The BIS Innovation Hub Hong Kong published its report on the MBridge Pilot, a multi-CBDC initiative for cross border payments involving the central banks of China, Hong Kong, Thailand and UAE. Twenty commercial banks were involved and another six central banks participated as observers of the central bank digital currency (CBDC) project.
During the pilot, which ran for five weeks to 23 September, $12 million in currency was issued and $22 million in trade transactions were executed.
Two of the highlights are using a custom-built permissioned blockchain and privacy.
The cross border rationale
It’s widely known that there are considerable frictions in cross border payments which are estimated to cost $120 billion in transaction fees annually, apart from the indirect costs of delayed payments.
Emerging markets tend to bear the brunt of the friction. The correspondent banking system has contracted since the last financial crisis, reducing the choice of intermediaries for making payments. Also, emerging market countries often transact in foreign currencies adding to the need for intermediaries.
MBridge is on a path to production
Circling back to MBridge, following the pilot, it is now being developed into a minimal viable product ahead of a production deployment.
During the pilot, more transactions were done via the People’s Bank of China, in part because Hong and China started the pilot earlier than the other central banks. And also because China was the only bank that automated the integration of CBDC issuance and redemption with its domestic CBDC system.
How it works
The solution uses a custom shared permissioned blockchain, which it boasts was developed by central banks for central banks. At the core are the four central banks that provide issuance and redemption of the CBDCs.
Commercial banks can request the issuance of CBDC against reserves in their local currency at their local central bank and also redemption. Additionally, they can hold foreign CBDCs. Commercial banks can initiate peer-to-peer push payments for trade in any platform digital currency. And they could also execute FX PvP with other commercial banks.
The FX transactions proved less popular for a couple of reasons. In part because some central banks required them to offload foreign CBDC by close of day. Additionally, there was a mismatch of RTGS hours, FX rates were determined off-bridge, and a need for more liquidity. These are all factors that will be addressed during the next phase.
Custom blockchain and privacy
The report contrasts the custom-built distributed ledger (DLT) with “other multi-CBDC projects, in which the underlying technology was built by non-central bank entities.” In fairness, the MBridge team didn’t build a blockchain entirely from scratch. It uses Ethereum’s Solidity smart contract language, the Ethereum Virtual Machine, and HotStuff+ for consensus. In a bit of irony, Hotstuff’s main claim to fame was as the consensus mechanism chosen by the Libra/Diem stablecoin project that was nixed following pushback by regulators.
In terms of privacy, a central bank can view all commercial bank transactions in which its local bank participates. For some central banks this is a significant expansion in the visibility of day-to-day transactions with potential privacy issues not mentioned in the report.
The permissioned nature of the ledger means each commercial bank only gets to see its own transactions. And sensitive data is stored off the ledger. For the next phase, zero knowledge proofs are being explored “to enforce stronger privacy against arbitrary central bank validators.”
Of the 20 commercial banks, several were foreign branches of the same banks. Hence the high profile banks involved were the five biggest Chinese state banks and HSBC and Standard Chartered.
Observer central banks included those from the Philippines, Malaysia, Indonesia, Korea, Sweden, Israel. Staff from the London BIS Innovation Hub also observed, as did representatives from the innovation center at the NY Federal Reserve Bank of New York.
Both multi-CBDC projects highlighted the policy challenges of allowing foreign banks to hold a CBDC when central bank money is usually only available to locally supervised banks.
A key issue not yet addressed by MBridge is how to support a foreign currency not represented by the participant central banks. The US dollar is the most obvious one.