ASX's new blockchain-based CHESS system: A marathon not a sprint

The exchange expects the real innovation to come after its 2021 go-live.
Written by Asha Barbaschow, Contributor

The Australian Securities Exchange (ASX) is well on its way to going live with an actual industrial-scale blockchain use case, and not just a glorified database.

The exchange announced in June 2016 that it was building a new post-trade solution using blockchain technology, which it had been working on since 2015. The solution is set to replace its legacy Clearing House Electronic Subregister System (CHESS) platform, which has been running for around 25 years. It's also written in COBOL.

There's about AU$2 trillion worth of equities registered in CHESS. Per day, around AU$5 billion is processed through CHESS, and 300-500 transactions and up to 1,500 messages are processed per second.

When it was commissioned in the 1990s, CHESS was seen as an innovative, world-leading solution. At the time, there were no globally accepted standards for data exchange, so the ASX created its own proprietary protocol.

The ASX is about 30-40% through its major functional build, planning for industry-wide testing of the new system in 2020, and after a few push backs, go-live for the new system is slated for 2021. According to ASX general manager of engineering and architecture David Campbell, the distributed ledger (DLT) project is a marathon, not a sprint.

Speaking at the Gartner Application Architecture, Development, and Integration Summit in Sydney on Tuesday, Campbell provided further information on CHESS, now the ASX is actually moving forward with actual progress.

Read more: Here's what to expect from ASX's blockchain-based CHESS replacement

"If I was to buy a BHP share from my friend Rob, there'd actually be 15 people in that chain of transfer and all that's actually happening is I'm giving Rob some money and the name on the title in the CHESS database is changing from Rob's to mine, yet there are 15 organisations in that chain," Campbell explained.

"That's what we call equity post trade and it represents about 13% of our business.

"We're re-platforming that with DLT -- we're talking about a real industrial-style, mission-critical solution."

CHESS sits at the heart of banks and brokers as they exchange information about trading; CHESS is constantly sending messages back and forth to reconcile the industry view of the clearing and settlement data for the Australian equities market.

Apart from the operator, everyone's operating on a copy of the data and not operating on the source of truth. If there's ever a change, everyone has to ripple through that change.

Campbell expects this to be a lot more efficient with DLT.

"For us, DLT-based CHESS is the next generation of market infrastructure, moving away from the notion of a central database messaging to distributed database shared with our customers, allowing for real time source of truth to be shared, whilst also preserving privacy and integrity," Campbell explained.

"It's a shared, replicated ledger. We have a copy, our customers have a copy -- anyone who takes a node has a copy of the ledger -- all synchronised. Each transaction refers to the previous transaction, and is cryptographically signed by a legal approver. In that case, that legal approval relates to trust in the system and a very important element."

In this instance, the ASX is lucky it is trusted as the default in the market, as trust is fundamental for the DLT solution.

"In a public blockchain solution, everyone has access to all of the data. In a permissioned DLT, what you end up with is shared business logic, but the actual data is only available to entitled users. So If I'm running a CHESS clearing and settlement solution, not everyone can see everyone else's holdings, but they can see their own and they can see their counterparties if they give each other access," he continued.

See also: Blockchain: A cheat sheet (TechRepublic)

The ASX wants its customers to be able to opt in, when it makes sense for them. If they don't wish to take a node, they can still communicate with the exchange through messaging. The only difference is they will need to do so using the ISO2022 messaging standard.

"One thing we are mindful of, though, is we don't want to force our customers to adopt a pace of change, which is faster than makes sense for them. So we see this as an incremental evolution, even though the underlying technology we're using is based on DLT, customers can still connect using messaging, and they can opt into the DLT solutions as it makes sense for them," Campbell said.

The benefit of taking a node, however, is having that golden source of data.

Connecting directly using the Digital Asset Modeling Language (DAML) will also allow customers to create shared workflows and innovate on top of the platform.

Digital Asset, the firm building CHESS with the ASX, has open-sourced the DAML SDK to enable customers to write applications.

Post go-live is when Campbell expects the real innovation to happen.

"Post go-live, we see investment by our customers, intermediaries, and technology providers into new solutions sitting on top of our DLT platform, additional products and asset classes. and exploring opportunities where synchronised source of truth data has real value," he said.

Joining Campbell during his presentation was Adrian Leow, a Gartner senior director analyst, who called the 2021 go-live of CHESS "pretty aggressive".

"This is a long run journey; it's not this overnight magical blockchain revolution," he said.

"Our time frame is more around 2023 -- looking at that kind of timeline, how you're progressing, that will be good to see if that can actually come through, because we're not seeing as many examples of things deployed into production yet."

"So if you can hit that date that will be really good use case to actually talk about the future," he said to Campbell.

CHESS will be completely decommissioned once the new system is ready. The ASX has no plans to run it in parallel.

"I can assure you we've got our best and brightest on it," Campbell concluded.


Editorial standards