Blockchain's limitations for clearing and settlement of securities
In an ideal world, distributed ledger technology would allow completely decentralized and privately controlled booking of property rights and transaction records without the intervention of a central authority. Could this replace the central booking systems of settlement infrastructure used in securities exchanges? If so, clearing and settlement could follow the decentralization of trade matching that has occurred in US and EU markets.
The triune strengths of encryption, control over booking by democratic rule of ledgers and integrity of ledger architecture should ensure that each booking is indelible and unalterable. While these guarantees may be enough for some types of systems, it is doubtful they will hold up under the volumes and pressures of a major securities exchange. Also see Benito Arrunada on this.
Let's look at decentralization first. It is hoped that many copies of the booking ledger would be maintained, making the market decentralized and democratic. However, even a market with very low turnover like the Stock Exchange of Hong Kong averaged turnover in 2016 of about 184 billion shares every day. The current matching system in Hong Kong can handle about 100,000 transactions a second. Who could maintain a ledger capable of such speed and volume? Clearly only a large broker-dealer bank. The democracy of ledger control would probably look more like an oligarchy of large institutions.
Would the encryption be able to guarantee that within this market oligarchy of ledgers, all bookings were perfectly safe? That sounds like a rhetorical question. As the strength of encryption is provided by what is essentially a difficult mathematical problem that would take enormous computing power to crack, what we rely on is the inability of a wrong-doer to marshal such computing power. Going forward, however, this is not a terribly robust hook on which to hang the property ownership recorded in the world's securities exchanges. Together with the military and financial industry, organized crime is one of the agents that actively competes to provide world-class hacking using the most powerful methods. The ultimate protection that would likely be used for a securities settlement ledger would therefore not be encryption, but merely a closed circle of persons with access to the ledgers, that is, the same oligarchy of large, mutually trusting institutions that are able to maintain the ledgers.
This leaves us with the reality that a push for use of blockchain in clearing and settlement of securities transactions would look very much like the decentralization of the market's front end since 2000: large institutions would replace centralized exchanges as the locus of regulated activity, making regulation more complex and expensive without providing benefits to the average investor. The network among these institutions may very well use blockchain or another distributed ledger technology, but the result would be neither open nor democratic. It would more likely be merely a complex network of ledgers owned and maintained by large, market insiders, thus removing booking from securities exchanges to their major participants. Improvement in the applicable technology might change this state of affairs, and encryption is not standing still, but for the moment promises far exceed delivery.
David Donald, Hong Kong