Project KhokhaCase Study

Increasing transaction volume and network resilience while maintaining confidentiality requirements for RTGS on Ethereum.

“Our goal with Project Khokha is to contribute to the global initiatives which assess the application and use cases of distributed ledger technology (DLT) through this collaborative effort piloted by theSouth African Reserve Bank (SARB) together with the national banking community.”

—Francois Groepe, Deputy Governor, South African Reserve Bank

For the banking industry, the cost of providing the utmost reliability, availability, and resilience against attacks or equipment failure is high. Additionally, reconciliation payments are inefficient and the burden of reporting on suspicious transactions falls on banks. The South

African Reserve Bank, in consortium with seven Commercial Banks, used Quorum, an enterprise grade implementation of Ethereum, in combination with the Istanbul Byzantine Fault Tolerance consensus mechanism, Pedersen commitments, and range proofs to process the typical daily volume of payments for the South African Reserve Bank — with full confidentiality and finality — in less than two hours.

Problem Statement:

How can central banks increase the resiliency of interbank payment systems while maintaining or reducing the overall cost of those systems? The South African Reserve Bank (SARB) wanted to replicate interbank clearing and settlements using a permissioned blockchain network to assess if the performance, scale, and confidentiality of payments in a real world simulation was possible with blockchain technology.

Solution:

In consortium with seven Commercial Banks, the SARB partnered with ConsenSys to build a proof of concept grounded in real world performance, confidentiality requirements, and diverse bank hardware on the blockchain without a single point of failure.

Goals Achieved:

1. 70,000 transactions in less than two hours achieving the transaction performance target.

2. 95% of block prorogation time in <1 sec and 99% propagation in <2 secs demonstrating that acceptable performance is achievable, despite the geographical distribution of the banks’ hardware.

3. Achieved privacy while meeting required transaction volumes. This was the first time that the IBFT consensus mechanism, Pedersen commitments, and range proofs for confidentiality were used with Quorum. Combined, these delivered a combination of scalability, resilience, confidentiality, and settlement finality.

3 Project Khokha Case Study

Problem Statement Story

Investigating more resilient and cost-effective wholesale payment systems

Interbank payments are critical for countries, and every bank has a mandate to protect the financial interests of their country. However, the cost of the required levels of resilience is high. If the central bank’s servers go down, regional banks would not be able to pay one another, and the country’s payment clearing system would collapse. To prevent such a scenario, central banks invest substantial capital in infrastructure and security. Blockchain technology now presents a more secure, efficient, and cost-effective alternative. If every bank were on blockchain technology, there would be no single point of failure: a country could maintain its interbank payment network even if one or several [nodes] were down.

Bank reconciliation in and of itself is an inefficient process. When Bank A needs to

make a payment, it sends a message to the Central Bank to make one on its behalf. The Central Bank then sends a message back to Bank A saying they’ve done so, and notifies the receiver, Bank B, that they’ve received money from Bank A. Bank A needs a mirror account to keep a record of their holdings at the Central Bank and continuously reconcile transactions against that balance. This single process takes four transactions to move money from one account to another. With a blockchain, everything happens on a distributed network. Since all banks are on the same ledger, they can pay one another directly, which simplifies the process and greatly decreases the odds of double spending, fraud, and network vulnerability to name a few.

Regulatory reporting is another burden that lays with banks. All banks need to report suspicious transactions and transaction details. With blockchain technology, they now have a distributed ledger that contains all transactions, with details of who paid whom, when, and how much. Banks can easily report

4Project Khokha Case Study

on transactions as they are all on the same ledger; regulators can pull data from a single source of truth with near real time reporting.

SAMOS is an interbank settlement system that was created on March 9, 1998 by SARB. The South African central bank owns and manages the system. The SARB initiated Project Khokha in the latter part of 2017 to trial interbank wholesale settlement using Distributed Ledger Technology (DLT). The SARB engaged ConsenSys as a technical partner on the project and worked with a consortium of banks made up of Absa, Capitec, Discovery Bank, FirstRand, Investec, Nedbank, and Standard Bank.

