The Apex Predator of Financial Market Infrastructures

Posted by

In an interview with CNBC, Anthony Pompliano asserted that “Bitcoin is the apex predator of financial markets”. Having grown at a compound annual growth rate of 200% for over a decade, Bitcoin beats all other asset classes when it comes to investment returns. 

Where does that leave Ethereum, the second largest cryptocurrency and so-called “global computer”? 

Despite also being a phenomenal investment, Ethereum is the apex predator of financial market infrastructures (FMIs)

We’ll paraphrase the Federal Reserve’s definition of FMIs as “multilateral systems among participating financial institutions used for the purposes of clearing, settling, or recording payments, securities, derivatives, or other financial transactions”. 

FMIs have been taking a lot of heat lately, but especially those involved in securities settlement systems. Back in January, Robinhood halted trading due to insufficient cash to meet a margin requirement set by their clearing house, the NSCC1.

Most people weren’t aware of a clearing house’s role in finance. The fact that a relatively unknown financial intermediary could limit people’s ability to trade their favorite stocks made it seem like these systems were broken. Crypto Twitter jumped on the opportunity to promote Ethereum’s DeFi (decentralized finance) as the solution

While I firmly believe DeFi is the future of finance, I don’t think we should automatically assume financial intermediaries are bad.

Today’s FMIs are the backbone of massive economic throughput. The Depository Trust and Clearing Corporation (DTCC), the parent company to the NSCC, processed $2.15 quadrillion in securities transactions in 2019 (DTCC), making it the largest central clearing counterparty in the world. 

Decentralized exchanges (DEXs), which use smart contracts to swap cryptocurrencies, saw only $300 billion of trading volume in comparison (Dune Analytics). Ethereum hosts a much smaller user base than traditional securities markets, so DEXs have yet to prove that its codebase can handle an equivalent amount of volume seen by today’s systems. 

But DEXs have proven that blockchain settlement works. DeFi has the power to radically transform FMIs, and the reason this transformation is needed is two-fold

  1. Many of today’s FMIs were created to solve a paper problem that no longer exists in today’s digital markets 
  2. Blockchain is the optimal infrastructure for digital markets

To better understand each of these points, let’s start with a brief history lesson on DTCC, a critical intermediary for the U.S. securities industry.

DTCC: A Short History

Prior to going digital, companies issued physical stock certificates to their investors to prove ownership. Those certificates were then exchanged between investors, brokers, and custodians to settle the purchase and sale of shares2. Trading volume grew so large in the 1960s, however, that the flow of paper became too much for Wall Street’s back offices to handle. 

Hundreds of millions of dollars of delayed securities transactions, misdirected dividend payments, and even the cease of operations for several firms were the result of this “paperwork crisis” that was experienced during that time period. 

The solution was a central securities depository. In 1972, the Depository Trust Company (DTC, note this was not yet the DTCC) was born. Its purpose: immobilize stock certificates in a vault and use book-entry delivery for securities settlement. 

The idea was very clever at the time—an investor provided their shares to a custodian, which was then deposited with DTC. If those shares were sold to another investor whose custodian also held an account with DTC, an accounting entry was booked to reflect change of ownership. This “custodian to the custodians” model greatly reduced the amount of paperwork firms mailed to each other. 

Book-entry delivery soon became the industry norm. Brokers began to charge fees for requesting physical certificates, so investors opted to hold their shares in “street-name”, allowing DTC to recognize the account-holding custodian as the new owner.  

The consequence of this shift, according to Charles Cascarilla—CEO of Paxos—is that “we gave up transparency of clear ownership and chain of title”. In a black-swan event such as Lehman’s default, lack of clear ownership can be a nightmare for customers trying to reclaim their funds3

But liquidity was more important than ownership in the 1970s, so DTC continued to thrive as an important FMI to the securities settlement system. It frequently interacted with the National Securities Clearing Corporation (NSCC), which sent instructions to move securities between custodians’ DTC accounts on settlement day. 

In the late ‘90s, a consensus grew among participants that DTC and NSCC should be integrated in a way that respected their similar—but not identical—rules and operating models. So in 1999, the Depository Trust and Clearing Corporation (DTCC) was formed as a holding company, with DTC and NSCC as its subsidiaries. 

Later (in 2003), the Fixed Income Clearing Corporation (FICC) joined as the third major subsidiary, solidifying DTCC as one of the most critical pieces of infrastructure for U.S. financial markets.

