This document discusses decentralized financial market infrastructure (dFMI) as a way to recreate the utility of centralized financial market infrastructures (FMIs) like central counterparties and payment systems, while reducing their systemic impact. It notes that previous "blockchain" projects have often recreated intermediaries rather than eliminating them. The document discusses how dFMI could automate processes using distributed networks instead of client-server architectures and intermediaries. It provides an example of Clearmatics' work developing the Utility Settlement Coin (USC) for post-trade settlement using distributed automation technology.
2. Disclaimer
● These slides contain names of specific companies. This is
for illustrative purposes only and is not intended to be seen
as an endorsement.
● The contents herein does not reflect the views or opinions of
anyone else but myself.
● I am not a lawyer.
6. Ground to cover
Why are regulated institutions interested in blockchains and
financial market infrastructure (FMI) and why have most of
the proposed designs recreated intermediaries?
This presentation discusses some of the history and nuance
of what has become vaguer terms over time (e.g., “DLT” as a
marketing term).
7. dFMI
This presentation includes a few of the problems with mono-implementation
architectures. It will briefly cover what FMIs are and what some of the fintech
participants are claiming they can do to change or supplement FMIs.
It also presents one model that bridges the research behind distributed,
automated processes with the requirements that regulated financial institutions
must fulfill.
The new term for this is decentralized FMI (dFMI), an acronym first coined by
Robert Sams and Rhom Ram through the evolution of the USC consortium
incubated by Clearmatics.
9. According to the PFMI, the three main infrastructures are
1. Central Counterparties (CCP): perform netting and facilitate value,
collateralization pooling, mutualization of risk, provide anonymity and
ensure that delivery-versus-payment (DvP) takes place as promised
2. Central Securities Depositories (CSD): beginning in the early 1970s,
these were set up during the transition from paper to electronic with
mainframes as a means to electronically track ownership of
immobilized securities
3. Payment systems
Examples: DTCC, CLS, Euroclear, ICE, CME, ACH
10. FMIs live in the background, are socially useful but
due to how they are setup and concentrated have
become systemically important
11. What if...
… there was a way to recreate the utility that specific FMIs
provide, but reduce and/or remove their systemic impact?
12. But just use a database, right?
Ironically, the quip that some maximalists use to criticize
“enterprise” efforts would actually lead to further
concentration and single-points-of-failure; e.g., their advice is
bad for society
Recall that FMIs to-date are governed and operate through a
singular entity and if the goal is to reduce systemic risks, why
would you advise participants to continue intermediating the
marketplace which further concentrates risk?
13. Previous attempts at creating mental models
Back in 2015, Robert Sams first coined the term
“permissioned” chain to describe a network with specific
attributes
17. But the “DLT” term became misused by many
different special interest groups and is mostly a
meaningless marketing term today
18. dFMI
- Information sharing can securely take place in full
compliance with the law while simultaneously enabling
market participants to mutualize infrastructure that would
otherwise be run by a single trusted party.
- Some of the services that intermediaries currently provide,
can potentially be replaced and/or augmented by
decentralized infrastructures or decentralized financial
market infrastructure (dFMI).
- dFMI also enables a change in business structure where
a re-alignment of incentives can take place such that the
firms taking the risks fully bear the consequences of these
risks.
19. And what kind of non-intermediated network can
enable secure sharing of facts and records?
In theory: B-words!
20. Stereotypical B-words
At a high level, it is a network that has:
- Global state
- Peer-to-peer validation (no intermediary)
- Containers (blocks)
21. B-words in action
A short list of live anarchic B-words that have these three characteristics:
● Bitcoin
● Ethereum
● Monero
● Cosmos
● Zcash
22. And what about that “enterprise” noise?
Note: not an endorsement!
23. Non-exhaustive list of “startups” focused on capital
markets
● Axoni
● Symbiont
● Digital Asset
● R3
● SETL (under “administration”)
● Peernova
● Chain (Interstellar)
● CobaltDL
● Rise Financial
● Clearmatics
● Peernova
● ConsenSys Solutions (PegaSys)
Note: this does not include consortia and “BaaS” which are separate topics
24.
25. Q: Are these “vendors” actual vendors?
A: Some want to be perceived that way...
Are their end-products commoditized?
Nope.
How are they unalike?
Great question!
26. Are some of them involved in chainwashing and
“blockchain” theater?
Most definitely!
Are any of them “brickstring” technology?
Yes!
