Virtu Financial has moved to expand flexibility for its clients when looking to trade algorithmically by allowing them to switch algorithms mid-way through an order.
James Osborne
Named Switcher, the new “algorithm of algorithms” as so-called by Virtu is designed to allow clients greater optionality for an order throughout its lifecycle depending on changing conditions.
“Clients have been asking for a more dynamic approach to algo wheel implementations that solves the constraint of only using one algo strategy for each order,” James Osborne, global head of algorithmic development at Virtu Financial told The TRADE.
“Switcher multiplies the power of an algo wheel by dynamically switching between multiple algo strategies as conditions change – the same way a human trader would trade orders if they had time to monitor and trade everything themselves.”
The new capability acts as a co-pilot for traders, helping them to manage low or no touch flow. The algorithm helps traders to select the most appropriate algorithms based on trader constraints, market conditions and the order’s characteristics such as size of order, the level of urgency set by the trader, and available liquidity.
However, the defining factor of the algorithm and the inspiration for its name is a capability that allows it to switch strategies mid-flight.
Switcher continuously consumes real-time market inputs and if required can switch between Virtu’s algos – such as Opportunistic, Flex POV, Catch, Oasis – autonomously as market conditions change.
Virtu confirmed the new offering would allow traders to focus more of their attentions on high touch flow and would address time consuming challenges around selecting a strategy in light of changing flow profiles.
The new offering will allow traders to be more flexible instead of being chained to one single algo strategy that could lead to sub-optimal outcomes.
Virtu confirmed that the new algorithm has been rolled out to a handful of clients in the US, EMEA and APAC over the past few months.
“We continue to receive strong client interest in the Switcher not only for traditional algo wheel implementations but also for discretionary trading strategies,” added Osborne. “The beauty of the Switcher framework is that it makes it very easy to build new strategies targeting other benchmarks or types of flow, allowing us to expand our partnership with clients to other segments of their trading flow in the future.”
“Future areas of research include using index/sector-relative trading signals to increase the model’s conviction when making algorithm strategy changes. Another interesting idea we are exploring is whether there are times when Switcher should stop trading – i.e. the decision of which algo to use is no algo at all. Historically execution algorithms were not designed with this in mind but stepping out of the market for a period of time is a perfectly valid trading decision that humans make all the time.”