The Securities Settlement System in One Picture

Besides DTC and NSCC, other FMIs play an important role in the securities settlement system. Here’s a summary of each, courtesy of Richard Brown in his essay, “A Simple Explanation of How Shares Move Around the Securities Settlement System”:

  • Registrar / Stock Transfer Agent: works on behalf of the public company (in this case, MegaCorp) to maintain a register of shareholders and keep it up to date. If MegaCorp pays a dividend, these FMIs are responsible for distributing it. AST Financial is an example of a transfer agent in the United States.
  • Market Makers: commit to the buying and selling of shares on stock exchanges for the purposes of providing liquidity. Citadel Securities is an example of a market maker.
  • Stock Exchange: brings buyers and sellers together in a central marketplace. The New York Stock Exchange is one example. 
  • Broker: buys and sells securities on an investor’s behalf, dealing directly with the exchange. Robinhood is an example of a broker. 
  • Custodian: handles back-office functions for brokers such as accounting, custody, and clearing and settlement. BNY Mellon is an example of a custodian.
  • Central Securities Depository: holds certificates in exchange for an equivalent entry on its electronic register. DTC is the largest central securities depository in the world. 
  • Clearing house: instructs the electronic settlement of trades to DTC and custodians and steps in between buyers and sellers as a central counterparty. NSCC is a clearing house.

When Brown wrote his essay seven years ago, Ethereum and DeFi were still non-existent. Since then, we’ve been exposed to new ways of thinking about FMIs and how blockchain can be used for digital markets. 

Blockchain and the Rise of Decentralized FMIs

The paperwork crisis was solved after DTCC, but there was another back office problem still lingering: separate accounting ledgers. 

Brokers and custodians all have their own accounting ledgers to record the purchase and sale of securities. Recording the incorrect number of shares traded (or the incorrect price) could impact a firm’s P&L by millions of dollars. This is why there are entire teams dedicated to reconciling the books of their firm with counterparties.

Blockchain solves the “separate accounting ledgers” problem. Over $1 billion of crypto assets are traded daily on Ethereum’s DEXs with no reconciliation team4. This is because more than 9,000 machines running Ethereum’s blockchain must previously agree to a transaction before it’s recorded on the ledger5

No broker, custodian, clearing house, or central securities depository is involved, either. A buyer or seller simply connects their online wallet to a DEX and books a trade. The DEX’s smart contract handles clearing and settlement, while the blockchain serves as a central token depository6. MetaMask is one example of an online wallet, while Uniswap is an example of a DEX.

These new infrastructures can be thought of as decentralized FMIs (dFMIs): decentralized because no single person or organization controls them and there is no infrastructural central point of failure7.

Summing it all together, here’s a picture of Ethereum’s DeFi settlement system in contrast to Brown’s diagram:

As you can see, DeFi does result in less intermediaries, but we can’t take this model and apply it to the traditional securities system. 

First, a soon-to-be public company will likely want to outsource the issuing of their shares to a firm that specializes in blockchain technology (similar to how banks are used today to handle the IPO process). 

Second, widespread blockchain adoption will likely require the success of multiple blockchains. Company A might choose Ethereum for its strong developer network, but Company B could choose R3’s Corda because it better integrates with their legacy systems. If these platforms cannot properly talk to each other, the effectiveness of blockchain as a shared accounting ledger is lost.

What’s needed are some basic principles for interoperability so that dFMIs can function across different technology stacks and scale to the level of today’s FMIs. 

The Baseline Protocol: Blockchain Interoperability

On March 4, 2020, an initiative called the Baseline Protocol was announced by EY and ConsenSys, a leading Ethereum software company. The Baseline Protocol’s goal is to standardize the technical requirements for enterprise blockchain adoption and establish a “Mainnet”, defined as:

“A state machine maintained as a public good that is always-on, pay-as-you-go, and which strongly resists anyone gaining control over the system, locking others out of valid operations, or tampering with the established record.”

John Wolpert

A helpful analogy is to think of the Mainnet as something like the Internet: a publicly available platform that hosts a variety of applications from which some data is public and other data is not. The Mainnet would also serve as the final check when different systems or blockchains must agree on their state. 

Is Ethereum destined to become the “blockchain of blockchains”? It’s still too early to tell, but the initiative’s press release highlighted (emphasis mine), “the protocol will support tokenization and decentralized financial services on the Mainnet in a way that doesn’t reveal corporate assets or activities to unauthorized parties, and it leaves enterprise data safely in traditional systems”. 