27. Confusingly called “enterprise” and “DLT”
DLT is kind of a meaningless term now, morphed into a marketing phrase
"Something something enterprise something something DLT"
It often encompases networks that do not have global state and are not peer-to-
peer
Examples are distributed messaging systems: Fabric (from Hyperledger),
GSL (from Digital Asset), or Corda (from R3)
That's not what the original meaning for DLT was but no point in dying on that hill
Still worth pointing out the etymology, history, and nuance
28. A later presentation will dive into those companies
and their efforts, so let’s take a step back and look at
market structure in general
30. Visible intermediaries
On the anarchic (public) chain side, we see very large intermediaries in the form of
deposit-taking organizations and custodians typically referred to as cryptocurrency
“exchanges” and “hosted wallets”
In a given day, a supermajority of on-chain transaction activity is movement from
one intermediary to another, most of whom have terms-of-service and KYC / AML
/ CTF obligations (e.g., permissioned endpoints)
And because nearly all of the miners / mining pools are identifiable, PoW chains
have turned into very expensive networks in which most participants are paying
for but not necessarily receiving “permissionless’ness”
31. Mono-implementation = intermedization
Even if code is nominally open-sourced, if network participants depend on one
specific group or vendor to maintain the codebase, you effectively have a single-
point-of-trust and a single-point-of-failure… they have intermediated the workflow
Why is this a problem?
(1) The group or vendor could use their monopoly as leverage in negotiating
roadmap changes
(2) The network could suffer from a bug or hack and if every validator uses the
same exploitable code, this could freeze the chain and even partition the
network
e.g., if Parity has a problem and geth doesn’t, the network can still
operate
32. More visible intermediaries
On the archic (“enterprise”) chain side, networks are frequently designed with the
assumption that identified, regulated institutions will operate validating nodes
and/or participate in the consensus building process
Because there is no Sybil problem, then there is no need for PoW
But in nearly every “enterprise” chain deployed thus far, most of the network
sponsors have created a single-point-of-trust and single-point-of-failure: a specific
vendor who operates the network and/or maintains a licensed form of the
codebase
If you have to rely on a single participant to run, manage, administer, and/or
govern the network, then you arguably don’t need to use a blockchain
33. Fundamental flaws with “enterprise” today
From a governance perspective, having a single company manage or administer
a network is antithetical to having a blockchain which definitionally is a
disintermediated network
- This gives the company or group an outsized influence on decision making
Why is it common?
Because decisions have to be made and someone needs to do it; even some
nominally decentralized networks have centralized coordinators for decision
making (e.g., Bitcoin Core). But may be other ways to handle (un)structured
“governance” (see Cosmos, POA)
34. Cont’d
From a business model perspective, most of the “enterprise” companies are
following the play-book of the traditional vendor-driven model (Oracle, IBM) which
involves licensing software.
The unit economics and revenue models are known and easy to comprehend.
But this typically recreates the same vendor-dominated tech landscape we have
today, market structure is left unchanged. And it can result in vendor lock-in which
impacts the governance of the protocol(s).
(e.g., vendor could use monopoly status as leverage)
35. Cont’d
From a technological perspective, having a single company manage or administer
a network and/or maintain a single implementation is a security hole
- If the company is compromised then this could result in the same issues
and abuses that maintainers of databases have
- At the protocol layer, if the company has no competition, then there is a
potential “Hold-up” Problem and the network can stagnant.
Why is it common?
Because it is familiar territory for investors, entrepreneurs, and developers
36. But what about consortia?
In terms of governance of protocols and specifications there are key differences:
For instance, both Hyperledger and EEA are non-profit organizations in which no
single entity has a permanent seat on the governing or TSC board; they also
provide a specification in which other implementations can be built on.
In contrast, the Corda Foundation is currently sponsored by one for-profit
organization (R3) and R3 occupies permanent board seats (2-of-11); the
enterprise version of the Corda “software” is proprietary
And there are many other consortia that lay somewhere between.
38. What is a solution?
How about: changing the market structure by changing the business model
And how can that be done in regulated capital markets?
By using technology to automate the processes that previously used
intermediaries and client-server architecture to accomplish
39. Clearmatics
● London based Tech R&D company founded
in 2015 focused on core infrastructure.
● Employing 40+ staff with deep expertise in
financial markets and blockchain engineering
● Developing distributed automation technology
for capital markets
● Provide enterprise software platform & tools
based on an Open Source Model
● Flagship project is the “Utility Settlement
Coin” (USC), a digital cash instrument for
settlements & payments
40. USC
● USC is an asset-backed digital cash instrument
(using a segregated account at a central bank)
designed by Clearmatics for use within global
financial markets
● USC is a series of cash assets, with a version for
each of the major currencies (USD, EUR, GBP,
CHF, etc.) and USC is convertible at a parity with
a bank deposit in the corresponding currency.
Benefits include:
○ Improved capital efficiency
○ Reduced settlement and systemic risk
● Launched in partnership with UBS in September
2015, the conclusion of Phase II that validated
the potential benefits of USC paved the way for
the introduction of new partners in Q3 2017 with
17 consortium members in 2019.
41. USC
Benefits
● Instant finality on settlement
● 24/7 to be compared against RTGS
● Bypass clearing houses and
intermediaries, greatly simplifying the
model
43. Appendix
Interrelated topics include:
● Central bank digital currency (CBDC)
○ “Money flower” model
● Central bank digital account (CBDA)
● Fedcoin
● “Narrow bank”
● Settlement finality
● Commercial bank credit risk
● Sovereign credit risk
Editor's Notes
February 2014
https://www.youtube.com/watch?v=VDRYZ122mXA
August 2016
https://www.youtube.com/watch?v=9mVcWps1VQ0