The business of securities settlement ought to look like a club night with Pete Tong. Just as the DJ and electronic music pioneer stands before a crowd effortlessly manipulating gigabytes of music from a single iPad, the back offices of the world’s financial institutions are supposed to allow traders to switch in and out of trillions of dollars worth of assets without skipping a beat.
In reality, their performance looks more like Rick Wakeman, the keyboardist from the 1970s prog rock band Yes, who would frantically wave his wizard-robed arms as he yanked cables out of a dozen incompatible boxes.
Harmonising the ledgers of the world’s banks requires scarcely less skill and energy. Records of trades have to be compiled and entered into those systems from the phone calls, emails and post-lunch handshakes that constitute the original bargains. Reconciling those trades and ensuring that all parties have the same records is a painstaking business. Electronic trade confirmations help, especially for standardised products, but in many markets paper tickets are still the norm.
Not all failures are as spectacular as those which culminated with Kweku Adoboli of UBS being found with up to $2bn of losses on unconfirmed deal tickets in his desk drawer. But mistakes slip through the net all the time, and cost money when they do. In the US Treasury market alone, about $50bn of trades “fail” every business day, incurring a charge equivalent to 3 per cent interest a year on the delinquent balances. Including private sector markets, the total volume of fails is likely to be in the hundreds of billions.
In the world of stadium rock, the problem of standardisation was solved long ago. You do not see Radiohead’s Thom Yorke posing in front of a huge stack of electronics resembling a telephone exchange, because in 1983 the makers of synthesisers and keyboards agreed on a common standard that lets you plug one into another without an impressive work of amateur electronics.
Financiers may not have noticed, but a technology has been invented that would solve their compatibility problem, too. Its name is bitcoin, although it could probably do with a new one; a whiff of sulphur still clings to cryptographiccurrencies because of their historical association with anarchists and the customers of the Silk Road online narcotics emporium. Richard Gendal of IBM has suggested that the underlying system should be called “shared ledger technology” instead.
So far as legitimate finance is concerned, the utility of this technology has little to do with its supposed ability to make transactions untraceable — a feature, if that is what it is, which can easily be switched off. It has everything to do with the “blockchain”, a public record of transactions that is updated whenever one person sends payment in bitcoins to another. Every bitcoin user can consult this shared ledger, and any copy is as good as another; there is no uniquely authoritative source. Clever encryption keeps the copies in sync, by making it easy to tell whether one of them has been doctored.
This is exactly what securities settlement needs: a way of recording every transaction once, sharing that record between the counterparties in exactly the same form, and updating in an agreed and standardised manner. It provides an easy way to set up the settlement systems for new or customised products.
It also automatically creates a full, agreed record of transactions, which would assist risk management and compliance functions hugely. And, with such a system in place, it should be straightforward to ensure that netting opportunities are never missed; if one trader sells units in an exchange traded fund and a colleague buys units in the same fund, their institution need not transact with any external parties.
There is the potential for cost savings that go well beyond eliminating mistakes. Back office employees earn salaries into six figures for reconciling settlements and confirming trades by hand. (One of the shameful secrets of financial services is how many people do jobs that could be made obsolete by a small computer program for pasting data from one window into another.)
The biggest obstacle to the adoption of “shared ledger”, of course, is one which has delayed much necessary investment in the past. It is unlikely that the banks will take on bitcoin in its current form, so they will have to agree on the design of a new standard. That means working together and trusting each other. Alas, those are habits that come less easily to bankers than to musicians.