Blockchain – Part 1 – Technology Overview
In recent years, two of the major drivers of change in the financial industry have been regulation and technology, and it is likely that this will continue for some time. There is one technology in particular that has been generating much interest (both to participants and regulators), active research and product development; a technology associated with a great deal of hype and expectation. This is the technology known as blockchain.
For Sinara, it’s always important we keep abreast of new developments that affect our industry, and be able to advise our clients accordingly as far as blockchain is concerned. As part of our ongoing efforts in this area, I attended the Blockchain Summit in London on 28 March 2017, a one-day conference bringing together a range of industry participants to discuss developments in blockchain technology. The moderated discussions covered not just the technical aspects, but other key areas such as regulatory, legal, and how it is being applied in the industry, barriers to adoption, and the biggest areas for growth.
In this first of a series of three posts on blockchain, I’ll give an overview of the technology. In the second post, I’ll discuss trade financing, capital markets, and cross border payments, some of the blockchain applications that were spoken about at the Summit. In the final post, I’ll talk about applications in banking (as well as central banks), self-sovereign identity, and regulation.
What is blockchain?
So what is this technology all about? We had a morning session at the Summit giving an overview to delegates, and I found plenty of articles on the Web going into varying levels of details. I’ll try to give a summary here.
A blockchain is an electronic ledger that can be shared in its entirety between disparate participants over a public or private computer network. It contains a growing sequential list of unchangeable, transparent records (called blocks) of the participants’ transactions, each one time-stamped and cryptographically linked to the previous one.
Blockchains are therefore a type of distributed ledger technology (DLT), though it should be noted that not all distributed ledgers necessarily consist of blocks or chained transactions.
A blockchain can only be updated by consensus between all the participants, and once a new block is appended to the chain, it can never be removed. At any point in time, therefore, it is guaranteed by design that the blockchain contains an accurate and verifiable record of every transaction ever made in the system.
This is done through a highly intensive calculation that generates a unique cryptographic hash code, or fingerprint, for each block. The hash code is based upon both the contents of the block as well as the hash of the previous block. This means the contents cannot be modified without invalidating the block’s recorded hash, making it ideal for applications in recordkeeping and auditing. If you were to try and modify a block in the chain, it would invalidate its hash, making it obvious that it has been changed. In addition, it would end up invalidating the hashes of all the other blocks down the chain, making editing infeasible (since you would have to recalculate them all, which is computationally expensive).
Furthermore, in a blockchain network, a copy of the entire blockchain is stored on many servers, and every server is in constant communication with all the other servers on the network. The copies of the blockchain on all these machines are constantly kept synchronised to ensure that they all contain a valid copy, and that any invalid copies are quickly rejected. Therefore, to add a new transaction into the blockchain, it must be accepted by all the servers, effectively meaning that authority and trust now exists in a peer-to-peer network.
This makes attacking the system even harder, since if you wanted to modify a previously recorded transaction, you would have to not only recalculate the blockchain, but update it on every server that holds a copy of it! In a sense, the more servers that share the data, the better the data is protected. The failure or corruption of an individual server has no impact on the availability or correctness of the data overall.
This approach is in contrast to the industry’s prevailing paradigm of centralised record keeping, i.e, when one database contains the master copy of the data, responds to queries and records new transactions. Sinara have certainly built a fair few of these over the years! With the centralised approach, the organisation who owns the database controls the data, and if the database becomes unavailable, so does the data. As a result, central databases require many layers of protection such as disk mirroring, backups, failover sites, encrypted data storage, encrypted connections, as well as physical access control. With a blockchain database, these measures will probably still be needed to an extent, but they become less critical.
Why is it useful in the financial industry?
A good way of thinking about the benefits of blockchain is by considering how ownership is currently recorded. At the moment, you require a trusted third party of some kind to be able to verify that you own something, whether that’s money held by a bank, equities held by a custodian, intellectual property rights by a patent/copyright office, etc. If you want to change ownership, you need to contact the trusted third party, follow the correct procedures, probably pay a fee, and they will transfer ownership by recording the change in their own database.
With a blockchain, the knowledge of who owns what is not held in one place by one organisation. Instead, it is shared with everybody, and there are lots of copies of this blockchain (let’s think of it as the “register of ownership”) maintained by different participants. All these registers are compared to one another on a regular basis to ensure they are in agreement and that any errors are fixed.
Whenever a transaction occurs, both the ‘buyer’ and the ‘seller’ have to broadcast the relevant details to the network. All participants with a copy of the register have to update it, after making sure that the new transaction doesn’t break any rules. The registers will then be compared to each other again. If there are any discrepancies or disagreement about the contents, the most common register on the network is accepted as the correct version.
In this way, any corrupted register will be ignored by the rest of the network, and it will have to accept the majority version or be expelled. The manipulation, errors or failure of an individual register will therefore have no impact.
So in order to move something of value over a blockchain (whether that’s cash, securities, or anything else), the participants must first all agree that the transaction is valid, which means no single participant can give an authoritative view on transaction validity. As mentioned before, a malicious user seeking to modify a blockchain couldn’t just go in and attack one system: they would have to attack every single computer holding a copy of the blockchain.
Blockchain therefore reduces or changes the role of a large number of intermediaries, allows more trust to be created, and potentially for costs to be reduced. It eliminates the need for many separate recordkeeping systems and therefore reduces the chances of reconciliation problems. Indeed, if you were to follow this to its extreme conclusion (and believe the hype), blockchain may eventually be the method for integrating the entire world’s record keeping!
A comment from one of the speakers at the Blockchain Summit that really hit home was that the present state of blockchain was similar to that of the Web in the mid-90s. Lots of potential, as yet unexplored, lack of standardisation, underdeveloped platforms, and still no ‘killer’ application. If developed further and widely adopted, it could be possible that blockchain could be just as revolutionary as the Web.
Whether this actually proves to be the case remains to be seen, but expectations are currently high. Last year, Deutsche Bank conducted a survey of financial market participants and found that almost 75% of respondents saw blockchain being widely used within the following 3-6 years. There are few banks and large financial institutions that don’t now have a Head of Blockchain, and indeed there were a fair few at the Summit.