The co-location scam of NSE could be the beginning of the end of centralised exchanges. Blockchain technology negates the risks of putting too much trust for delivery and liquidity on a centralised platform that is open to manipulation.
The NSE co-location scandal refuses to die. If one zooms out a bit, at the heart of the scandal is the very architecture of centralised exchanges. One place where everyone shows up to buy and sell – that’s just too attractive a prize for those willing to break the law for economic gain.
Till recently, there was no real alternative. A market simply meant a place where strangers can come together to buy and sell things. And for capital markets that value scale, this meant a centralised counterparty. How else will I be sure that the stock I just bought will indeed be delivered to me? The tiered approach of a centralised exchange and brokers plugging into it for their respective clients (or sub-brokers) enabled a staggeringly large number of strangers to trade with each other.
Counter(party) is over!
This system works well, so long as one is willing to place a great deal of faith in the exchange itself. After all, if I can’t trust the exchange, there is not much guarantee of my stocks or money coming to me after the trades are settled. In most situations, this faith is not misplaced. However, when it is violated, it has the potential to threaten the very stability of the entire capital market. Given these high stakes, regulators are rightly wary of anything that may jeopardise an exchange.
That means vigilance, audit costs and most importantly, monopoly or oligopoly!
It is ironic that NSE finds itself in the middle of the co-location scandal since it was set up in the first place as a competing exchange to BSE, which seemed excessively dominated by a select group of brokers to the institutional investors of yore. History, it seems, has a strange sense of humour!
TINA or TIA?
Are we in the familiar territory of There Is No Alternative? What can market participants do? Some or other exchange is needed to coordinate the order books and trades, after all. However, There Is an Alternative – especially given the emergence of blockchain technology and clear demonstration of decentralised exchanges as well as atomic swaps in the crypto domain.
What makes a centralised exchange attractive? It has liquidity, a single counterparty and economies of scale in transaction costs.
Can we hope to replicate these benefits in a decentralised architecture?
Many d-exchanges
The core idea is simple: today’s centralised exchange is replaced with multiple (maybe hundreds of) d-exchanges, each of which works on the principle of atomic swaps for immediately settling trades.
What is an atomic swap? We have all been using it, or at least our banks have been. We are used to this in the money transfer domain – with the familiar but mouthful of a name: Real Time Gross Settlement (RTGS).
The banks do not trust each other and route all their mutual transfers through the central bank. This used to be done on a net basis at the end of each business day till quite recently (and still is, for cheque settlements). However, when it comes to RTGS, the banks carry out immediate transfer, backed by suitable collateral, and still through the central bank.
An atomic swap works similarly – the exchange of money for assets happens through a single transaction that either works or fails, with no possibility of only one half working. The parties to the transaction either go home with their own assets/monies or that of the other party, never both.
If a technology platform allows individuals to participate in sending orders which are matched and then settled immediately through an atomic swap of money for assets, we are done.
What happens to liquidity?
There is a strong reason why the tens of stock exchanges India had (including one in Delhi and Kolkata) boiled down to two. Liquidity is hard to bring and it tends to be self-fulfilling. Liquidity begets liquidity.
A design with hundreds of d-exchanges is impractical unless they can share liquidity. This is challenging in the current framework because the exchanges have no incentive to build direct bridges and each bridge is a costly affair to set up and maintain.
In the blockchain world, these are trivial problems. Since no one has to trust each other, the bridge-building is extremely simple. That leaves the economic incentives to enable bridges. That boils down to customers’ demands. An exchange owned by its participants would be far more willing to build bridges as it helps them vs one owned by purely profit-driven shareholders. This also answers the question of why there should be hundreds of d-exchanges and not just a few. It would literally be a case of to each her own, when it comes to trading in an exchange – everyone would trade in one they own a part of (through a Decentralised Autonomous Organisation structure).
How far away is this?
This vision seems a bit utopian and far-fetched if one starts from the current capital markets. Regulations, investor behaviour and setting up the systems – it is all too daunting.
Come at it from the crypto world, and this seems like yesterday’s infrastructure. Somewhere between the two, we have a realistic possibility of an ecosystem of d-exchanges for stocks within the next five years.
Given the slow pace at which most regulations evolve, it is quite likely that cross border stock d-exchanges will precede ‘onshore’ ones. To put it simply, Indians might trade US stocks in d-exchanges while Europeans trade Indian stocks in a similar manner, way before the locals can sign up for the same.
The beginning of the end for centralised exchanges is here!