January 29, 2018
By Harry Stahl, solution architect, Institutional & Wholesale
Distributed ledger technology (DLT) – generally referred to as blockchain – is picking up steam in terms of both enthusiasm and potential adoption. As discussed in my last blog, there are a number of areas where it can be quickly implemented and quickly show results.
However, anyone considering investing in a DLT project should consider a few challenges before jumping in. It’s not all about the technology:
1) Security and entitlements. DLT is built around sophisticated cryptography. This has been successful, but the business model of cryptocurrencies is permission-less and much simpler than in capital markets, where distributed ledgers will be permissioned, and where protecting the privacy of transaction participants may also require the use of private keys. Even if the cryptography can never be broken (always a question) it will need careful thought to design and execute systems that balance security and entitlements concerns on a large scale. Just as important is the governance: who gets access, who may use private keys, etc. This is a critical dimension for Anti-Money Laundering and Know Your Customer and is outside of the DLT itself.
2) Functionality. While DLT has proved its value in bitcoin, other domains also add complexity to some of the core behaviour. For example, as the ESMA report on DLT points out, for centrally-cleared OTC transactions, systems compute and settle position margins and collateral requirements on a net basis. Also, market participants routinely transact margin finance with externally-financed assets. This is not to say that DLT cannot support these processes through smart contracts or other techniques. But it is always worth a detailed assessment of the full functional capabilities you need.
3) Revising transactions. Cryptocurrency transaction have been designed to be immutable, which has worked well, though there have been dramatic, if rare, exceptions. The experience with securities transactions, however, needs to accommodate corrections the arise from operational errors at various stages by various parties.
4) Performance. Some DLT processes are not as fast as you’d expect even on the standard Internet, largely because you’re dealing with complex cryptographic processes. DLT can be locked down further and made more secure, but this can slow things down. Witness the Cryptokitties collapse. Specifics count when thinking about performance, but definitely consider the volume and timing of flows when evaluating DLT, and include performance benchmarks and stress testing early in prototype or pilot program.
These are few of the obvious dimensions to consider. But don’t be put off either. Think about where DLT might hold promise for your organization. Then create a SWAT team that includes technical, functional, and regulatory experts – not just your tech evangelists – and take a practical and realistic look at blockchain. If you quantify the benefits and uncover the gotchas early, you can build a business case you – and your stakeholders –can believe in.