Robert Sams, the founder of a technology company called Clearmatics, likes to present a slide of the Soviet Union's logo, the hammer and sickle, but with a "C" from the "CCCP" filled in. The plan is to highlight centralisation happening in today's derivatives markets as they move towards central counterparty clearing houses (CCPs).
"It's a bit of cheap shot," says Sams. "But even if you disagree with my entire thesis, hopefully I will have been successful in cementing into your subconscious this connection, so whenever you think of CCPs from now on its CCP- Soviet."
So what is Sams' thesis? His focus is on over-the-counter (OTC) trading in derivatives. OTC trades take place between two parties directly, facilitated by broker-dealer networks using ISDN lines, as opposed to taking place on central exchanges that supervise the trade, mitigate credit risk in case of defaults and publish prices. Since the financial crisis, to which OTC derivatives contracts undoubtedly contributed, a wave of regulation is set to make these markets less opaque and address credit risk; CCPs bear credit risk removing it from the buyers and sellers either side of a trade.
In very simple terms, centralisation of OTC trading threatens some of the desirable features of that market, such as pre-trade price discovery and the customisability of OTC contracts, while arguably concentrating risk within another set of institutions (CCPs) that are "too big to fail".
Sams believes there is a third way. OTC derivatives can be traded, using computers synced up to all agree on the automated life cycle of a financial contract, which he calls a "Decentralised Clearing Network" (DCN). This group of member-owned utilities would exist as a legal entity, but a special purpose entity to custodise initial margin and default fund assets, owned by clearing members without any CEO or executive.
The DCN uses shared business logic inspired by the decentralised blockchain technology, Ethereum: essentially a consensus-reaching computer shared by a network of users.
At the recent QuanTech conference in London, Sams' company Clearmatics and OpenGamma demonstrated a proof of concept showing how two counterparts can collectively value in real-time a portfolio of FX swaps using blockchain technology adapted from the Ethereum codebase. The demonstration was delivered by Nick Zeeb, lead engineer at Clearmatics, and Marc
Henrard, head of Quantitative Research at OpenGamma.
A few days before the Clearmatics demo, Barclays' Smart Contract Templates demonstrated an interest rate swap using R3's Corda platform. IBTimes asked Zeeb about the differences and similarities.
He said: "Corda, as far as we are concerned, is an interesting concept of how to do it. We think the concept of keeping the blockchain as a way of doing consensus computing is still important. Yes, you can solve the privacy problem by just having individual nodes in a particular agreement try and agree, but they are also likely to disagree, and the chance that they would fail to come to consensus is much higher because they are both involved in the actual contract.
"So we think that having it done on a consensus computer like a blockchain is safer because everyone else who doesn't have a direct stake in the contract is helping decide the output. That introduces a new problem, which is of course the one of privacy, so that we want to solve by using some of these more advanced techniques like homomorphic encryption.
"That way you get the best of both worlds – you get privacy but you also get a much more likely chance that the fairness of the contract will be computed and computable in the blockchain without having to rely and fall back to the existing legal system, because the only people agreeing on the contract in the Corda case are the people who are involved in it, and they are the most likely to disagree. So that's like a blockchain of two nodes right, so if one disagrees that's it."
The presentation by Sams explored in detail the evolution of OTC markets, and delved into how distributed ledger technology could and should be applied to them. Sams said: "I think that the bilateral market as it was prior to the financial crisis had problems at its core that definitely needed to be addressed. I also think the model of CCP clearing as it currently exists and is envisaged over the next several years is bad for the market and, I think, bad for society."
Among the more desirable innovations of this market is the non-proprietary nature of OTC derivatives contracts. These are not financial products; nobody has any IP on the specification of an interest rate derivative, which is just a standard and commons that belongs to the market.
However, there is a natural tendency for vertical integration between the pre-trade and the post-trade side of the market, noted Sams, so that when everything moves to CCP clearing there is a good case being made that the price discovery side of the OTC market will move away from the dealer decentralised and competitive dealer model, to one that's centralised based on a single limit order book.
"I would say that if we use the futures exchanges as a benchmark of innovation in the derivatives space, I think we would have to conclude that the futures exchanges are not particularly good at innovating contract specs and I don't think the CCPs will be much better.
"We are already seeing with the basis between these swaps that are cleared on one of the large CCPs versus another that the market is starting to organise around pricing and valuation, based on the venue in which the contract is cleared.
"Combined with the vertical integration scenario, it's not hard to see how we might move onto a model where contracts become financial products. It might be an exaggeration to say that the swaps market will eventually transform itself into a series of futures contracts, but something between the reality as it is today and that scenario seems likely to me."
Some of the inspiration behind the Clearmatics DCN approach comes from the way derivatives markets operated in the early part of the last century; the first CCPs operated under a market mutualisation model where the clearing members owned the CCP and had governance over collateralisation requirements that the counterpart had to have.
In the run up to the financial crash, the bilateral market evolved to the point where people held highly leveraged positions without exchanging up-front collateral, using the balance sheets of the counterparts rather than collateral for counterparty risk mitigation.
Sams said: "I think there is a story here about how the market evolved and eventually became monopolised by the investment banks, because if you are going to use the credit of your counterpart for your counterparty risk mitigation, then you are going to obviously want to trade with the best credits in town.
"And who are the best credits in town? It's an investment bank, and the reason it has the cheapest cost of funding is because the government will not let it fail."
There are unanswered questions about the transparency of CCP default rules. The initial margin gets hit first and then the default fund, but if that is blown through, is it the equity of the CCP that takes the hit, or does the taxpayer come in and bail it out through a mandatory recapitalisation, or state funded recapitalisation.
Sams emphasised that the whole blockchain space is about is about distributed automation. It allows us to automate processes that organisations have to share between them. Post-trade is the obvious example, because you don't trade with yourself.
"It's an inter-institutional bit of automation that needs to be done. If we can do that automation without centralising it, without having intermediated automation; if we can share the automation amongst ourselves, that actually has some really powerful implications."
A centralisation paradox can befall blockchain design, depending on where the distributed ledger is located in the post-trade lifecycle – a point that has escaped most people talking about this space, noted Sams.
"If the distributed ledger technology (DLT) comes only at the end of the post-trade life cycle, then some other technology or technologies are automating the post trade processes up until that point, and the DLT could end up turning a distributed, industry-wide golden record into an intermediary technology service – even if the DLT is itself a technology commons, open source and everything.
"This could have the rather paradoxical consequence of actually concentrating rather than decentralising post-trade intermediation.
"When we use the term 'distributed ledger' and the emphasis is on the fact that it's a ledger like a distributed data store, the word 'distributed' is in there so it's not supposed to be a centralised thing.
"But in actual fact, the kind of architectures that people are talking about will lead to a centralising influence. Even if the golden record is something that is on a distributed data store, you still have the whole post trade process; it's quite complicated and it differs from market to market.
"But it's a process that needs to be automated and if the automation of that process isn't based on a model of distributed automation, but some other type of technology, it doesn't matter that the distributed ledger is distributed.
"You are still going to have that centralising influence and I think the kind of standardisation required to maintain that golden record is going to lead to even more centralising concentrated intermediation - if we go down that route.
"It's not just the ledger that we care about, it's the computations that amend this shared data store that need to be distributed in order to have the implications that we seem to think that they will have."