Given that Ethereum is already host to a booming DeFi ecosystem, a strong network of developers, and the upgrade to Ethereum 2.0 promises to solve its scalability issues, I predict that Ethereum will be the “root chain” of decentralized finance. 

Here’s an example of how a blockchain-based settlement system could look with dFMIs and blockchain interoperability:

While this diagram may look just as dense as the diagram for traditional FMIs, it’s important to remember that blockchains reduce complexity in the system by promoting clarity of ownership, shared consensus of transactions, and programmatic clearing and settlement of trades, thereby reducing the number of centralized intermediaries involved. 

There are two intermediaries that weren’t in the original DeFi diagram, but which I added back:

  1. the transfer agent
  2. the market maker

If you recall, transfer agents will likely serve a critical role as blockchain specialists for companies that want to outsource the issuing of their security tokens on-chain. 

As for market makers, the reality is that they will continue to play an important role in decentralized finance, providing liquidity and arbitraging price mismatches across trading venues. As DeFi continues to evolve, however, market makers may be quite different from the ones we’re familiar with today8

Decentralized Points of Failure

If there’s anything the Robinhood case taught us, it’s that financial intermediaries are bad when they are big and centralized

Nassim Taleb, author of Antifragile, has written extensively on the dangers of central points of failure in systems:

“Distributed randomness (as opposed to the concentrated type) is a necessity, not an option: everything big is short volatility.”

Nassim Taleb, Antifragile

Consider what happened with NSCC demanding so much cash from Robinhood that they literally had to raise $1 billion in a matter of hours, freeze trading for a majority of their customers, and even sell positions for those on margin. When market volatility hits a certain  point, the survival of FMIs are threatened, triggering risk-reduction protocols that actually increase volatility for investors9.

The asymmetrical impact of NSCC on securities markets went beyond financial gain or loss—people lost trust in the entire system.

Large, unanticipated margin demands create volatile shocks to financial markets.

This is where decentralization comes in: shocks to the system are distributed across various dFMIs. Liquidity is important, but now we need resiliency

Other than blockchain platforms like Ethereum and R3, firms like Circle, Securitize, Paxos, and BitGo are helping build the infrastructure needed to move everything on-chain and facilitate the rise of dFMIs. 

If Ethereum 2.0 delivers on its promise of scalability, we may see an explosion of DeFi that pulls a majority of securities liquidity into crypto. Companies that want to fundraise via public capital will have the appetite needed to go public on-chain, forcing regulators and legacy institutions to act faster than anyone would have predicted.

Notes

1 Margin is cash that clearing houses require their participating brokers to deposit to cover possible losses in the event a broker defaults. Clearing houses match trade instructions between two brokers and step into the middle of a trade in a process known as novation. Novation makes the clearing house the new counterparty to each broker, facilitating one net settlement at the end of the day. This was a huge operational gain back when shares were paper stock certificates.

2 Where brokers primarily serve the client-facing function of handling customer orders and dealing directly with the exchange, custodians service back-office functions such as share custody, accounting, and clearing and settlement. It is not uncommon for a firm to serve as both broker and custodian. 

3 Prior to DTC, it used to be standard legal advice for investors to request physical certificates and place them in their own vaults.

4 DEX volume continues to increase, more than doubling on a 24 hour basis at the time this essay was published. For the sake of conservatism and simplicity, I decided to stick with the $1 billion figure.

5 For more information on these machines running software aka “nodes”, see Ethereum’s documentation on nodes and clients here: https://ethereum.org/en/developers/docs/nodes-and-clients/.

6 DEXs are unique in that, besides using smart contracts to execute trades, they use “liquidity pools” that investors can buy and sell against. A token holder can deposit (aka stake) their crypto directly into a liquidity pool and earn fees proportional to a DEX’s trading volume. 

7 For additional reading on the meaning of decentralization, see Vitalik Buterin’s Medium essay: The Meaning of Decentralization. It’s worth noting that online wallets don’t meet the technical requirements of decentralization. I chose to include them as a dFMI for two reasons: (1) they are a needed component for using DeFi, and (2) their operating model, which relies on the blockchain to store a user’s crypto assets, is very different from traditional custodians.

8 For more information on the evolution of market making in DeFi, check out Haseeb Qureshi’s essay on Medium: Unbundling Uniswap: The Future of On-Chain Market Making.

9 For more details on the downward spiral that FMIs can create in financial markets, see Craig Pirrong’s essay titled: Battle of the Borgs.

Thank you to the Compound Writing members who reviewed this piece: Art LapinschSam Auch, and Andy Louis-Charles.