Building Truly Distributed Settlement Systems

On The Power of Locality and Credit

This is Part 1/N of a series of blog posts by Settle Labs aimed at introducing the ideas of credit networks. In this post, we discuss credit networks at a high level, and briefly tease at their power in transforming the financial ecosystem and unlocking financial value.

We’ve obviously seen the disruptive power of cryptocurrencies, whose underlying mechanism is a global consensus protocol that tracks all value transfers between parties. However, there is another powerful construction of value transfer, which fundamentally departs from our current model of DLTs (decentralized ledgers), both at the technological level and at the value level, by adopting the idea of trust defined as a function of socioeconomic or financial trust relationships.

First, a brief primer on value transfer with a cryptocurrency. When Alice sends Bob 10 tokens (BTC, ETH, LTC, XRP, etc), there is an explicit global agreement procedure that runs between all validating participants in each respective chain. Eg., if Alice sends 10 BTC, then every miner in the Bitcoin network validates this transaction, ensuring that it is committed, with low probability of reversing, as long as Bob waits sufficiently long. In cryptocurrencies, without exception, the trust function is implicitly defined as follows:

1. Bob does not trust that Alice will both send him 10 BTC AND not spend the 10 BTC elsewhere.
2. Bob does, however, trust that the set of Bitcoin miners as a whole will process the transfer of 10 BTC to his account correctly, and will prevent Alice from ever spending her old 10 BTC on anyone else in the future.

Minus some small subtleties, this is effectively the same implicit trust function that is defined for credit card transactions as well. The merchant, Bob, who is accepting a transaction from Alice, does not trust that Alice will actually pay him for the delivery of goods, but trusts that the card processor (in this case Visa) will do the proper thing.

Takeaway: in both cryptocurrency payments and credit card payments, the recipient of the money does not trust the sender, but instead trusts some intermediate transaction processor. In crypto, it’s miners. When buying a latte at Starbucks, it’s probably Visa. Indeed, the value proposition of crypto is in displacing the centralization of Visa through a more decentralized transaction processing mechanism.

While decentralization of the transaction processor already brings in some great benefits, we want to truly break the mold and disrupt the entire infrastructure. To do that, we can ask ourselves the following question: what if, instead of placing our trust on a single common payments processor we leverage preexisting relationships to create inter-party settlement channels (credit in return for debt)? Such a network can then additionally leverage the induced competitive marketplace of various fully independent and decentralized channels for liquidity, giving rise to highly efficient, cheap, scalable, and private transaction processing.

This is where IOweYou (IOU) credit networks come into the picture as they model trust among financial entities in a network through a directed, weighted graph, where the value of each edge shows the amount of IOU credit that a user is willing to extend to another. The key functionality of a credit network is to settle payment between two users that neither have a direct credit link nor a common currency between them but that are instead connected through one or more paths of transitive trust, where the success of a settlement transaction depends on the availability of sufficient credit along the credit path.

Key Observation: Unlike cryptocurrencies, credit networks do not require global consensus: they are an ideal target for a decentralized architecture where each user maintains their own credit links locally and checks that their inflow and outflow of credit do not change without their explicit consent. Settle realizes such a distributed credit network offering atomicity and privacy for users’ transactions in real time.

Maybe even more importantly, the ability of performing transactions through credit links with implicit counter-party risks has deep implications. Credit networks are already an abstraction that we use in our daily lives:

The vision of Settle Labs is to actually unlock all these types of implicit credit transactions we perform on a daily basis as actual code, potentially unlocking vast amounts of dormant financial value. We will expand on the market implications and network effects of credit networks, as well as the types of financial instruments that they give rise to in a future post. In the meantime, as Andrew Miller succinctly puts it:

Although payment channel networks (PCNs) such as lightning networks also facilitate similar path-based transaction facility, there are significant differences: PCNs are restricted to a cryptocurrency and inherently limited in terms of liquidity restricted to that currency, while links in credit networks are only restricted by the counterparty risk the nodes are willing to take. Furthermore, unlike PCNs, credit networks are highly amenable to geographical regulatory restrictions and compliance issues. In fact, Settle demonstrates that we can achieve the same compliance and risk management with a distributed (ledger-less) credit network as with a ledger-based solutions like Ripple and Stellar.

In summary, credit networks are a robust system for real time gross settlement and aptly abstract the current financial infrastructure. In future posts we discuss the research of Settle Labs in solving all the issues with privacy, routing, liquidity, and KYC compliance.

Appendix

Here at Settle Labs we are building the first truly decentralized and privacy-preserving credit network. Please follow us to hear more about our underlying technology, the current banking infrastructure, and partnership announcements.

Follow us on Twitter! If you’d like to see a video discussing credit networks in more technical detail, see the discussion by our CTO, Prof. Aniket Kate, here. Finally, here are some of the research papers accompanying Settle Labs.