our regulator’s requirements.” 

While the ASX sees potential in a tamper-proof shared ledger for high-value assets, it is not enamoured of the notion that all participants should see everything. 

Richards explains: “Almost from day one, our strong hypothesis has been that only the regulators should be able to see everything. Entries will be encrypted and parties will have private keys, but they would only be able to use them to see full details of transactions to which they themselves are party.” He adds: “It might be possible for a party to confirm that another entity has claim to an asset of a certain value but not to identify that asset, effectively a zero-knowledge proof.” 

Blockchain enthusiasts in the banking industry will continue to watch events in Sydney closely. Post-trade settlement for all manner of assets – loans, mortgages, derivatives, collateral, corporate bonds – is one of the key use cases quickly identified last year as open to potentially big efficiency gains from blockchain. 

Those big gains may come from using blockchain to reduce clearing periods and settlement times in slow-moving markets where netting is difficult and collateral is tied up against counterparty performance risk for days and even weeks: markets such as corporate and mortgage whole loans. And here the ASX is conducting a real-time experiment in one of the first cash equities markets in the world to have been fully dematerialized, one that is reasonably well vertically integrated with self-contained banks, brokers, asset managers, exchange, clearing counterparty and regulator. 

The experiment is prompted by the collective decision to get ready to replace Chess as it approaches the end of its natural life. In the background lurks the question of how fast technology allows clearing to become and the balance between who gains and loses from speedier settlement. 

Huw van Steenis, Morgan Stanley

Van Steenis notes that the primary reason for multi-day settlement periods is regulatory, legal and market practice, which enables a broader participation by retail investors. “Current technology could deliver T+0 settlement today in a broad range of asset classes if regulatory and legal rules allow for it. You don’t need a blockchain to deliver T+0,” he argues. It may not even be desirable. “Markets with T+0 today appear to have less liquidity and more volatility than markets with a settlement window for several reasons, chief among them [being] that in a T+0 settlement window there is no shorting.”

Shorter settlement times might help investment banks that want to turn over their now much smaller amounts of capital as rapidly as possible but hurt custodians that benefit from carry on cash and collateral.

Richards says: “Australia has joined much of Europe on the move from settlement in T+3 to T+2. While we have never said that we would move the whole market onto T+0, there may be a business opportunity to offer settlement options with durations shorter than T+2.” However, he adds: “It’s important to remember that many of our clients get considerable benefit for the operational and balance-sheet netting that comes with T+2 net settlement.” 

It would be intriguing if the blockchain allowed for different options on settlement time. 

For the ASX, as for every financial market participant experimenting with blockchains, there remain interesting questions about managing the boundaries and connections between any private, permissioned distributed ledger for recording changes of ownership in a specific asset class and associated systems, for example those handling payments. The Reserve Bank of Australia and the Australia Payments Clearing Association are working on a new payments facility due in 2017 that will enable any Australian citizen with an Australian bank account to transfer payments to another in real time, as fast as sending a text or email. It remains to be seen how the new equities post-trade clearing and settlement system will tie into that.

Richards returns to the potential of distributed ledger technology around portable identity. “It would seem likely that you could use distributed ledger technology and cryptography to inject considerable efficiency by mutualizing identity data that is for now stored in fragmented, siloed systems and enhance privacy and security characteristics. We have, for example, been talking to the tax authorities about how, if we could capture and use more granular information about beneficial ownership of Australian equities, it might be a basis to auto-populate tax returns.” 

Next year could be the game-changer: 2017 is the year in which Charley Cooper, managing director of R3 CEV, says the wholesale financial services industry will see the first deployment of commercial applications of blockchain on a limited scale. 

R3 CEV is doing two things. It is developing an underlying distributed ledger technology platform called Corda, first details of which emerged in April and which R3 CEV says is heavily inspired by and captures the benefits of blockchain systems and which choreographs workflows between firms without a central controller. However – rather like ASX – Corda has no unnecessary sharing of data. Only those parties with a legitimate need to know can see the data within a financial agreement.

