The TRADE 67 - Q1 2021 | Page 78

[ A L G O R I T H M I C T R A D I N G S U R V E Y ]

Bank of America

Bank of America achieved an increase of 24 basis points in this year ’ s survey , raising its average score from 5.43 in 2020 to 5.67 in 2021 . Impressively , BofA has improved in all fifteen categories across the survey , with the highest increases seen in the data on venue / order routing logic or analysis (+ 0.41 ), reduction in market impact (+ 0.40 ) and customisation features (+ 0.39 ) categories , which indicates the strong direction of travel that the firm has undertaken over the past year . Despite improvements however , the US bank underperforms the survey average by 0.14 points overall . Aspects of service such as ease of use ( -0.26), access to dark pools ( -0.26), increased trader productivity ( -0.24) and reduced market impact ( -0.24) are perhaps the most notable . Furthermore , BofA scored below the category average in all categories , with the exception of data on venue / order routing logic or analysis , where its score of 5.82 came in 12 basis points above the category average of 5.70 .

The most commonly used algorithms for BofA by long-only respondents were implementation shortfall ( single stock ), dark or alternative liquidity seeking algos and VWAP . In terms of how the firm might improve its algo offering , clients commented that they would like to see more “ strategies for trading small-caps ”, whilst two UK-based large asset managers call for more “ innovation ”. A handful of clients also suggested that they would like to see more customisation features to support passive trading and liquidity provision . Seventy percent of the funds that rely on BofA for algo execution indicate they trade at least half of their portfolio value by algorithmic means . Just under half ( 46 %) of participants manage over US $ 50 billion in assets . For the most part the long-only respondents were from either European or North American markets , though there were also a small handful from Asian centers such as Singapore and Hong Kong .
BANK OF AMERICA RATINGS FOR ALGORITHMIC PERFORMANCE
Increased trader productivity
Reduced market impact
Execution consistency
Cost Speed Anonymity Price improvement Customisation
5.73 5.56 5.70 5.70 5.70 5.77 5.64 5.45
Ease of use
Order routing logic / analysis
Customer support
Execution consulting
Dark pool access
Flexibility and sophistication of SOR
Algo monitoring
Average score
5.76 5.82 5.89 5.47 5.66 5.65 5.54 5.67
KEY STATS

5.89 Highest score

( customer support )

5.45 Lowest score ( customisation )

0.41

Most improved ( order routing logic / analysis )

0.03 Least improved ( ease of use )

78 // TheTRADE // Spring 2021