I used to work, many years ago, as a broker of Canadian equities in London (don’t judge, it was the ‘80s). It was so long ago that I confess that I don’t remember much about the process, other than that we would write down our clients’ orders on bits of paper, and hand them in at the end of the day to the head of the desk. From there they would get passed on to the settlements department, which had its own floor, it was so large. Several days of stock price and currency movements later, the settlement would go through. I never thought to question the efficiency, I just assumed that that’s what it took to get ownership transferred.
When I started studying the blockchain, I assumed that aha!, here was a way to save billions in tied up money, simply by reducing the settlement time to seconds rather than days. I assumed that the blockchain’s transparency and immutability would make the cumbersome checking of ownership and payments unnecessary. Matching, verifying and netting would be reduced to code. So it needn’t take several days, right?
As with most financial concepts, it’s not that simple.
To see why, let’s start with a simple example. You and I are together in a café. I have a share certificate in my hand, and you want to buy it. You hand me the cash, I hand you the certificate, and we have instant settlement.
But wait a minute: how do you know that that share certificate is not fake? How do I know that the cash is not fake? My doubts are easier to resolve than yours (I just happen to have a fake bill detector in my pocket). How are you going to check that the certificate was not forged or copied? That’s going to take you a while. We no longer have instant settlement.
Back in the day when share certificates were bearer items (pieces of paper that belonged to whoever held them, much like the €20 in your pocket), checking authenticity could have been enough for the transaction. But now you need to know for sure that I have the right to sell that certificate, because if not, once I’ve disappeared with the cash, you could find that the rightful owner appears and wants the sale declared null and void. How do you do that?
And while you’re doing all your checking, I could decide to change the price. Markets move, after all. How do you prevent that from happening?
The settlement is getting more complicated, right? Now let’s throw in the delicacies of electronic payment, and electronic transfer. To the same doubts about authenticity, we can now add doubts about settlement. Even if we’re satisfied that we are who we say we are, and that we have the right to send the digital certificate/payment, I’m not going to transfer ownership until you send the money, and you’re not going to send the money until I transfer ownership. Who can help us with that?
And, since it’s unlikely that I hold my digital certificates, I have to route all instructions through my certificate custodian. And, how did you and I find each other in the first place? Suddenly a lot of other parties are getting involved, and each is going to want to verify and check that everyone is who they say they are.
So, a lot of verification is going on. And it’s that verification that takes the time.
In theory, the blockchain can help us with verification, in that if something is stored on the blockchain, it’s fact. But is it? Let’s presume that I, personally, don’t have the power to publish my share ownership on the blockchain. My custodian would have to do that. (Why he would do that is an interesting dilemma, since once he has published all his clients’ shares to the blockchain, he is out of a job.)
And how do we know that my custodian will publish the correct information? That my share holdings are 100% intact (nothing has been siphoned off), verifiable, and in a universally accepted format? Who decides what that format is? Would that not be a very, gasp, centralized decision? For a technology that rests on decentralization?
Let’s assume that we figure out a way for my custodian to blockchain all of my holdings and manage to keep his job. So, he transfers the digital ownership of the shares I want to sell to my broker, who is in contact with your broker. Why is my broker going to trust your broker to make the payment on your behalf? It’s easy enough if they know each other or have done business before. But what if you are in Azerbaijan and I am in Iceland? (And let’s not even go into how long it would take a payment to get from Azerbaijan to Iceland…). Right now regulation makes this all work relatively smoothly. And regulation insists on verification and re-verification of the facts and identities. It’s very unlikely that just putting it on the blockchain would be “good enough” for the regulators.
And we shouldn’t want it to be. Efficiency is great, yes. But when it comes to large sums of money, reliability is more important. And when you think of all of the verification that needs to happen for a trade to take place, settlement of two or three days doesn’t sound like that much, after all.
It is possible that the financial sector could come up with a way to encode verification, identity and transactions. But to do it going back far enough for it to be useful would entail a colossal cost. And to do it in a uniform manner in a fiercely fragmented business would require almost magical management skills.
But that’s not to say that blockchain settlement is not a good idea. Some asset classes take longer to settle than others. So, it’s quite possible that we’ll end up seeing separate settlement practices for separate types of deals, and some of those may well use the blockchain. I can especially see blockchain-style settlement being used for future asset classes that haven’t even been invented yet. Building a new settlement system from scratch would be an excellent opportunity for the blockchain to show its power. Converting current systems? Not so much.
And while decentralized trading sounds efficient and, well, democratic, we need to take a look at the steadying role a centralized control can play. Imagine what could have happened in September 2001, and again in the 2008 crash, withoutSEC interference. And the need to reverse a trade, either because of error or because of contractual conditions, is a relatively common occurrence, and something that the blockchain is not set up to allow.
The “blockchain as a new settlement system” dream is appealing. But as with much of the information and talk surrounding this powerful technology, it is largely hype. An improvement on the current system would benefit liquidity and profits, and make the sector more resilient to shocks. And the blockchain does have a role to play in making payments and data transfer more efficient. But it would be a mistake to assume that technology alone is the barrier to an instant settlement system. There is much to explore, though, and much to learn, and things can always be improved, indeed should be if the solution is practical. The sector and the blockchain will find ways of working together if we can move away from sweeping proclamations and towards practical applications that can help today.