[ A L G O R I T H M I C T R A D I N G S U R V E Y ]
( 52 %). The largest year-on-year increases were recorded in ETFs, up 13 % and fixed income, up 6 % compared with 2024. Most traders report that they use VWAP TCA( 57 %) to measure performance of their algorithms, as well as implementation shortfall TCA( 53 %), followed by liquidity capture( 49 %).
Rating algo performance The top five algorithm features identified in this year’ s survey remained fairly consistent from 2024: ease of use, customer support, anonymity, increased trader productivity and execution consistency( Figure 1). While most categories showed modest growth from the previous year, four areas declined, with speed showing the steepest decrease at-0.10. On the other hand, algo monitoring capabilities represented a significant increase at 0.22, validating the growing trend of firms looking to measure the overall effectiveness of algorithms.
Factors for algorithmic usage Respondents’ reasons for using algorithms are presented in Figure 2 as a percentage of responses from 2023 to 2025. The top five reasons are ease of use, reduce market impact, consistency of execution performance, greater anonymity and increase trader productivity, and combined, they account for about 52 % of all responses. Potentially indicating the growing need for reducing implicit trading costs, the largest increase was recorded in greater anonymity, which jumped by 2.47 % to reach 9.49 %. On the other hand, and perhaps not too surprising since the
Figure 2: Reasons for using algorithms(% of responses) |
Feature |
2025 |
2024 |
2023 |
Ease of use |
11.71 |
12.00 |
12.03 |
Reduce market impact |
11.07 |
10.35 |
10.87 |
Consistency of execution performance |
10.30 |
9.89 |
9.33 |
Greater anonymity |
9.49 |
7.02 |
8.25 |
Increased trader productivity |
9.36 |
10.17 |
8.83 |
Lower commission rates |
8.82 |
7.11 |
8.37 |
Flexibility and sophistication of SOR |
7.78 |
6.95 |
7.67 |
Better prices( price improvement) |
6.89 |
7.74 |
7.48 |
Higher speed, lower latency |
6.55 |
8.36 |
6.28 |
Customisation capabilities |
6.40 |
7.78 |
8.44 |
Algo monitoring capabilities |
4.67 |
6.23 |
6.05 |
Data on venue / order routing logic or analysis |
4.18 |
3.89 |
5.17 |
Results match pre-trade estimates |
2.77 |
2.50 |
1.23 |
race for capturing that extra microsecond advantage has certainly dissipated, the category of higher speed, lower latency showed the steepest decline at-1.81 %, pushing this factor into the bottom half of the list.
Using multiple brokers for algorithms is the norm Following in the footsteps of past surveys, we definitely see a positive correlation between a firm’ s AUM and the number of algo providers used( Figure 3). Large firms managing $ 10 billion to $ 50 billion in assets came out on top with 5.21 providers, overtaking firms with more than $ 50 billion in AUM, which recorded 4.81 providers, showing a slight decline of-0.04 from the previous year. The biggest decline came from those firms with $ 250 million to $ 500 million in AUM that saw a huge drop of nearly two providers(-1.71) to settle down at four providers. However, one could reasonably argue that the previous year’ s score of 5.71 might have been an exception for this particular AUM group, since surveys conducted leading
Figure 3: Average number of providers used by AuM( USD billions) |
AuM( USD billions) |
2025 |
2024 |
2023 |
More than $ 50 billion |
4.81 |
4.85 |
5.17 |
$ 10- 50 billion |
5.21 |
4.37 |
5.00 |
$ 1- 10 billion |
4.63 |
4.12 |
4.89 |
$ 0.5- 1 billion |
4.80 |
4.40 |
4.60 |
$. 025- 0.5 billion |
4.00 |
5.71 |
1.67 |
Up to $ 0.25 billion |
2.50 |
3.00 |
2.25 |
Not answered |
2.87 |
3.88 |
3.71 |
62 // TheTRADE // Q2 2025