The Finternet

A new paper from the Bank for International Settlements proposes the creation of a unified ledger and the tokenisation of everything. If it is successful, this could represent the most fundamental re-imagining of the banking system since the Medicis.

This is a slightly longer, link-enhanced version of an article that first appeared in the Mint. You can read the original here.

Last year I wrote about how we might reimagine the modern financial system by creating a unified ledger. I based the article on a speech by Agustin Carstens, General Manager of the Bank for International Settlements in which he argued in favour of digital ledgers and tokenised deposits:

What this would immediately mean is that there would be two separate digital ledgers - one for CBDCs and another for tokenised deposits. However, because these systems are digital we can, Carstens argues, connect them to each other through a unified ledger on which both public and private digital money can interact - unlocking new efficiencies that were impossible until now.

Last week, the Bank for International Settlements published a paper that proposed the establishment of ‘the Finternet’—a brand new digital framework that uses modern technology protocols to re-imagine how the financial system might work. If implemented, this will change the global financial system in ways the world has not witnessed since the Medicis of Europe.

All commercial transactions—the buying and selling of goods and services, making of investments, taking of loans, etc—operate within distinct regulatory environments that clearly specify what can and cannot be done in relation to those transactions. These environs are distinct from each other, overlapping only when a payment needs to be made or a transaction recorded. In almost every instance, they are designed so that the financial assets they regulate (the land title registries, the record of customer deposits with a bank, etc) remain distinct from the rules that govern them.

While this system has served us well for centuries, as commercial transactions have grown more complex, the inefficiencies in its original design have slowed down operations, increased costs and restricted competition and innovation.

The Finternet seeks to address these shortcomings using protocols and technologies similar to those that underpin the modern internet to connect these different ecosystems to each other. Rather than relying on traditional clearing systems and messaging chains, it proposes the tokenisation of financial assets, allowing them to transact seamlessly over digital ledgers that are designed so that they unify these different domains.

At the core of the proposal is the notion of tokenisation—the representation of financial and real assets in a digital form. A tokenised asset contains not just information pertaining to that financial asset, but also the rules that define what transactions can be performed on it and how. By unifying the asset and its governance rules, the Finternet would enable a range of transactions that were not previously possible.

Tokenised asset transactions are expected to take place on a unified ledger, a shared programmable system on which various different financial asset markets—central bank money, commercial bank deposits, company shares, government bonds and real estate assets, for example —can be managed and exchanged. The unified ledger is not a single centralised system, but rather an interoperability framework that is capable of connecting all digital ledgers that conform to the unified inter-ledger protocol. This would ensure the integrity of transactions and consistency across different ledgers, providing finality through strong technical guarantees. A transaction completed anywhere on the unified ledger would become irreversible everywhere.

Since users could open accounts on one ledger and perform transactions on another, tokens could be traded directly among holders without the messaging systems and intermediary institutions that are currently required. This opens up new settlement possibilities with reduced counter-party risk and no collateral requirements. As a result, registered assets can be immediately transferred with little or no reliance on external verification processes. Since tokens are programmable, operations and obligations can be embedded directly into financial asset. Since they are composable, it is possible for multiple transactions to be bundled into a single executable package.

Streamlined in this manner, financial transactions will become cheaper, faster and safer. Complex financial agreements can be automated and executed directly without intermediaries. Atomic settlement will allow different legs of a complex transaction to settle simultaneously with no counterparty risk. This will allow for the development of a range of new financial instruments (cross-border trades and multi-party asset swaps) as well as new investment products (tokenised portfolios and fractional ownership rights in real estate).

That said, various challenges remain to be overcome. While it is easy to see how this sort of system will work in the case of easily de-materialisable financial assets, such as money (central bank currency and private bank deposits) and shares (which we already transact in de-materialised form), it will be much harder to operationalise for real world assets (like a piece of jewellery).

In the first place, we will need to find an effective way to tokenise such assets so that they cannot be sold simultaneously on the unified ledger as well as in traditional offline markets for cash. Unless we find a way to solve this, the unified ledger will fail to adequately address the double-spending problem that blockchain technology was partly designed to address.

Where the law requires transfers to be registered, this is easy enough to do. All it will take is a statutory amendment to convert government registers (such as those that need to be maintained in relation to real estate transactions) into tokenised ledgers. All transactions, whether completed online or offline, will thereafter have to be recorded on that ledger, ensuring that it remains the single source of truth in respect of the title to and interest in that property.

What is that much harder to do, however, is to make this system work for movable assets. Consider priceless works of art and other such tangible moveable property for which no public registries exist. Comprehensive tokenisation will be hard to implement - but we have to make every effort to make it work.