Crossing the trust boundary with DLT pt1: Consortiums

Earlier this year I wrote about the main reasons for using Distributed Ledger Technology (DLT) to solve your use case. In summary these were:

Trust Boundary Extension: does your use case need trust between parties who would not normally trust each other? Does an agreement or consensus need to be reached?

Disintermediation: is there a middleman that is not adding value for money? Is a new business process being prohibited due to the lack of an authority figure?

Proof of provenance: are you seeking to prove that the supply chain an asset has been on (or is going to be on) is valid and true?

Transfer of value: does your use case require instant transfer of value?

Embedded business logic: are you looking to add business logic into the transaction?

Efficiency: does your use case replace low value manual processes? Are you looking to remove reconciliations between parties with data in common? Do you need to remove exception queues?

Reflecting on these further, I can’t get away from the fact that in the end, it all comes back to Trust.

For example, a platform has to fill a trust gap when checking the provenance of something if  one of the parties contributing data about the product’s journey isn’t trusted. One of the criteria for deciding on which DLT platform you should be using, is whether it should be a public or private / permissioned distributed ledger. This is strongly influenced by the level of trust in the participants. In a public platform like Bitcoin, Ethereum or EOS there is assumed to be no trust. As a result there is a strong focus on full decentralisation and eventual consensus mechanisms (Proof of Work, Proof of Stake etc) as one assumes that there are actors in your business process who are actively trying to subvert it - The Byzantine Generals problem is a nice metaphor for this.

When considering consortiums, even though they are usually made up of selected actors who may be competing entities, there is a base level of trust. This base level is characterised by:

• each party understands the business and the market they are in,

• each party will probably act rationally in any given situation

• ultimately each party’s long term bilateral relationship has value

This means that a permissioned or private ledger is appropriate as a fully decentralised model is not necessarily required to achieve the same benefits for the consortium. In practice this means that for the enterprise the consensus protocols can be lighter weight. For example consensus is provided by the the Notary and Gatekeeper nodes in a Corda network or the Committer in the Digital Asset Platform. These are centralised components.

The implications for disintermediation is profound but rather than removing intermediaries, their role can be slimmed down. Their services need to be only those that are absolutely necessary or those which are compelling. For example in a bond issuance use case, someone needs to permission issuers and traders on to the platform and run any of the infrastructure required by the permissioned DLT. However, other services that they offer (e.g. registrar or reconciliation services) and therefore fees they charge should be on a strict value adding basis. I.e. intermediaries will not be able take advantage of monopoly or legacy positions.

These lighter weight consensus protocols for permissioned ledger platforms allow the transaction throughput to go up, as you are essentially trading the trust requirement for speed. In fact a recent slew of articles have demonstrated that speed in these use cases has become more of a hygiene factor rather than a differentiator, with both Digital Asset and Hyperledger demonstrating sufficient throughput to comfortably cover significant asset classes requirements. The magic number at the top end seems to be in the payments space with Visa’s 24k per second, although there is debate at whether this is real and it almost certainly is more about transaction recording than finality of settlement.

So far so good with consortiums but how does trust affect the individual organisation? In Part 2 we look at whether there are use cases for the individual organisation and what trust means inside one.