“There will not be a single blockchain on which all wholesale financial markets run,” Cooper tells Euromoney. “Instead, rather like the internet, which is a series of inter-operable platforms communicating through TCP, there will be a number of interoperating distributed ledger platforms including Corda, Ethereum, Eris, Chain. These are already in operation, will be quite robust by 2017 and are learning to interoperate. All manner of commercial applications for the distributed ledger will run across this fabric layer. We are working with our member banks to develop a number of these, broadly relating to pre-trade information and post-trade processing, always aiming to fit the solutions to the actual requirements of our members.”

The pitch for R3 CEV, as a member-funded consortium, is that it can develop distributed ledger technology that fits the business requirements of a heavily regulated banking industry, rather than create something technically brilliant that banks cannot actually use.

Cooper describes the key founding principle of Corda – that not every piece of information should be distributed among all participants on a shared ledger – as “massively important”. He says: “The notion that every market participant gets to see every entry in the ledger simply does not work for the banking community. Why should any participant see details of transactions to which it is not a party, unless it is the regulator or an exchange or clearing member of a market?”

And there is also a very practical issue, says Cooper: “Sending an excess of information to third parties in encrypted form that they cannot even read, heavily burdens a shared ledger, impacts its scalability and may make conventional blockchains a nonstarter in most markets. Simply altering that one aspect of the shared ledger – not distributing every piece of data to every participant – could be a key turning point in adapting this technology to the real-world problems that financial institutions need to solve.”

Members of the blockchain venture capital community often talk sceptically about the governance of a banking consortium seeking to develop new technology solutions that could potentially damage the interests of certain members. One tells Euromoney: “Don’t expect R3 CEV to move fast or even to develop the killer blockchain app because that is likely to hurt the very banks funding it.” 

Partly to counteract this, the company has split banks into separate working groups, reasoning that no one bank should be working on each use case and that banks should not have much of a say in developing blockchain applications for markets they do not operate in. While some use cases being tested are of interest to all banks, such as payments and regulatory reporting, many aren’t: custodian banks looking at new ways to digitize transaction and ownership records, trade finance banks looking at that market, and so on. 

Charley Cooper, R3 CEV

Members of R3 CEV are working on many potential applications for the shared ledger, most intensely on: trade finance; smart contracts in fixed income, where the first large-scale tests have been done on commercial paper and credit default swaps; payments; regulatory reporting, especially relating to identity, know-your-customer and anti-money laundering; and reference data. 

“We have tested smart contracts in commercial paper and swaps, reasoning that the general logic and lessons learned might apply across many asset classes,” says Cooper. “We have also been looking at a whole suite of risk-management products. There is a massive opportunity to reduce operational risk and better manage credit, market and compliance risk.”

One of the biggest problems banks faced in the financial crisis was the inability of hundreds of internal databases to collate in an accurate and timely fashion aggregate exposures to market and counterparty credit risk. Today operational risk is the one that regulators have squarely in their sights, demanding banks hold yet more capital against it.

And while the sceptical venture capitalists may well be right that R3 CEV is an unlikely venue to produce blockchain applications that disintermediate the entire banking system outright, it might well produce real-world uses for the technology that can profoundly change how financial markets work.

Cooper, who previously worked at Deutsche Bank and State Street, says: “The reason why you see so many of the world’s largest banks coming together through R3 CEV is that they realize the development of this technology, which clearly has disruptive elements, is inevitable. They also understand that the power of this technology comes from the network effect, so the more users, the better for everyone. The banks can’t just walk away from it, refuse to invest in it and hope it withers on the vine. They can either get involved and help shape application of this transformative technology to their highly complex and heavily regulated markets, or they can sit on the sidelines and pray that whatever comes out doesn’t hurt them.

“Faced with that choice, most banks would like a seat at the table,” he says.

While central banks and industry regulators may be shy of participating in individual banks’ blockchain experiments for fear of being seen to favour commercial enterprises they oversee, they may be more willing to work with a large industry consortium. Cooper tells Euromoney: “We are in discussions right now on proposals with a small number of central banks, securities regulators and payments regulators to participate in experiments in our labs.”  

