Waters Wrap: ASX’s Chess DLT meets calamitous fate—what can be learned?

“We began this project with the latest information at the time …”

This was the opening line in a statement by Damian Roche, chairman of the Australian Securities Exchange, after news broke that the exchange would take an AUS$245-255 million (~$165-170m) pre-tax write-off after halting its Chess blockchain project. This coming after a review by Accenture found problems with deadlines not being met, excessive complexity with what was completed, and issues in governance between ASX and Digital Asset, which was tasked with building the Chess replacement.

So it is that I find these words fascinating: “We began this project with the latest information at the time…”

The “time” in question was 2016, when ASX first announced it was seriously considering distributed ledger technology to replace Chess, its settlement system that was rolled out in the 90s. It was at this “time” in 2016 that ASX invested about $15 million in Digital Asset as part of the DLT provider’s $60 million funding round. Later that year, it increased its stake from 5% to 8.5% after investing roughly another $7 million. At this “time”, most stories written about Digital Asset prominently noted the vendor’s then-CEO—Blythe Masters—who hailed from JP Morgan, where she spent nearly 30 years in various senior capacities. At the “time”, Masters’ resume brought immediate credibility to Digital Asset. And when the “time” came in 2017 for ASX to tap Digital Asset as the builder of the new Chess platform, the exchange’s decision was imbued with Digital Asset’s (and Masters’) reputation.

The period between 2016 and 2018 saw the peak of financial blockchain mania, but as we chronicled earlier this year, many of the projects announced during that time didn’t make it off the ground. And from the moment ASX announced it would replace Chess with a DLT solution all the way to the moment it announced the write-off, naysayers were abundant—and loud. I linked to the story already, but if you haven’t read it yet, the Australian Financial Review has done a fantastic job of chronicling the skepticism over the years (And if you’re a regular reader of this column, you’d know that I’m not the biggest fan of blockchain/DLT.)

So “at the time” that it dove into DLT (all blockchains are DLTs, but not all DLTs are blockchains, FYI, but I use them interchangeably to make things easier), ASX thought that it could distinguish itself from other exchanges by being very early adopters of DLT. While the technology was successfully underpinning a growing crypto economy, its practical use cases for critical, high-volume exchange technology were less evident.

ASX made a gamble on an unknown technology, and that technology led to a massive loss. In the headline I used the word “calamitous”, which carries a lot of weight, but I think it’s fitting; not only does the exchange have to write off hundreds of millions of dollars, industry participants had already sunk their own millions into getting their systems prepared for the new Chess launch, which, even long before this latest announcement, led to clashes with banks, asset managers and industry groups.

Not only that, but ASX’s announcement noted that its forecast for fiscal year 2023 capital expenditure “is expected to reduce by approximately $15 million as the Chess replacement team is downsized to focus on solution design.” That’s a nice way of saying there will be layoffs before the holiday season. Furthermore, trading firms in Australia have no idea how it’s going to actually replace this quarter-century-old, mission-critical system. And if you want to talk about reputations being burned, ASX now has stink on it, and so does Digital Asset, and so does the entire Australian marketplace.

As one exchange executive told me, the DLT believers have fallen off over the last few years—long before the ASX announcement.

“One major factor [in ASX’s decision] was probably the fact that DLT technology has some fundamental limitations. I have encountered very few use cases for DLT that couldn’t be solved more easily with a simple database. And another challenge added to that is trying to integrate the solution into the industry pipes (client infrastructure, etc.).”

Fool’s mate

With all that said, it doesn’t mean that the ASX is going to cease to exist, nor Digital Asset, nor the Australian marketplace. Roche did note that there were “significant technology, governance and delivery challenges that must be addressed,” and they’ve named Tim Whiteley as the project director for the Chess replacement. He’s held various business and tech roles at Commonwealth Bank (2000-2013) and Westpac Banking Corporation (2015-2022). So at least he would seem to know what an end-user is expecting out of the exchange.

The question for Whiteley and ASX, then, is this: what can the industry learn from this fiasco?

To answer that question, I reached out to two people who have been very vocal in their dissent of blockchain/DLT: Virginie O’Shea, founder of Firebrand Research, and Brennan Carley, managing principal of Proton Advisors. For what it’s worth, I trust what they say because they have been saying publicly what so many of my bank and buy-side sources have said in private when it comes to this particular subject.

In a week that included juicy revelations about FTX’s implosion, O’Shea says the ASX announcement was the least surprising bit of news from the last seven days. Ever since 2016, the Chess replacement project has been beset by user problems and significant delays.

