Bonds on blockchain: an idea whose time hasn’t come
Capital markets have always been among the early adopters of new technology, but blockchain for bonds is further away than you might think. As always, humans will get in the way.
Any bank worthy of the name is now furiously rebranding itself as a technology company, and capital markets bankers know which way the wind is blowing. Even if, deep inside, bankers want nothing more than a chance to get on a plane, pitch over a good lunch, and sell bonds over the telephone, they’ll be paying lip service to blockchain, distributed ledgers, big data, machine learning, algorithms, automation and all other “disruptive” technologies. Nobody wants to be branded the Luddite who didn’t see the possibilities inherent in the industry, and Silicon Valley’s glamour eclipsed Wall Street's a long time ago.
That’s part of the thinking behind Commonwealth Bank of Australia’s bold claim to be the first bank to issue a bond on the blockchain. The issue, an auction for Queensland Treasury Corporation, was sold back to the issuer, and was never clearable, fungible with any other QTC issue, or even recorded in the Australian state’s debt total.
But it was a proof of concept. CBA recorded the major contractual terms of the bond in digital format, including a tweak which made the coupons automatically payable. The blockchain settles and delivers the bonds instantly and irrevocably — investor bids at auction are binding. That’s a real step forward, and counts as a victory, though plenty of other blockchain and securities trials are underway. Ex-JP Morgan banker Blythe Masters’ Digital Asset Holdings is working on loan and derivatives settlement, and partnering with the Australian Stock Exchange on equities.
Deutsche Börse and its clearing house subsidiary Clearstream have several projects going on while nearly every international bank is a member of the “R3” consortium (though Goldman recently quit). Bonds are a different asset class, but from a tech wizard perspective, all financial instruments, securities or not, are just contracts and ownership ledgers.
When modems were modish
The tech wizard perspective, though, is exactly what’s generated such high expectations for the blockchain, which are unlikely to be fulfilled in the next decade. Consider previous rounds of technological innovation to see why.
The internet, for example, existed in embryonic, but adequate, form by 1994. It wasn’t terribly fast and didn’t have much content, but it was perfectly useful for, say, bond trading and new issues. The number of kilobytes in a pricing message or new issue announcement would travel through a dial-up connection just fine, even if any website with pictures was a challenge.
Yet here we are in 2017, with only around 40% of bond trading done in a natively electronic format. Much of the rest, it’s true, goes over Bloomberg messenger, but this is hardly the automated tech utopia of 1990s dreams. The “e-bond” issues which enjoyed a brief vogue in the early 2000s, were electronic in the sense that playing solitaire on your PC counts as playing computer games — they didn’t exactly explore the possibilities of the format.
The problem is human behaviour and long standing market practice, and this will be the problem for bonds on blockchain as well. To derive any real benefit from the technology, all of the different blockchain systems under development — for payments, for central bank currency, for securities settlement, for auction, for trading — must talk to each other, and create the same levels of trust and verification between blockchains as within them.
To put it another way, even if you have a distributed ledger for say, trading French government bonds, and can settle immediately in a trusted, private environment, you still need to make this settlement system feed into whichever system you use to deliver collateral to back derivatives trades, or to the system which identifies repo collateral, or to the system which you use to hedge new issues (as well as internal risk systems, regulatory databases, and so on).
If that works out, then it will make the movement of collateral much more efficient. But history isn’t encouraging. Europe’s securities settlement systems — a broadly homogenous set of institutions in a single political block — are still not fully connected.
The ECB’s Target 2 Securities system has rolled out in four waves so far, and a fifth is planned. This project took years and cost millions, and, like blockchain, promises a more efficient future. But the technology to do it has been around for decades. Only human co-operation and cost has stood in the way.
In electronic trading, too, it’s easy to see what a good system would look like. A central marketplace which can incorporate buying and selling interest from anyone interested in a given asset class, anywhere in the world, with settlement through the main clearing systems, liquidity provision from all the main market makers, and trading protocols which balance firm prices with information leakage.
But instead, almost every bank went through a phase of starting its own “click to trade” portal, followed shortly by dozens and dozens of start-up e-platforms, each offering an individual twist on trading protocols, each competing to attract liquidity from market makers, and few offering anything like the depth and immediacy that would make them useful for investors wanting to move institutional quantities of bonds. The technology to make them all talk to each other, or to route orders to the most efficient platform, is comparatively trivial. The problem of human co-ordination and willingness, not so much.