From ‘Blockchain Hype’ To A Real Business Case For Financial Markets
Banca IMI; Bocconi University
March 21, 2016
Introduction: Blockchain Hype vs Blockchain Seclusion?
There has been a lot of noise in the press about the great potential uses for financial markets of Bitcoin-related technology, that could be extracted from the Bitcoin world and applied to existing markets to increase efficiency dramatically. Later, there has been a lot of noise about the fact that there is no actual use but all boils down to a generic enthusiasm called Blockchain Hype, and Bitcoin is the only reality where such technology can be fruitfully used.
This paper shows that there are real business cases for improving financial markets based on the lesson learnt from cryptocurrencies, but, differently from what the hype-enthusiasts say, they are not application of a technology to the existing business model of financial markets. They are reforms of the business model itself. What needs to be exported from the world of cryptocurrencies are aspects of the market organization, inspiration for a different accounting and legal system, and some aspects of the technology. These can be a huge contribution towards more robust, efficient and stable markets, but the process cannot be immediate and effortless, and can only be achieved within a market-wide strategic view.
One crucial misunderstanding here is the idea that Blockchain Technology can be exported to financial markets as they are to make them more efficient. This is meaningless; Blockchain technology was created to change some trust-based business processes to make them less reliant on trust; without structural changes in this direction the best of Blockchain technology is lost and just the inefficiencies are left. This misunderstanding is the perfect partner of the idea that Blockchain technology cannot be used outside the Bitcoin world. This is equally meaningless; Bitcoin was created to attempt a level of independence from trust sufficient to allow players to be anonymous and do without any legal protection; other business solutions based on a level of trust intermediate between Bitcoin and traditional finance can use similar technology and yet be very different from Bitcoins. But we must ready to use the concept of trust in a totally different way, as a way to analyze the different parts of a business process and the reason for its current inefficiencies and risks.
In the next we develop these concepts first in a parallel analysis of cryptocurrencies and financial markets. Then we focus on a specific business case regarding the collateralization of financial derivatives, that we describe bottom-up including quantifiable benefits in reducing costs, capital and risk. It is an example where the use of cryptocurrency technology is not more important than the business ideas developed in the analysis of cryptocurrencies; yet it was unconceivable before examples of distributed ledgers, smart contracts and oracles were visible in marketplaces. In fact, it was first presented in Morini and Sams (2015), in an introduction of the Blockchain innovation for the derivatives world.
The misunderstanding about “trust”
Notice that the term “trust” is often used in the Bitcoin debate in a radical way, opposing a totally trustless anarchist model to a corporative model based on absolute trust. None of them really exists. Even Bitcoin has created peculiar elements of trust in new players like a stable group of core developers or miners; and financial markets have never been based on absolute trust in counterparties or central bodies. The radicalism of the debate has hidden the fact that different business models are associated to different levels of trust; trust can be hidden in many passages in the working of a market, and can be eliminated or reduced in some without disappearing from others. More than a generic term for ideological debate, trust can be used as a precise concept to understand the features of a business model, and how that model can be positively reformed; without forgetting that any removal of trust creates some form of disintermediation, of some institutions or of some functions within institutions, and in this way it requires changes to the business model, and often to the legal, regulatory and accounting framework.
An example of unnecessary element of trust is the reliance on the agreement between two counterparties about the exact representation of a deal without any automation enforcing this agreement, not even in critical cases. Many markets are still crippled by this feature. This can be addressed with elements of distributed automation similar to those seen in cryptocurrencies.
What are the problems of financial markets that we want to solve?
Financial market transactions are still based on the logic of “consensus-by-reconciliation”. Every player gives its own representation of a transaction in its own accounting systems (ledger) and its own IT systems. The only proof that this representation is correct is coincidence with the representation given by the counterparties. The confidence in the legal validity of the contract in all its aspects is crucially dependent on trust in this coincidence, which in practice needs to be verified more than once. This verification requires a number of steps, such as confirmation, affirmation, communication to central bodies, and other reconciliation passages along clearing and settlement.
This is an objective bottle-neck towards more efficient and reliable markets. Current reconciliation steps slow the process down even if the technology enables very fast communication. They also drive costs up. Furthermore, the need for this kind of reconciliation leaves open the risk of disagreement and litigation, making the process uncertain and increasing the capital requirements for members. It is a system intrinsically inefficient that has never been seriously reformed in decades, for lack of incentives and no visibility of a technological and organizational stack suitable for a change. Even if many bits of the fundamental technology to solve it were already available in the past decades, just think of the internet giving a really shared information platform, this had never been applied to changing the foundations of some transactions. Now there is visibility of a different business model in the cryptocurrency example, together with a full technology package enabling it.
