Using Fixed Income Algos to enhance traditional trading and workflow

08/19/2021

In the run up to Fixed Income and FX Leaders Summit APAC, we managed to speak with Jeremy Son, Vice President - Fixed Income Trading at AllianceBernstein. He shares with us on the topic of Fixed Income Algos in a bid to enhance traditional trading and workflow. Ultimately, the fixed income desk of the future is about data and automation to bridge the gap between asset classes, desks and ecosystems.


What exactly do you mean by “fixed income algos”?

The term "algorithm" refers to a method of applying a predetermined set of rules to handle incoming data or a given situation, typically in the context of computerized, automated processes. In the recent years, we have witnessed meaningful developments in the usage of fixed income algorithms in intelligent pricing of securities (algo pricing), automatic execution of orders that meet predefined criteria (auto execution), automatic response to request-for-quotes (auto responder) and parcel trading of larger standing orders (tactical parcel execution).

This is a huge area of growth for the future of fixed income trading. I believe that a meaningful portion of a buyside trader’s day-to-day tasks can be emulated by an intelligent machine, which in turn enables the human trader to move up the value chain and spend more time analyzing market technicals and handling sensitive orders in today’s highly volatile, illiquid markets.


What are the benefits of fixed income algos?

At the risk of oversimplifying, I would like to summarize the main benefits of fixed income algos into the following three aspects:

First, workflow automation - mundane trading practices can be efficiently handled by a machine much more quicker and accurately than by any human trader. Trader performance is often measured against arrival price - the price at the time when an order arrives at the desk. We should not underestimate the importance of faster execution, which can add tremendously to the overall execution quality and not to mention the amount of time saved by auto execution on the part of human traders who now have more time to analyze market technicals and devise the next generation of algo processes.

Second, transaction cost reduction – every investor, whether institutional or retail, has experienced and understands the burden of transaction cost, which eats away from every penny from the hard-won profit. In this sense, algos are truly empowering, because they allow us to be liquidity providers instead of liquidity takers, thereby taking bid-offer spread from the market instead of the other way around. Tactical parcel execution also allows us to break up a large order ticket into smaller sizes that minimizes bid-offer spread.

Third, alpha generation – this is an area where algos get really interesting, but admittedly it is still in it’s early days. We can take the strategies developed by successful tactical traders from other asset classes like equities and futures, teaching the machine to buy the dip or sell the rally in a trending market or dial up the aggression level in a momentum market. We can introduce factors like moving averages, volume spikes, RSIs to enrich our algo intelligence.


Compare and contrast fixed income algos with equities and FX?

Fixed income algos are lagging their counterparts in equities and FX by at least about a decade. The roadmap for fixed income algos therefore is relatively straightforward, as I see the development trajectory taken by equities and FX as a model to be emulated.

Within the buyside, fixed income traders have focused on automatic execution of low touch orders as a low hanging fruit, while equities and FX traders now navigate through a wealth of sellside algos that provide them with a sophisticated set of parcel execution strategies designed to capture centralized liquidity with a degree of internal matching and alternative liquidity sourcing. We are hoping that forward-thinking buyside firms like AllianceBernstein can take the charge in adopting algo innovation amid fixed income’s decentralized trading landscape, as we have the benefit of seeing prices from the majority of the market, whereas the sellside liquidity may be somewhat limited to their proprietary client flows and interdealer channels.


What are the major obstacles for developing fixed income algos?

Fixed income marketplace suffers from a number of factors that disadvantage its participants from embracing electronic algos compared to their equity and FX counterparts. First and foremost, the culture among fixed income traders tends to be relatively more conservative and resistant to technology. We would need to see more innovation in trading technology and openness to embrace change on the part of both the buyside and sellside. Another major drawback is the lack of a centralized exchange where participants can buy and sell publicly and anonymously. Electronic algos would thrive with such a concentrated liquidity pool thanks to better price visibility and reduced information slippage. That being said, the fixed income market has evolved in the past several years thanks to a rapid growth of digitization, most notably in all-to-all trading and portfolio trading, which have laid the foundation for further progress on the algo front.


Has electronic transformation for fixed income algos been hampered or accelerated by the pandemic trading environment?

At first I thought the pandemic situation would be a short-lived phenomenon presenting an extreme stress to our technological infrastructure and remote-working capabilities. But as time passed by and a return to normalcy proved to be painstakingly slow, I realized that the changes to people's behavior and mindset were no longer temporary, but have become permanent. On the whole, the asset management industry is now more alert than ever about technological enablement. Likewise as a firm, AllianceBernstein is more focused than ever about leveraging technology to enhance the way we operate and differentiate ourselves amongst the crowd. Fixed income trading has taken a similar path, and we are seeing a tremendous uptake of interest in algo solutions both inside and outside the firm.


What is AllianceBernstein doing on the fixed income algo front?

AllianceBernstein is a proud leader of electronic trading in fixed income. After introducing the state-of-the-art liquidity aggregator ALFA about five years ago, we have continued in our journey to pioneer the next-generation tools, such as bot-to-bot trading, portfolio trading and zero touch trading. Most recently we launched AbbieX, which is an OEMS (Order and Execution Management System) capable of releasing orders to multiple venues and dealers simultaneously in order to maximize breadth of liquidity and lower transaction cost.

Looking ahead, we are focused on developing fixed income algos, particularly an algo pricer and intelligent algo trader. Stay tuned on this front, as we hope to have more exciting news soon!