Blockchain impacting Clearing and Settlement
There is a strong use case for replacing legacy Clearing and Settlement infrastructure with a blockchain solution to provide a more streamlined and secure trading and settlement process. The inefficiencies in the current infrastructure creates multi-layered processes across pre-trade, trade, post-trade and custody with reconciliations being completed by every participant at every level.
It can take 3+ days to complete a transaction through these cumbersome procedures which are expensive, inefficient, time-consuming and prone to risk.
With so much to gain from improved processing, many stock exchanges are experimenting with blockchain-based solutions and in the past few years several high-profile projects have been announced including:
NASDAQ Linq: liquidity in private securities
Japan Exchange Group: trading in low transaction markets
The Korea Exchange: start-up equity trading
Deutsche Börse Group: cross-border securities transfers
SETL: payments, settlement and clearing of cash and other financial instruments
Euroclear: settlement system for the London gold market
Australian Stock Exchange: clearing and settling trades
Following several years of testing and a lengthy industry consultation process, the ASX is the first major exchange to announce it will replace their legacy infrastructure with blockchain technology. In this paper we examine whether their proposed operating model and implementation approach will realise the full potential a blockchain based exchange has to offer.
Before commenting on the proposed ASX model it is worth noting that the most successful blockchain use cases have the following characteristics:
Creating a decentralised immutable record that can be accessed by participants without the need for them to maintain their own records
Providing a trusted environment for exchanging data and conducting transactions between untrusting participants
Providing the opportunity to tokenise assets
Allowing participants to maintain anonymity
The Use Case for a blockchain-based exchange satisfies all these criteria however there are a number of challenges that must be overcome in the operating model for the benefits to be realised.
The most significant of these is that maximising the power of blockchain requires a decentralised structure in a regulated environment that currently demands a centralised institution to support market operations and preserve confidentiality.
To resolve that challenge while still meeting market needs, the ASX is proposing to use a ‘split ledger’ system. Before we get into that, let’s take a closer look at the stock exchange environment and the Centralised versus Decentralised debate.
What is the right blockchain architecture for Clearing and Settlement?
For the purpose of our analysis:
In a centralised architecture, a single party has the authority to approve and append new transactions to the blockchain
In a decentralised architecture, multiple parties maintain the integrity of the blockchain by appending and validating new transactions
Currently, the Clearing and Settlement industry structure is marked by centralised institutions such as Central Banks, Central Securities Depositories, Central Counterparty Clearing structures and Centralised Collateral Management Systems.
Following the implementation of the Dodd-Frank Act in the US and the European Market Infrastructure Regulation (EMIR), the importance given to centralised institutions has only been elevated as over-the-counter (OTC) derivative transactions must be cleared through centralised market infrastructures.
Does full decentralisation work?
Given that as background, some argue that there is no real place for blockchain as a decentralising technology in a stock exchange environment. Taking the most prominent blockchain as an example, the Bitcoin blockchain is fully decentralised and open to all participants. Distributed consensus is one of its greatest strengths, but also its weakness if it would be used to replace stock exchange operations.
First, the computational intensity of the Proof-of-Work consensus approach would make it too slow.
Second, Bitcoin miners compete to process transactions by solving mathematical problems and get rewarded settlement fees for doing so. Those miners prioritise the order of transactions to be cleared based on the fees offered and the difficulty of the problems to record it in a new block. In the stock exchange context, this could mean that investors trading stock would have to compete against each other to have their transactions cleared faster than others.
Third, even though transaction data would be pseudonymous, financial institutions would not likely be able to accept the level of transparency. If a hedge fund were to sell large positions on a prolonged basis, it could trigger a market rally or collapse under circumstances which today would be viewed as insider trading.
But taking a more disruptive-friendly view, a fully decentralised blockchain operation would unlock the full potential of the technology to help reduce costs, speed up the processing of transactions, and secure the integrity of data at all times.
One way it would achieve this is by eliminating the need of an intermediary because the rules, business logic and regulations would be enforced with each trade through Smart Contracts enabled by a blockchain such as Ethereum. In this case, the blockchain network itself would act as the enabler for all transactions in the market.
