Immediately after the March mayhem, I had a call with a friend who works in prime brokerage and custodian departments at a large bank:
X: These days I receive calls from our friends in other banks to check the background of our clients (asset managers).
Me: Why is that?
X: Markets are very volatile. These friends are worried if clients will settle the cash on large trades. You know that cash settles T+2?
Me: Yes, but 2 days are standard for cash settlement, right?
X: Well, these are two very long days. After a few volatile days, people are worried that someone might default.
Markets are still volatile and it is a suitable time to share our learnings around possible roles of technology in managing settlement risk.
Prime brokers and custodian departments in banks exist to serve different types of asset managers that range from pension funds to hedge funds and REITs. Some asset managers use leverage, which can be more than 10 times the actual cash deposited in their account for trading. Asset managers buy financial instruments and collateralize these to support the leverage. Prime brokers are responsible for carefully monitoring the risk for each client and ensuring that total collateral covers the losses on the client’s portfolio each day. Whenever the total value of the deposited cash (capital) and collateralized securities is less than the loan provided by the prime broker, a margin call is made. Further, asset managers and prime brokers may hedge the risk exposure with highly correlated securities by taking opposite positions. Asset managers generally deal with multiple prime brokers and custodians. The true risk exposure of an asset manager is not visible to anybody outside the firm in real-time. Large banks only see a part of their portfolio and can monitor risk on that section.
In volatile markets, things can break at multiple ends. Correlations used for hedging the positions disappear, price swings incur large losses and margin calls are triggered. However, even in these volatile markets, firms need to continuously trade to avoid further losses, because of liquidity constraints, hedging needs, and possibly serving a major client with years of relationships. Many of these trades require billions of dollars of payments in exchange for the delivery of assets. Now imagine a scenario when asset prices are extremely volatile, firms don’t have a complete picture of risk exposure of their counterparties and a firm enters in a trade to receive payment at T+2. These are the market conditions when the frontline of risk management and trading teams at large banks strenuously work to ensure that things don’t go haywire. Every prudent investor and responsible market participant wants its cash settled as soon as possible. Certainly, the creation of clearing houses has mitigated some of the T+2 risk (at significant cost).
If immediate cash settlement can save billions of dollars for vigilant market participants, one may wonder why there is a T+2 or T+3 settlement in the markets. Another question will be what does it have to do with distributed ledger technology (DLT)?
Every trade of a financial instrument creates an obligation for one party to pay cash to another party. For payments to process faster, one needs an accurate and live view of two things: payment obligations and cash liquidity.
When a trade is created, it goes through a complex cycle before getting settled. The details of a single trade may be shared between a minimum of 3 to a maximum of 6 or 7 different organizations before settlement. Note that under Basel standards, the least regulated products like bi-lateral derivatives also require a third-party custodian. Once a trade is created between two parties, it gets entered into their individual databases. Intra-day or at the end of the day, these trades get sent to other parties involved for confirmation. The technology is efficient enough to share the details of the trades between multiple parties instantly. However, the last thing one wants to do is send an incorrect billion dollar payment to a counterparty. To minimize the possibility of errors and rectify any reconciliation issues, T+2 and T+3 settlement takes place in the industry. One or two extra days between trade creation and settlement are there in case trades don’t reconcile among different parties and disagreements need to be resolved.
The immediate question is what gives rise to these reconciliation issues? Human error is one reason but that’s not the biggest factor. We have extremely reliable and advanced technology available to exchange information but how is it adequate in ensuring completely error-free multi-party workflow?
The key issue with messaging based technology is that it doesn’t ensure the immutability of data when reconciliations take place among multiple parties. A second missing factor is the immediate, authoritative acknowledgment of reconciliation across multiple parties. Imagine a scenario where party A reconciles with B and then B shares data with C. With current market infrastructure, C will separately need to reconcile with A as nothing guarantees that data received from B are accurate. There are well-defined industry regulations around trade matching and confirmation. Further, if C doesn’t reconcile with A, B will not know for another day or so about the issue. Multi-party reconciliations with fragmented systems leave room for errors that require T+2 and T+3 cash settlements. If one tries to reduce settlement time in the capital markets network using available messaging based technology, the accuracy and precision requirements will increase drastically as it approaches instant settlement. In the next paragraphs, we explain how distributed ledger technology (DLT) can play a role in expediting settlement. However, in the below diagram we share a comparison of the cost of infrastructure to enable faster settlement with incumbent technology vs DLT. Note that this chart is based on intuition and back-of-the-envelope calculations, and requires careful verification.
