What is driving demand for automation in the fixed income market?

There has been an increase in automation across the fixed income infrastructure and these changes are creating opportunities to increase revenue, as well as reduce operating costs.
31 May, 2022

Fixed income infrastructure has long experienced waves of increased automation, and a firm’s success in this space has often been tied to how well prepared they are to respond and leverage these changes. It is fair to say that we are in a particularly active period of change. There has been a fundamental reshaping of the traditional manual relationship-focused processes with new means for counterparties to engage, to source liquidity, to trade and to process their fixed income needs. COVID accelerated the demand for increased efficiency and control, while newer SaaS providers are continuing to ride the wave of cloud adoption. There has been a swell in the re-engineering of workflows and an increase in value extraction from workflows that had traditionally been part of the middle and back office. These changes are creating opportunities to increase revenue, as well as reduce operating costs, as the industry braces itself for a return to normal trading conditions.

Fixed income product usage is big and growing. In 2020, OTC derivatives notional grew by 5% to roughly $430 trillion outstanding. US Treasury average daily trading volume grew by 1% to $603 billion, European sovereign trading by almost 4% to $82 billion, and JGB growth a whopping 17% to around $96 billion.

Fixed Income Products are Facing Increased Demand for Automation

Fixed income product usage is big and growing. In 2020, OTC derivatives notional grew by 5% to roughly $430 trillion outstanding. US Treasury average daily trading volume grew by 1% to $603 billion, European sovereign trading by almost 4% to $82 billion, and JGB growth a whopping 17% to around $96 billion. There has been significant growth, but fixed income automation is not high. Looking at the US as an example, fixed income trading is 25% bigger than equities, but automation is estimated at 50% versus 90% for equities.

Market participants in fixed income are struggling with ongoing costs and control pressures made worse by COVID. It is not surprising that firms have also become leerier of spreadsheets when working from home. On the one hand firms are focusing on standardization and interoperability as clients value this. On the other hand, is really one word: cloud. All told, this is driving the need for innovation. On the practical front we are seeing examples of robotic process automation and on the more esoteric front, we are seeing technology such as digital ledgers, blockchain and smart contracts. From every corner the demand for automation is increasing.

A Changing Climate – LIBOR Transition

Against the backdrop of consolidation and automation, rates businesses face the sea change of no longer using LIBOR. The end of this era will have a very deep impact across the spectrum of fixed income. The shift to other reference rates, such as secured overnight finance rates (SOFR), affects products including OTC derivatives, short-term borrowing, floating-rate borrowing/lending, and other financing activities.

As the end of LIBOR approaches, it should not be used for any new contracts. The replacements for where LIBOR is used for existing contracts is expected to be in place for 2023. This is going to affect OTC derivatives and other products.

The Alternative Reference Rates Committee (ARRC)published two documents, first in 2019 then updated in 2021, where they recommended using the Secured Overnight Financing Rate (SOFR) for the USD market. The document from the ARRC mentions a couple of features which are relatively new for rates calculations, namely lookbacks and lockouts. SOFR notes and loans require lookbacks and lockouts because the payment of the SOFR coupon is calculated using daily compounding. It is paid in arrears which means that this coupon is paid on the same day as the last rate is published without delay. This lack of payment delay leads to the introduction of the lookbacks/lockouts mechanism, because when there is no delay, you have to pay the coupon on the day of your last SOFR rate, which is part of this coupon calculation, is yet to be published.

In a recent survey conducted by Quantifi, featuring Celent, participants were asked which type of rate they thought would be prevailing in SOFR notes in two years. As expected, going forward most market participants (66%) would prefer to calculate SOFR coupons in a forward-looking way i.e., similar to LIBOR when coupon is known at the beginning of the accrual period. For the two term rates in the poll, the majority of respondents (41%) believe that the “SOFR -in-advance” rate rather than the CME SOFR Term rate (25%) will be the prevailing rate in two years’ time. It is worth noting that currently “SOFR-in-advance” is predominantly used by agency bonds with the coupon calculated at the beginning of the coupon period, based on the SOFR average (can be 30, 90, or 180 day) quoted by the Fed two days before accrual starts. This highlights that even though there is a mismatch of the rate calculation period and coupon period, the convenience of the “SOFR-in advance” rate calculation makes it more attractive for bond market participants.

Uncleared Margin Rules

We are entering the last phase of uncleared margin rules. In response to the global financial crisis, regulators decided to put in place margin requirements for bilateral OTC derivatives. The goal is to reduce counterparty risk by introducing initial margin. This is causing substantial operational challenges. Not every firm in Phase 6 will be required to post margin, they will however need to monitor whether or not they are breaching the threshold that requires them to post margin. This is increasing operational effort as well as margin costs and will draw into question the profitability of using uncleared products.

