The Finance Hive Global Pulse: FX Algos & Automation

May 2023 in Industry Reports

The Finance Hive recently partnered with Bloomberg to produce a fascinating report that pulled together quantitative insights from survey responses obtained from 52 buy side heads of desk which were then aggregated and analysed. The full report provides perspectives on how automation has impacted the role of the buyside, how FX trading desks are balancing technology with the human touch and where algos sit within more automated workflows. It is some of the key findings from the last part of the survey that we are focusing on in this edition of FXAlgoNews.

The need to manage volatility and the cost effectiveness of algos has caused a boom in buy side usage. However, a lack of clear differentiation among providers has left the buy side with difficult decisions around which ones they should use, and how they are best put into effect. In total, 92% of buy side surveyed made use of algos for at least some of their trades with only 8% not using algos at all for any of their flow. 27% of the group are only executing up to 5% of their flow via algo. In contrast the second largest group at 15% are executing as much as 60% of their flow via algo, closely followed by 10% of the group who sit between 20-30% of their flow. Overall US respondents executed less via algo with the average percentage of flow equalling 25% whereas the European cohort are executing on average 35% of their flow via algo. US asset managers also executed up to 10% more of their flow than their hedge fund counterparts, and this is mirrored in Europe which displayed similar results.

Expected change to flow executed via algo

36% of those surveyed expected the amount of flow executed via algo to increase over the next 12 months, with 59% expecting their flow to remain the same, and a very small group at 5% expecting their usage to decrease. For those expecting a decrease, respondents cited concerns as to market volatility and a perceived loss of buy side control as the major factors influencing this decision.

The majority of those expecting algo usage to increase (50%) came from the group currently executing 0-5% of their flow via algo. This shows a clear correlation that once the buy side begin to execute via algo overall usage and percentage of flow allocated increases in turn. This is further shown by the fact that the remaining 50% in this grouping currently sit at 20-30% and 50-60% respectively.
40% of those expecting their algo usage to remain the same came from the group that already utilise algos for over 60% of their trades, with the remaining 60% falling between 50% and 10%, with the majority executing up to 30% via algo. None of the respondents placed algos as the top investment priority for the desk over the next 12 months. Ranging from not at all to a large focus 41% of respondents stated that their firm had some focus on algo investment, 29% a slight focus and only 18% a large focus with the remaining 12% of stating that there would be no investment in algos at all.

Types of algos used

56% of participants utilise either a dark liquidity seeking or implementation shortfall algos, with only 13% making use of VWAP and TWAP algos respectively. This clearly shows a buy side bias toward algos that enable them to minimise market impact and information leakage.

Satisfaction with providers

Interestingly on a scale of 1-5 with 1 being the lowest level of satisfaction possible 52% of those surveyed were very satisfied with their algos, rating their providers as a four out of five. Only 4% voted that their algo providers were terrible with 9% giving five out of five and 35% being neither satisfied nor dissatisfied with the service and performance they received. This shows a clear buy side opinion that current market provision of FX algos is of a high standard. European respondents also seemed happiest with their providers; with an average of rating of four out of five as opposed to US respondents’ three of five.

Top considerations when choosing algos and providers

A reduction in market impact was the most important consideration noted by those surveyed with 57% of participants selecting this as their most important area. There is then a steep drop off with the next major consideration being execution consistency and price improvement at 11%.

Dark pool access is the second most important consideration when choosing an algo provider for respondents. This supports earlier analysis that the buy side are looking to gain access to alternate sources of liquidity with 31% of the group using dark liquidity seeking algos. Ease of use, order routing logic and increased trader productivity were joint third for the survey group when considering which algo provider to use with 13% choosing each respectively.

Algo improvement areas

The largest area for improvement noted with 41% of the group was for their algos to have an easier integration with their existing techstack. This is not a surprising result as ease of integration and lack of disruption to workflow are primary concerns for trading desks.

The second largest area for providers to improve upon is to provide greater levels of transparency to the buy side with 31% of buy side heads citing this as a key development area. The third most important area for development cited by the group was for there to be clearer differentiation between algo providers.

Roadblocks to onboarding more algo providers

When asked what the largest roadblock to onboarding more algo providers was participants overall said that their typical trade size simply did not justify the use of more algos and interestingly in Europe, the group on average stated that other projects were taking priority as their main blocker.

A potential explanation here could be the already relatively high level of satisfaction across both the US and Europe leading to less of a push internally to onboard more providers – if it isn’t broken, why fix it? Previously cited concerns with workflow integration are also a reason for many. This perhaps is a simple fix for brokers if they are able to differentiate themselves from the competition more as previously requested from the buy side, and explain how they can work within individual desks’ workflows. For more information about this report please