Cooper replies with a look that says he gets this question a lot, not least from the consortium’s members, mainly banks as well as insurance companies and other financial institutions.
His response involves both the foundational nature of R3’s work, and its mandate to deliver the ledger without involving crypto-currencies – although he adds the consortium expects to have a full enterprise version 1.0 of Corda ready by the end of the year.
First, the team has pivoted away from focusing on apps to an emphasis on the underlying operating system.
“A product built to address a specific use is like an app on your iPhone,” he says, cautioning that his engineers regard the analogy to be simplistic, but it gets the gist across. “You have to build the phone before the app can be used. Without a tech stack beneath, the iPhone would be just an incremental change over other smartphones.”
For R3, its distributed ledger, Corda, is the operating system plus a middle layer of core components that allow the ledger to interact with outside software and perform basic functions such as identifying users, transferring value on the network, and registering a trade or an asset.
On top, there is the user-facing version of Corda, which is akin to an app store, enabling things that banks actually want to do, such as facilitating trade finance, conducting know-your-client processes or reconciling an interest-rate swap.
Corda is about the operating system and interoperability. Unlike the blockchain solutions being developed by firms such as the Depository Trust and Clearing Corporation or the Australian Stock Exchange, which involve creating the full technology stack, Corda is leaving the app development to third parties, be it the banks themselves or vendors such as Calypso or Thomson Reuters.
Changing metaphors, Cooper suggests the better reference for R3 is Microsoft Windows, an ubiquitous operating system for all manner of enterprises.
“Initially, when David [Rutter, CEO] began the firm, we thought we’d focus on apps, but we couldn’t find a foundational distributed-ledger technology fit for purpose: our members didn’t think the solutions were ready,” Cooper said. “So we built the code from scratch” and partner with specialists for the apps.
The firm, in consultation with its members, decided to focus on building the distributed-ledger technology. But this is why Cooper and his colleagues continue to field questions about use cases: “The technologists at the banks get it, but sometimes the P&L folks do not, because they don’t see an immediate return to support a budget…we’re not focused solely on cases, but on building the foundation.”
No crypto-currency, please
The second reason why R3 prefers not to jump into a discussion around use cases is its mandate from members to develop distributed-ledger technology in isolation from digital currencies.
“Our clients said that for this to be scalable and deployable, they can’t rely on a virtual currency,” Cooper said. That meant building Corda to support smart contracts and a protocol for transactions that didn’t involve a crypto-currency or a digital token bespoke to the network.
This is distinct from the original blockchain that was invented to facilitate bitcoin, or the digital tokens underpinning Ethereum, Ripple and other DLT.
However, at the same time, Corda is being developed in order to support the advent of central banks issuing digital currencies – drawing a distinction between handling currencies issued by third parties, as opposed to generating a Corda-currency to facilitate network operations.
“We’re not replacing the status quo and everything must comply with existing regulation,” said Cooper, who is responsible for R3’s legal and government relations.
Technically, therefore, Corda isn’t a blockchain: there’s no underlying currency it mines and distributes, and validating transactions is not based on proof-of-work consensus like Bitcoin, or in other DLT under Hyperledger or Enterprise Ethereum Alliance. The reason is privacy. With a Bitcoin blockchain, consensus would require the entire network to see every transaction in order to validate the blocks being mined. On Corda, only a narrow circle of members, their counterparties, and trusted third parties (such as an exchange, a clearing house, or a regulator) can see a transaction.
Permissions for privacy
Moreover, to augment security, validation will require 100% consensus among permissioned players, rather than a simple majority of a broad network.
This arrangement will still enable Corda’s DLT to eliminate middle and back offices for reconciliation and confirmations. It also allows transactions to be reversed if all parties agree.
“You can’t have confidential and private banking in a pure blockchain,” Cooper said. In addition, whereas the consensus method on blockchain has no KYC – anyone on the network can create or receive value through a block – the Corda version requires stringent legal bona-fides for anyone to secure a node, while Corda’s technology maintains the system.
But this emphasis on legal permissions is a big reason why R3’s development has been slow, and why it takes longer and costs more to integrate its apps into a member’s existing systems.
As Corda nears its debut, however, R3 is also changing its business model. It launched with a membership model, providing advisory services to its banks, and charging them fees for ownership of the intellectual property and a seat at the development table.
But earlier this year, R3 completed a $107 million Series A fundraising, at which point members could turn into equity shareholders; and five large banks also dropped out.
Now over 40 banks are shareholders; others continue to pay consulting fees but Cooper says they will probably turn into clients, and R3 into a more traditional vendor selling enterprise software and integration services.