Chain CEO Adam Ludwin supported this level of disintermediation in an interview with TIME Magazine when he said:
“When people say blockchain technology will change Clearing and Settlement, what that really means is that blockchain technology will make Clearing and Settlement redundant. It’s as if I gave you a 10 dollar bill, and then asked you how do we clear and settle that payment. You would look at me funny, because it doesn’t make sense.”
While it is technically possible, it would be challenging to actually achieve this level of decentralisation in a stock exchange environment given the regulatory restrictions and vested interests held by the incumbent financial institutions.
Additionally, for transactions to effect the trade of securities through Smart Contracts likely requires the tokenisation of securities to enable near-instant transactions with cryptocurrencies, bypassing the tedious banking processes involved with money transfers today.
But if not fully decentralised, how could a blockchain-based solution still capitalise on its potential to preserve data integrity and create a safer more efficient system? Well, the ASX is proposing to solve the problem with a permissioned network that uses a split ledger.
ASX replaces CHESS with blockchain-based platform
ASX operates at the heart of Australia’s financial markets and is one of the top ten securities exchanges in the world operating the largest interest rate derivatives market in Asia. By its own admission, ASX stands at the bleeding edge of innovation as they were the first one to dematerialise securities in the early 90s.
In 2015, ASX announced it was exploring blockchain technology to replace its Clearing House Electronic Subregister System (CHESS). CHESS was one of the world’s first electronic settlement systems when it was rolled out in 1994. It provides clearing, settlement and other post-trade services for the $2 trillion cash equity market in Australia.
Over the last few years, it became apparent the aging and inflexible CHESS was in need of a radical upgrade, and in 2017 ASX confirmed it was planning to rollout a blockchain-based system run by US-based start-up Digital Asset. The intended launch date is scheduled between Q4 2020 and Q1 2021.
Balancing current market needs with future blockchain potential
On 27 April 2018, a consultation paper was released by ASX that detailed the new system’s targeted ‘Day 1’ functional requirements, non-technical business requirements, blockchain architecture, connectivity options, and implementation plan for the years leading up to the switchover date.
Over the past 18 months, consultation with stakeholders has identified 50 new business requirements relating to diverse areas such as account structures, pre-settlement, settlement, and corporate action processes.
Because the ASX is tasked with maintaining the efficient operation of Australia’s stock exchange, the replacement system also aims to meet high non-functional and technical requirements including:
Availability & Resilience: ASX is targeting the system to be reliably available 99.95% of the time during published operating hours, exceeding the current CHESS target of 99.8%
Recoverability: the new system must be recoverable without data loss to the backup site within a maximum time of one hour
Scalability: as the network extends, the system needs to support the processing of increased daily market trading and number of users connected to the blockchain using a node
Security: all integration channels between ASX and users will be protected by strong authentication and encryption, as users will only ever receive data that they are entitled to receive
Performance: the initial implementation will be tested to 10 million trades per day (CHESS peaks at 5.4 million), or a sustained rate of 463 trades per second (BitShares Decentralised Exchange claims 3300 transactions per second). ASX will maintain the T+2 settlement timeframe as the default for equities but says the new system will provide the possibility to settle some trades in one day, and potentially even sooner.
While the final functional scope will only be released late July 2018 after additional consultation feedback has been received, analysing the ASX consultation paper together with Digital Asset’s whitepaper reveals just enough to assess whether the new ASX system meets those market needs while still maximising the power of blockchain.
ASX blockchain Architecture: split ledger
Following the ASX and Digital Asset reference material, the blockchain system will be designed as a permissioned ledger where ASX will host a central database and retain the authority to append new blocks of data.
Pre-approved exchange participants will take a node to that database, to confirm instantly that their own records match those of the ASX. This differs from a permissionless ledger, like Bitcoin and Ethereum, where anyone can read or write to the blockchain.
The ledger will be broken out into two subcomponents: the Private Contract Store (PCS) and the Global Synchronisation Log (GSL).
Private Contract Store
Each participant node in the network will have its own privately segregated PCS, which contains all validated contracts to which the user is a party. In this context, the term “contracts” refers to business logic, including transaction parameters, rights and obligations, reflecting the encoded terms of legal agreements by which participants are bound.
