Sometimes, signs of a broader conceptual technology challenge show up unexpectedly in a very neat, concise package. That was the case this past month when a pair of news items—seemingly contradictory in nature—popped up regarding the implementation of blockchain in the foreign exchange (FX) market, in particular, a massive project within the bank-backed market infrastructure for FX, known as CLS. Discussing them isn’t to impune CLS, so much as to examine the consequences for market participants as the industry bounces wildly from unbridled optimism to incremental stops-and-starts around ledgers.
First, it’s worth a brief discussion of the news—each side of which touches upon CLSNet, the utility’s new trade processing platform that has been in development for around two years. Just recently, Financial News in London reported that CLS “banks pick old tech over blockchain for FX service,” citing reluctance from major bank users to jump into CLSNet’s ledger-based processing, given its current state of readiness. CLS chief strategy officer Alan Marquard noted that security concerns were a leading factor in those decisions—and a reason that its core settlement service, underpinned by traditional database technology, was being offered in parallel.
Whatever the motivations, it was a bit of a rare admission from a major industry player that despite significant investment, time, and energy expended, blockchain still isn’t sufficiently robust enough to go live in a serious way. As Marquard candidly put it, “building on new tech is always more expensive. We took the step and made this investment. Nobody has a crystal ball.”
On the other hand, only two days later, IBM—which has partnered with CLS using its Hyperledger Fabric—was out in the press discussing a new connectivity and application portal that had grown out of the CLSNet project, called LedgerConnect. In FX context, the idea behind the portal is to enable participants to fill in some of the gaps using ledgers—automated trade data reconciliation, removal of data duplication in netting processes—that run up to completed trade settlement. Ram Komarraju, head of innovation and solution delivery at CLS, explained to CoinDesk that “there is a lot of trade processing we do for banks and buy-side firms, without getting to the last mission-critical aspects of settlement itself”—though he also noted that CLS will remain agnostic, and indeed expects to deploy multiple ledger technologies over time. Meanwhile, a spokesperson from IBM said the computing giant is releasing LedgerConnect now to get past the growing perception of “proof-of-concept fatigue” in financial services.
If you’re a little confused by these two bits of information (or their timing), you could be forgiven. One could reasonably read the PoC fatigue comment as a friendly call to the industry to throw some cards on the table—and IBM aren’t wrong about the PoC strategy probably leading to more inertia at this point than innovation around ledgers. It’s time for more live environments. Yet, it would also seem a curious moment to do that when, if the FN reporting is to be believed, many CLS users are still skeptical enough of the basic operational risks of blockchain to steer clear.
Without speculating on that bit of intrigue, the situation does clearly point to a larger, more delicate challenge for FX trading desk technologists. The question, as often, goes to the why. The FX market is heavily fragmented and extremely volume intensive; therefore, it makes sense to at least try to bring ledgers to bear for certain currencies (after all, blockchain was originally designed to manage a store of value, albeit a stateless and virtual one). CLSNet is among the biggest blockchain swing so far. But for those managing internal reconciliation systems and books of record, conflicting signals around blockchain can lead to the making of an even greater mess out of a problem that was already fairly messy. It demands their systems must be highly flexible and adaptable to change—and in fact, must be able to cope with switching back and forth as necessary between PoC-style pilots and core legacy platforms in short order. Sloppy coding and manual workarounds just won’t cut it when a major industry utility like CLS is oscillating between present and future-state technology.
The fact is that we rarely see headlines like “banks pick old tech” because, frankly, no one wants to be perceived as stale and staid in a world where talk of disruption reigns. But the more honest explanation is that some banks remain on legacy platforms (rather than actively choose to do so) because their own tech constraints make it difficult to test or adapt to the fresh alternative. Long term, is the CLSNet distributed ledger better? Is its security really that troublesome? Many firms simply may not know—and that question is probably at the heart of both sides of the CLS-IBM news.
Call it less about blockchain and more about that “crystal ball” mentality Marquard mentions. Less about being bold - more about being ready.