The most popular criticism of the application of DLT to capital markets is that whatever the technology achieves, it can be done by a database. One may consider solving the problem of reconciliations by having a single database that all the market participants can look into. However, anybody who understands the dynamics of the market would know that this gives rise to various business problems:
- Who will maintain this database? How does one ensure that the party managing the database will not tamper with the natural workflow in extreme circumstances?
- There is a non-zero probability of databases being hacked or tampered in a real-world scenario. How can we convince all the market participants to trust this centralized database all the time? If market participants don’t trust this database, they will go back to maintain their internal ledgers.
- If the solution is to have a neutral third party to manage a centralized database of transactions, this model already exists in the market. The role of a neutral third-party is played by custodians, clearinghouses, or CCPs in different scenarios. In spite of these, reconciliation issues exist and every market participant maintains its own internal ledger.
- When data is centralized with a third party, the impact of any data breach can be huge. This reduces data security.
Completely removing the possibility of errors and reducing settlement time requires all the steps from trade creation, confirmation, and reconciliations to take place on a single immutable copy of the trade across multiple parties. Distributed ledgers are effectively a single database shared among multiple parties and, secured by cryptographic tools to enhance its status to a tamper-proof ledger. This technology provides several other advantages that can address the business problems mentioned above:
- No single third party is managing the “shared database”. Every party on the network has its role to play through engraved business rules inside the smart contracts. No centralized party can have any influence on workflow in extreme circumstances.
- Each party maintains a node of DLT that is always in sync with counterparties. The trade entry, confirmation, and settlement amount calculation take place across relevant network participants. There is cryptographic verification of actions on the network and each participant can trust data in its own node.
- All the actions are legally binding and there is a detailed audit trail shared among all relevant participants. If a manual mistake happens, there is always a scope of mutually agreed amendments.
- Data is distributed among transacting participants and this drastically reduces the impact on any potential data breach and enhances data security.
- Since this technology drastically cuts down reconciliation issues, it has the potential to enable T+0 or even instant settlement in the capital markets.
Fairom and several other startups are focused on building technology that can provide an accurate and synchronized view of obligations across capital market participants. However as mentioned above, the second piece in the settlement puzzle is the clear view of liquidity. Utility Settlement Coin by Fnality and JPM coin are projects that are working on solving this problem. Independent Stablecoins could also contribute to a solution. Fnality’s white paper nicely explains how a distributed ledger technology based solution introduces efficiency in the wholesale market. While solving both the problems of obligations and liquidity is the holy grail, progress on each aspect also introduces significant efficiency in the markets. We will share some posts in the coming weeks about how solutions like Fairom create value.
Removing the settlement risk from capital markets is not going to happen overnight. While organizations have a much better understanding of the value proposition of DLT in capital markets, there are major business challenges in facilitating the adoption of this new technology. These challenges range from reimagining the existing business processes to changing the current infrastructure, which has cost billions of dollars to build. As an example, it took more than two years, once the decision was made, to move the settlement time from T+3 to T+2 in Canada on existing infrastructure. The move to T+1 or something closer to real-time was thought to introduce too many operational challenges to the market given the state of current capital markets infrastructure. In spite of these challenges, among the candidate technologies, distributed ledger technology is the most suited to pave a path for almost zero settlement risk in capital markets. If we are to develop a truly digital capital market we need truly digital tools that support real-time and immutable confirmation, settlement and event management for trades.
Thanks to Jean Desgagne and Brian O’Donnell for feedback and suggestions on the post. I also thank Ahmad Sghaier for several insightful comments around a comparison of incumbent technologies with distributed ledger technology.