The PCS contains a historical record of all executable contracts (both active and inactive) pertaining to a participant and must be paired with corresponding active evidences in the GSL.
Global Synchronisation Log
ASX will append blocks of data to the GSL, a shared, replicated, append-only log of hashes referring to sensitive transaction data and execution commitments. These hashes are one-way cryptographic functions which can be proven to accurately match a party’s data but do not contain any information about the confidential data itself nor the parties involved.
Users who connect through participant nodes can read and verify the integrity of the GSL and independently verify private contractual information relevant to them in their own segregated PCS, but it cannot be used to reverse engineer the contents of those transactions.
The GSL is a communication layer designed to deliver network-wide integrity guarantees and transparency typically associated with fully decentralised systems. This establishes a common and complete set of valid transaction data that, when combined with the corresponding private data in the PCS, comprises the full ledger.
What the split ledger design achieves
The split ledger design meets the requirements of both financial institutions and regulators around confidentiality and security, while still maintaining some of the benefits gained from a decentralised blockchain system.
However, it doesn’t create a fully trustless system as ASX is still the only centralised authority to have a complete view of the GSL and can validate its integrity.
And if it’s correct that the PCS is indeed a separate editable off-chain database that stores actual transactions while the GSL verifies and authenticates data in the PCS, does the new system adequately solve the duplicate reconciliation problem if each user is still updating a personal ledger?
Compare that to a fully decentralised exchange (DEX) like BitShares for example. The BitShares exchange uses a single technological implementation to effect clearing and settlement of transactions, requiring no involvement from a central authority.
In a way, it resembles the scenario that Adam Ludwin described earlier: clearing and settlement is no longer necessary as a separate process, as it becomes part of the transaction itself. In other words, a DEX gives effect to transactions whereas a system like the one ASX proposes merely records transactions.
That is perhaps simply the reality of a central market infrastructure provider implementing blockchain while it needs to preserve existing power balances and the roles each entity plays in the market.
It is certainly a massive upgrade from the aging and inflexible CHESS system, but it doesn’t maximise the power of blockchain.
ASX blockchain utopia
While market adoption of the new system will take a phased-in approach that allows users to choose how they connect to the system – by sending and receiving messages in a similar way they do today which doesn’t require them to maintain a PCS or GSL, or by taking a node and interacting through an API – ASX CEO Dominic Stevens envisions a future where the platform plays an important role as an innovation engine.
That is because market participants are being encouraged to learn Digital Asset Modelling Language (DAML), a new computer programming language developed by Digital Asset which could be used to build new applications on top of the ASX blockchain system.
This could mutualise financial infrastructure and workflow processing between institutions because they would all be connected to the same system using the same programming language. Hypothetically, if a custodian or broker were to develop a DAML application that works on the ASX system, it would also work for their clients.
It’s a big stretch, but it does mean the platform harbours the potential to have industry-wide impact as the catalyst for interoperability at an unprecedented level.
Not disruptive, yet still a milestone
Confirmation that the blockchain system to replace CHESS will be switched on in 2.5 years marks a major milestone for the financial services adoption of blockchain globally, and indicates the Australian industry is ready to make the blockchain real.
However, for a 5-year project that begun in 2015, the advancements and capabilities set out in the consultation paper seem somewhat minimal. As of writing, the ASX blockchain system does not build on decentralisation, Smart Contracts or tokenisation to maximise the power of blockchain. Perhaps it will be part of the system’s evolution in the years after the switchover date.
Granted, it is simply more challenging for a regulated platform such as the ASX to implement true transformational change, however, as Bill Gates once observed, people overestimate how much things will change in five years and massively underestimate the degree of change in ten.
The challenge for the ASX is that what was a brave decision in 2015 already looks like it may be dated by the time it is implemented and straight up legacy within a couple of years more. That may not have been too problematic if they had taken the industry with them but hoping for gradual adoption means the ASX has likely forfeited its leadership role before it has launched.
That may then expose them to the disruption they have not delivered.
This is a conundrum for large, long lead time tech projects in a rapidly evolving space and, from the materials available, it does not appear the ASX has addressed it.
Unfortunately this is a lesson likely to be learnt again where the limited adoption of dynamically evolving technology is locked into a point in time by regulatory limitations.