Investment bank Goldman Sachs has expanded the use of algorithmic trading in its fixed income division in a move that will treble the number of bonds included in the programme to 7,000.
The bank is also planning to spread the use of algorithms from corporate bonds to junk bonds later this year.
The industry's use of trading algorithms has grown in recent years but has largely been focused on equities and favoured by high frequency trading hedge funds looking to trade at high speeds.
However both large investment banks and investors are now looking to expand their use of algorithmic trading to other asset classes in order to automate more of their activity and free up traders for larger and more complex transactions.
Goldman's algorithms trawl the bond market for publicly available pricing data on bonds which is then used to automatically generate prices for investors.
The technology has "allowed us to address [thousands of] inquiries in a systematic and automated fashion and enabled our traders to focus more on more challenging situations," said Amy Hong, head of market structure and global credit at Goldman to the Financial Times.
Goldman is not alone in its bid to automate more of its bond trading activity. Stricter regulations on the market and tighter capital adequacy rules have cut margins for dealers and investors and led many to overhaul their systems in favour of more automation, particularly around more commoditised parts of the market.
A 2016 survey by Greenwich Associates stated that 80% of US corporate bond investors in the US now use some form of electronic trading.
But while Goldman has also grown the size of its bond trading activity, becoming one of the top three participants in sub-$1m debt instruments (also known as oddlots) on the bond trading platform MarketAxess, other parts of the bond market are still proving somewhat resistant to the use of wholly electronic trading and some dealing firms are insistent that human intervention is still a necessary part of bond trading.
Michael Healy, co-founder at Millennium Advisors told the Financial Times that his firm uses technology to assist traders pricing bonds, but that execution still requires human intervention. “The way dealers win more trades, oddlots or otherwise, isn’t by adding incremental automation. You do it by simply pricing more bonds, more aggressively — the same way dealers have always gained market share."