As banks toil away with their blockchain experiments, they should remember that they are not alone. Other industries are exploring uses for shared ledger too. And the extent to which they change the technology for tracking the provenance and ownership of assets – property, commodities, traded goods, cars – that might touch the banking system as security for loans or prompts for payments, means banks must keep up to speed.

One banker tells Euromoney: “We’re taking to auto-makers who are looking at ways the shared ledger could authenticate and time stamp not just changes of ownership but also when a car has been serviced, which parts have been replaced and when, all of which might impact the secondary market and residual values.”

Kirby at Factom has been working with banks and non-banks, including governments, on blockchain projects. “The consensus computing underpinning blockchain may be hard to do but it’s no wonder that everyone who even starts to understand it soon becomes very excited by it. When you boil it down, it’s a data management technology that allows multiple parties to agree to a single version of the truth, to time-stamp that, and then never need to take their haggles over who said what to whom and when back to some central authority in the middle of the system. It lets you trust the math, not the people,” he says.

In some ways banking is the least amenable industry to blockchain technology because it is built around central regulatory authorities at the core of each market. They ultimately have to bless market participants’ adoption of any new technology. And blockchain technology is, among other things, an existential threat to central authorities as well. The regulators are what blockchain computer scientists sometimes call the asshole in the middle of traditional database systems, the central authority to which subscribers appeal. The blockchain does away with the asshole in the middle.

The financial services industry also has a fixation on confidentiality and a reluctance to share data, as evidenced by the development of Corda and the work of ASX. It is hard to see it having much relevance to systems for trading high-value financial assets.

Greenspan at MultiChain says: “One of the areas where we see the most interest is lightweight financial systems, in which the problem of information leakage is less of a concern. This might imply participants exchanging low-value goods, such as loyalty points, or a few participants that are not competing directly with each other making occasional transactions, such as corporate pension plan sponsors exchanging assets with each other rather than via intermediaries.”

Blockchain venture capitalists are starting to look away from developed world financial markets. Greenspan says: “We have noticed a greater openness to blockchain in the developing world, for example in Asia, where privacy and confidentiality may be less of a concern than overcoming the potential for corrupt officials to tamper with and falsify title deeds and records of ownership.”

Factom has been working with the government of Honduras on a project to put its land registry on blockchain. It is also working with Chinese municipal authorities on using the shared ledger to manage smart cities’ data derived, for example, from traffic sensors and power producers.

Peter Kirby, Factom

“The blockchain may have been pioneered by bitcoin, but at its core it is essentially about moving data and record keeping, not about moving money,” says Kirby. To an extent, this is semantics. Banks are little more than vast collections of databases of financial transactions. That is why venture capitalists have thrown so much money at computer scientists to create blockchain services and products to sell into the vast banking market.

It is a natural step from using the blockchain to record changes in the land registry, to property titles and record keeping relating to mortgages and indeed any other collateral. “In 2008, the world learned the impact on global financial markets and the economy from a problem that began with terrible record keeping for US mortgages,” says Kirby. “Today, partly in response to the consequent regulation, a US mortgage can be 1,000 pages of legal documents covering tax forms, credit appraisals, title records, servicing history.”

Factom suggests that when Bank of America bought Countrywide, the biggest mortgage provider in America, it had to transfer 10 billion pages of legal documents. It was lousy scrutiny of those records that ultimately cost Bank of America $17 billion in legal settlements. Factom is developing a system for banks to enter each original mortgage document in digital form and each record of subsequent payments onto directory blocks, hashed for reference onto the blockchain and tied to business logic that flags up a missing credit appraisal or title document before allowing verification. It can also audit records to check if they have been backdated or tampered with.

Kirby says: “The record-keeping requirements imposed on banks after the crisis have almost frozen parts of the banking industry because the potential punishment for failure to comply is so great. But it turns out that compliance with regulatory record-keeping is just about collecting data and time-stamping it. A distributed ledger can help do that. We are not quite ready to announce anything yet, but we are working on very interesting partnerships to bring this to market. The data layer runs on the bitcoin blockchain now and in the future on Ethereum or Ripple.”

It is a big prize: data collection, record keeping and monitoring for every financial instrument used as collateral for banks to lend against. And banks are determined that they should be the recipients of it.