The Demand for Centralized Clearing

Here, we look at US Treasuries as a proxy for what might happen to rates globally. US Treasuries are a concentrated market. If we divide the dealer activity into quintiles, 75% of all activity comes from the top two quintiles. More interestingly though, non-dealer principal trading firms are as big as dealers. Principal trading firms are highly automated, non-broker dealer firms and they make up more than 60% of electronic activity. For some product lines, such as on-the-run treasuries, it could be as high as 80% – 90% of all electronic activity, and they make up over 50% overall. So, principal trading firms are as big as primary dealers in terms of providing liquidity. Against this backdrop and in response to significant dislocations over the last few years, there have been calls for centralized clearing for outright and financing trading activities. The goal of these calls is to reduce the technical pressures, primarily settlement risk, however, it is going to allow more entrants as it will lower the barriers to entry for non-traditional, likely automated providers.

If automation is going to be necessary, what will it look like? Figure 1 outlines a high-level functional overview of a front office workflow.

Market Analysis: develop pricing understanding, patterns and algorithms

  • Lends insight into business and client performance
    • A central requirement as the industry becomes increasingly data-driven
    • Uses current and past market data and client interactions
    • Outputs include market direction signals, definition of target bid-ask spread and likely client specific bias

Price Formation: generate orders and respond to pricing inquiry

  • Based on a historical and current market view, positions, risk, and, in the case of customer inquiry, analysis of the quality of the relationship
    • Protocols include request for quote (RFQ) and request for stream (RFS) as well as placing orders and aggressing on the market in central limit order books (CLOBs)
    • Components interact with venues, single-dealer platforms, and APIs and rely on algorithms developed under market analysis

Execution: automated market and customer interaction

  • Venues or via API or single-dealer platforms
    • May execute post-transaction hedging
    • Components interact with venues, single-dealer platforms, and APIs and rely on algorithms developed under market analysis

Inventory Update: update inventory based on order executions

Risk Update: update risk profile based on position changes and, potentially, changes in market level.

Implications – Raised Table Stakes

Any firm entering the fixed income trading market or trying to gain market share faces a daunting task because of the combination of the development and maintenance required to maintain sophisticated automation and the concentration of liquidity providers. James Gorman, CEO of Morgan Stanley, called out this challenge in a 3Q 2021 earnings analyst call, where he said, “Frankly, the technology is so demanding and complex that it’s pretty hard to be a wannabe or a new entrant trying to break into the group that is dominating the global flow of capital markets.”

It is a sobering statement and now looking at what is happening in US Treasuries, Celent noted that 75% of the liquidity is coming from the top two quintiles of liquidity providers and 50% of the market activity is coming from highly automated principal trading firms. This really draws into question whether firms should be competing in all markets as liquidity providers. For a firm looking to become a liquidity provider, the stakes have been raised and they are only going to get higher. No matter how these firms put into play, they are going to have to leverage best-in-breed solutions. This means, in some fashion, firms are going to be using third party tools. Gone are the days when one would use a proprietary solution end-to-end, or a third-party solution engine. Third party solutions speed up time to market, speed up time to quality, as well as come with the side benefit of shared development and maintenance expenses. Celent believe this leads to the end of monolithic technologies.

Two areas where Celent have seen examples (for rates) of use of best-in-breed solutions is in business and sales analysis as well as venue connectivity. Understanding business and customer performance, identifying toxic relationships, creating sales opportunities from analysis of the market, dealer inventory, and a client’s past timing behavior is no longer a “nice to have.” The demand for this analysis is external as well as internal. Clients meet attempts to maintain or expand relationships with detailed data analysis about the sell-side firm’s performance as a liquidity provider. Business and sales analysis are tools that help us understand one’s business.

Clients are analyzing dealer firms too. They are trying to understand whether the dealers’ trading activity is beneficial or detrimental to them. Therefore, when dealers talk to clients about increasing activity with them/providing them greater liquidity, they should be expecting clients to be analyzing whether or not you are successful at managing liquidity. If not, the client may cut you back.

We operate in a data driven world. On the venue connectivity front, firms need to manage multiple venues and multiple protocols and there is really no value-add to managing a standalone activity to provide that connectivity. There is too much variation in protocol and too much variation in venues to make a standalone solution viable. Instead, Celent are seeing the opportunity for third party solutions to provide consolidation and transformation of the various venues into a single protocol. By letting one vendor manage all of the connectivity and then connecting to it, means firms can standardize across the multiple protocols and venues.

Path Forward – Build for Collaboration

So, what is the path forward? Environments and technologies are going to be more collaborative and cloud is going to be a key driver for this. The shift to the cloud is laying the foundation of increased interoperability. With firms using best-in-breed solutions, the next step is to integrate and bring together multiple elements and components of those solutions. Interaction, historically, has been at the end of a process, system collaboration is rapidly evolving to mid-process rather than at just the beginning or end.

Collaboration is also going to change the way we develop technology. The industry is set to see less reliance on custom proprietary, end-to-end solutions, or internal systems. In place, we will see a growing use of open source. Technical components of open source are already widely used. What we are seeing now is the advent of open source, functional components becoming commonplace. The mix of third party, open source, and internally developed components enables strategic focus of technology resources to create true proprietary advantage.

Let's talk!

Speak with one of our solution experts