What exactly are the solutions that may come out of the Bitcoin experience?
Bitcoin and the other experiments of crypto-currencies or crypto-transactions are based on a single accounting and reporting system, a Distributed Ledger. With a Distributed Ledger, the reconciliation bottleneck is avoided since there is a consensus algorithm that verifies transactions and gives to them a unique representation on the ledger, collapsing all reconciliation steps into a a single initial passage.
Further reconciliation steps are unnecessary when there is a single authoritative deal representation for all the parties. It is this business model that makes transactions so fast for bitcoins, more generally than any specific piece of technology. This insight is useful for ledgers in financial markets too, even if financial players may need distributed ledgers different from the Blockchain, which is a peculiar implementation where all transactions are reported together, visible to all, and their time-order is defined through a sequence of blocks.
For advanced financial markets, distributed consensus can be applied also to a deal made up of many payments, like a derivative or a bond, through the concept of a Smart Contract, which is a piece of program code, in a given computer language, executing the transaction agreed at inception between the parties. This guarantees the enforcement of consensus, namely that the deal will follow the agreement taken at inception between the parties. Bitcoin have only basic smart contracts, but other cryptocurrencies like Ethereum have smart contracts written in a Turing-complete language, which means it can do everything that a normal computer does.
Notice that this is a further step towards a different and more advanced model of the market. Not only the accounting/reporting of the transaction moves from individual representation to an authoritative distributed representation, but also the contract stops being two pieces of papers to be implemented and represented in separate ways but becomes a unique manager of the transaction signed (cryptographically) by the interested parties. Financial contracts are already translated by parties into software running on IT systems; what was missing were working examples of a technology where the piece of code could become the contract itself and not one of the many representations of it given by the parties. When the unique smart contract signed by the parties manages directly the flow of the transaction, there is a further reduction of delays and risks of disagreements and misalignments.
But all these goals can be obtained just via a central database and computing grid on one server.
For many of the above goals, the answer is: of course. But a computer/database shared among all the players of a market is a centralized solution, with all the well-known limits of centralization. These limits are a central topic in the state machine replication approach: centralized systems are usually more efficient from a technological point of view, but they are not fault-tolerant. In abstract terms, this means that failure of the central server is failure of the entire system. In economic terms, this unpleasant fact has additional consequences. In case of centralization, there will be an administrator of the database/hardware, and this institution would bear a great operational risk, the risk of the entire network, thus demanding an equally great power on controlling and changing unilaterally the rules. Centralized solutions create monopolies that drive the business costs up because the monopolist has not the right incentives to contain them; additionally, in finance centralized solutions also generate a concentration of financial risk that drives up – correctly – both the regulatory burden and the amount of risk-management provisions such as collateral.
A centralized database also raises the likelihood of legal disputes; it would be easy to accuse the administrator of tampering with the ledger. Since the ledger must report the situation of everyone and yet belong to no-one, a distributed ledger appears a more natural solution. It avoids the need for a central body and also reduces the legal uncertainties. The ledger downloaded by one party is the official ledger as much as the version downloaded by someone else. They are all replications of the ledger, there is not one central database and many duplications, which is a situation providing ground for uncertainty, reconciliation delays, and legal disputes.
And what if the database was “fully replicated and distributed”?
The technology of Distributed Services (and the state machine replication approach) that developed in the last decades are certainly a crucial part of the solution. There exists database technologies that try to keep away from the risks of centralization and date back to much earlier than Distributed Ledger technology. One can find works on fully replicated distributed databases that date back to as early as the early 90’s, like [1]. The evolution of the technology has brought to popular distributed solutions like DVCS (Distributed Version Control Systems), of which Linus Torvalds’ Git is the best-known.
The Bitcoin Blockchain evolved in the same stream of technological advances, partially based on the same cryptographic solutions. It is a relevant example of radical economic application of this form of technology, and in this way it showed how this technology applied logically to a market brings about a fundamental change in market organization.
Bitcoin found a decentralized solution for chronological tracking and time-stamping that was suitable for its peculiar context of building a market from scratch based on pseudonymous players. Even if this solution cannot be exported rigidly to different contexts like current and foreseeable financial markets, Blockchain is the natural turning point of distributed technology to take inspiration from when building Distributed Ledgers for financial markets, without ideological distinctions between distributed ledgers with blocks and proof-of-work, and distributed ledgers that may be different in these respects. An additional reason to keep more than an eye on Blockchain in evolving existing financial markets is to keep a standard compatible with other Distributed Ledger solutions that have different privacy and validation requirements, cryptocurrencies included.
See full PDF below.
