Executive summary

Sell-side institutions have been facing mounting imperatives to limit overheads and optimize resource allocation, especially over the last ten years as margins have tightened across the financial sector.

Investment banks in particular have addressed this by upping their spending on automation technology with a view to remaining competitive against non-bank market makers who have, thus far, tended to be more agile and dynamic, allowing them to take market share from their larger institutional cousins.

While COVID-19 has seen a pronounced spike in sell-side interest in automation, a more general trend of innovation and rising buy-side liquidity demands is fueling a longer-term drive away from inefficient manual processes. The result of this is that the fixed income market is increasingly resembling the equities market in terms of its employment of these technologies.

As buy-side innovation and liquidity demands continue to grow, the sell side has little choice but to adapt. This process begins with sell side firms identifying the parts of the trading workflow that will most benefit from automation, with price discovery taking poll position. Also suited are smaller, everyday or odd-lot trades. The less reliant the sell side is on manual processing for these trades, the more time traders will have to dedicate to larger, more lucrative endeavors.

Unlike their more digitally savvy competitors and buy-side counterparties who have been able to set the pace of their shift to automation, the sell side transition has necessarily been more reactive — and therefore rushed at times.

Predictably, there is an acute human cost to this, and late-to-the-game sell-side firms have a harder job than others assuaging the fears of workforces concerned about how the implementation of machine learning, artificial intelligence and cloud-based data management could affect their employability.

These concerns have been amplified in the wake of COVID-19, as many banks have reduced their headcounts to maintain profits.

How humans and machines will co-exist on the trading floor is a matter worthy of consideration, and it may take some time before the benefits outshine the desire for caution. Bond substitution, for example, is something many sell side traders are challenged by on a daily basis. There is little if any downside to automating the process of trawling through reams of data for a replacement bond.

In-house data management is another area where automation can add value, with implications for optimizing how sell-side firms are purchasing their data, how risk is viewed holistically across different business groups, and how regulatory requirements are adhered to. An in-house data management warehouse can ensure more efficient, compliant infrastructure is built in at the

very start. While there may be cost implications in doing this, there is little reason to think that automation in this arena will be anything less than helpful to existing workflows, let alone represent of threat to human jobs.