ConsenSys was brought in to build on their experience with the Monetary Authority of Singapore (MAS). ConsenSys played a key role in Project Ubin, an effort to digitise the country’s currency on the Ethereum blockchain, initiated by MAS. ConsenSys’ role as architect and technology partner on the Quorum workstream for Phase 2 of Project Ubin was leveraged in Project Khokha and enabled this project to be established quickly.

Solution Story

Building a smarter solution with blockchain

There are three ‘waves’ of work for this type of initiative. Wave one is proof-of-concept on the Ethereum blockchain. Wave two takes the previous work and investigates the interconnected issues of scalability, resilience, confidentiality, and finality. Project Khokha forms part of this second wave and was designed to provide a realistic test of a DLT-based wholesale payments system. In particular, it investigated whether confidentiality can be achieved at scale. Wave

three is just beginning, and diverges into areas such as securities settlement and cross-border transactions.

This was the first time Quorum was used with the Istanbul Byzantine Fault Tolerance (IBFT) and Pedersen commitments to create a distributed ledger between participating banks. The tokenization of the South African Rand (ZAR) enabled a domestic payment system allowing participating banks to pledge, redeem, and track balances on the blockchain without a centralized central bank system—that is, without a single point of failure. This ensured that the solution provided robust confidentiality while enabling the transaction throughput required.

Project Khokha had four phases. Under phases three and four, know your customer (KYC) and anti-money laundering (AML) payment information could be shared between sending and receiving banks, as well as the Financial Intelligence Center (FIC), by enabling them to decrypt the relevant parts of the transaction.

5 Project Khokha Case Study

The Istanbul Byzantine Fault Tolerance:

A consensus mechanism which is an alternative to Proof of Work in an Ethereum network. Like other consensus algorithms, IBFT ensures a single, agreed-upon ordering for transactions in the blockchain, but also provides added benefits for enterprises, including increased throughput and settlement finality.

Whisper for private messaging:

A secure peer-to-peer messaging system built into Ethereum. At the start-up of the network, each bank exchanges encryption keys with every other bank via this channel and these keys are then used in transactions between those banks. The SARB has a copy of all the keys so that it can view all the transactions.

1 2

The Project Khokha Quorum solution used several finality, privacy, and confidentiality mechanisms:

6Project Khokha Case Study

Pedersen commitments:

Shown to be effective in ensuring full confidentiality of transaction amounts. Pedersen commitments allow the sender of the payment to commit to a value without being able to change it once the commitment is provided. Pedersen commitments are perfectly hiding and computationally binding under the discrete logarithm assumption.

Range proofs:

A type of zero-knowledge proof that enables one party to prove that a secret value (like one contained in a Pedersen commitment) is within a certain range.

3 4

7 Project Khokha Case Study

Goals Achieved

Pushing new boundaries to achieve privacy with required transaction volumes.

Project Khokha represents a significant contribution to the global DLT body of knowledge, as this was the first time that the IBFT consensus mechanism and Pedersen commitments for confidentiality were used with Quorum. Results show that the typical daily volume of the South African payments system could be processed in less than two hours with full confidentiality of transactions. IBFT, Pedersen Commitments, and range proofs delivered a combination of scalability, resilience, confidentiality, and finality.

Project Khokha, implemented on Quorum, demonstrated the practical test of a network across diverse bank hardware, pushing new boundaries in terms of confidentiality with performance in a regulated environment. For ConsenSys, this was another professional engagement with regulated customers that involved hands-on training and a successful deployment process for a Quorum testnet on varying hardware.

The aim of the project wasn’t to replace SAMOS. Rather, it aimed to evaluate the use case of DLT in wholesale payments for interbank settlements and, therefore, to build on and enhance the work done in other countries. SARB joined other central banks around the world that are experimenting with distributed ledger technology such as Brazil and Canada. The project was a significant step forward for Ethereum and Quorum, showcasing how the banking industry stands to benefit from blockchain technology.

This system can be developed to enable other uses beyond wholesale settlement. Examples include the exchange of tokenized money for other tokenized assets, like bonds or securities” — SARB

8Project Khokha Case Study