“The path of least resistance is the easiest to deploy when it comes to technology—blockchain is the antithesis of that,” O’Shea says. “It requires a significant structural and process change to be implemented properly and there are absolutely no guarantees it would work at the scale required for a market infrastructure at its current level of maturity.”

Additionally, to go back to Roche’s “information at the time” statement, the nascent technology was believed to be more cyber secure than traditional databases. In fact, O’Shea notes, “there’s ample proof that it isn’t any more cyber secure than existing technology—in fact, crypto platform hacks have been rampant over the last year.”

O’Shea doesn’t have much hope for blockchain’s future as firms realize that the tech is expensive and the cost/benefit is tough to justify when implementations take years and adoption is still low.

“It might find a home somewhere in lower volume and less mission-critical areas perhaps, but I’m sure the next technology fad will be buzzing around the market by that point,” she says.

Carley also says that he wasn’t surprised by ASX’s announcement. He says that the key takeaway is “don’t do technology for technology’s sake.” To go back to a point I referenced previously, ASX was looking for a bit of cachet when it went DLT—becoming the first exchange to replace a mission-critical platform with a DLT substitute.

As the exchange wrote in its statement, “[the Accenture] report identifies complexity in the integrated solution design, including in the way ASX requirements interact with the application and underlying ledger.” The thinking back at the “time” that ASX went DLT was that blockchain would streamline workflows and make settlement more efficient. That does not appear to be the case.

Cloud is well established at solving real problems, with demonstrated functional and cost benefits. And it is fit for purpose for a centralized securities depository.

Brennan Carley

As Carley notes, “It was clear from Day One that ASX’s use of blockchain was a tech solution in search of a problem to solve—you have to always start with a well-defined business requirement and solve for that.”

Another problem with blockchain is the marketing hype was (is) so great and the shiny new toy can serve as a siren’s call for desperate executives looking for a silver bullet to fix complex post-trade problems.

“Innovation is great, but mission-critical, core systems—like Chess—are not the best place to take innovation risks. Try something less critical first,” Carley says. “ASX clearly made the original decision because they wanted to be seen as cool, not because the tech was the right choice.”

Finally, Carley says a key learning moment from this debacle is to manage the number of risks you take with an ambitious project. “This project involved a new system, using a new technology, and from a new vendor. Maybe take one of those risks—but taking all three at the same time just compounds the risk of failure.”

I also believe—and this is me recklessly speculating—that after years of delays and bad press, there were executives inside of ASX who were pleading for a change in strategy. Unfortunately, we don’t have boots on the ground in Australia, but I would love to read an insider account of ASX’s original decision to choose DLT as its replacement platform for Chess; I would love to know about the tumult that led to a change of CEO midstream; and I would love to read about the final reckoning that led to the decision to take an AUS$245-255 million write-off of the project. (For what it’s worth, I’m not ready to talk about Digital Asset just yet, as I’d like to better understand what some of the governance issues were between the exchange and the vendor. As I previously wrote, though, it felt to me that Digital Asset was focusing more of late on its Daml programming language for digital assets, than hardcore DLT technology.)

Now, an Accenture report led to this final decision—from Roche: “There are significant technology, governance and delivery challenges that must be addressed.”—but I’d be willing to bet that there were executives in the room saying something to the effect of, CME Group has partnered with Google; Nasdaq has done something similar with Amazon Web Services; and even Cboe has partnered with Snowflake to improve its analytics capabilities…none of these major announcements involve DLT, but are all about cloud. 

While cloud and blockchain aren’t mutually exclusive, the biggest exchanges in the world appear to be signing exclusive contracts with the major cloud providers, and not the blockchain startups. “Cloud is well established at solving real problems, with demonstrated functional and cost benefits,” Carley says. “And it is fit for purpose for a centralized securities depository.”

I’m of the opinion—and this opinion is based on the feedback from a wide variety of senior technologists in the capital markets—that while DLT can be an effective tool for internal workflows, we’re a decade in and DLT has yet to prove efficient for industry-wide infrastructure that’s viewed as high-volume and mission-critical. On the other hand, Amazon, Google, Microsoft, and IBM have proven their abilities to handle high volumes—not just in finance, but in areas like retail, national defense, and health care.

“We began this project with the latest information at the time…” To me, the lesson is, once again, that what we know at this point in “time” is that cloud is the future—and ASX’s Chess play should serve as a warning to other ambitious industry-wide important projects.