Misaligned views led towards $250M ASX CHESS project failure

Significant technology, governance and delivery challenges were some of the key findings revealed in Accenture’s report into Australian Securities Exchange’s (ASX) $250 million CHESS project failure.

As a result, the ASX is going back to the drawing board and scrapping its original intentions to use distributed ledger technology at its core. 

“The independent report coupled with our own assessment work, confirms a number of significant challenges associated with aspects of the CHESS replacement project,” ASX managing director and CEO Helen Lofthouse said. “These findings provide valuable inputs to helping us determine a revised solution. We have some work to do before updating and consulting with stakeholders more deeply.”

Lofthouse further clarified the $245-to-$255 million derecognition charge reflected the uncertainty of the future value of the current solution design. 

“Our clearing and settlement licenses are critical. We take feedback seriously and examine it carefully,” she said. “We are aiming for the best solution for the Australian market and will be wide-ranging and thoughtful in our analysis of the options.”

As a result, the ASX has hired Tim Whiteley as project director for the next phase of the CHESS replacement, reporting to Lofthouse. It will also establish an industry forum to provide input and receive regular status reporting on the project, complementing the existing business committee. 

As the next step, the ASX will undertake further planning to understand and evaluate solution design options and potential impact. All stakeholder activity on the project will be paused and the industry testing environment closed. 

In 2017, following two years of evaluation, the ASX confirmed it was spending $50 million in replacing its decades-old Clearing House Electronic Subregister System (CHESS) with blockchain-inspired distributed ledger technology (DLT) developed by US software firm Digital Asset.

CHESS has been used to record shareholdings and manage the clearing and settlement of equity transactions in Australia since it was introduced in 1994. 

The exchange wanted to take the opportunity to replace it with “a next-generation post-trade platform using contemporary technology”.

The appointment of Accenture in September came as the ASX pushed back the CHESS go-live date “based on current information”.

Delivery model, execution and what went wrong

Accenture’s review of the system shone a spotlight on the lack of communication between both ASX and Digital Asset (DA) matched with lacking accountability. 

The report pointed out deficiencies in execution rigour and the lack of a clear understanding of progress compounded by misalignment and expectations between ASX and DA. 

The siloed execution and reporting between them resulted in misaligned views of status including delivery progress, risks and issues. 

This was further hampered by insufficiently detailed alignment of scope, delivery plan and resourcing resulting in missed dependencies and opportunities for optimisation. 

Test and internal release management processes also lacked industrialised tooling along with management of vendor accountabilities including inconsistent information, tracking execution outcomes and quality-related metrics, the report said.

Functional and nonfunctional requirements were also misaligned and sometimes conflicting, along with granularity/quality, delivery timing and scope, which the core issues highlighted the consequences of this. 

The Accenture report was broken up into three stages focusing on the six core issues identified by ASX and DA; the delivery assessment featuring a solution review looking at the architecture and use of Daml and VMware Blockchain and a set of 45 recommendations that were grouped into 12 areas of focus on ways of working, software delivery, quality engineering efficiency and solution design. 

The six core issues identified by ASX and DA were holdings, batch settlement, bulk process support and resiliency, bilateral matching, issuer notifications and support for ex-transactions. 

During the build and testing phase of the new technology, the report found that the overarching strategy and approach deviated from effective testing strategy definition and documentation leading to significant retesting and longer feedback cycles. 

Despite the ASX’s wishes to implement a DTL system, the DA development and system test were not executed on a ledger-based environment, making it difficult for the ASX team to validate it, leading to delivering non-functional capabilities.

Quality management was also inconsistently observed throughout workshops and program artefact reviews and the change management team was unstaffed due to recent departures.