Both the algo and TCA template introduced under the last FX Global Code review were highly beneficial and have contributed to more informed discussions and questions from our clients. In particular, the way it has generalised some quite specific and advanced concepts would not have been reached without the standards and the recommendations to publish such information.
Having algo disclaimers on your website is an excellent start, but we should not expect clients to browse through their liquidity providers websites. Making sure that the salesforce can spread that valuable information to clients during meetings and conversations is key to continue enriching the dialogues and helping the industry in providing greater transparency.
The algo discussions we have with clients can be divided into two main categories. The first category is how our algos work and the different parameters that a client can choose. This even encompasses the logic we use to select our venues and what we look at in the execution logic in order to always strive for the best possible outcome for the client. When we evaluate counterparties for trading, there are a number of metrics to look at. This can include:
- Fill ratio – this statistic tells us if the prices that are observed are actually tradable. If the fill ratio is low, meaning that the prices are often not tradable, the expected algo transaction cost might be too optimistic.
- Cost-of-cancel, defined as: “The price movement in the second-best market (B) from the time an order was sent to the best market (A), cancelled and re-sent to Market B.” Where fill ratio tells us what the true transaction cost estimation is, cost-of-cancel tells us the estimated opportunity cost of getting the priority between venues wrong. Fill ratio and cost of cancel are metrics trying to describe what happens at the time of execution.
We also want to understand what happens in the market after we execute: Does any venue show signs of market impact? If the executed amount ends up at a counterparty that would immediately hedge this flow on a lit market, other market participants might pick this signal up and thus change the way they quote in the market, anticipating even more flow.
This is, of course, counterproductive to what the algo is trying to achieve and is monitored with great care. We minimise market impact by internalising the Scandi flow through our unique Scandinavian and international client franchise and carefully selecting the venues on which we trade. In other currency pairs, this careful selection is to make sure that we find counterparties that are internalising the flow, or venues with the lowest market impact, and thus would not affect the markets.
The second category of discussion is about a specific execution. In these conversations we are going to look more at the external factors that can impact the choice of the execution. Looking at a concrete example with the help of data gives valuable insight to our customers about the execution they are considering. It is some of the information that a pre-trade TCA can provide to you with. A usual question from clients is: “When should I use an algo?” These metrics help clients make that informed decision.
For example:
- What is the current spread in the market and how does that stand against historical spreads? A wider spread would advocate a more passive algo.
- What is the current risk transfer price and how does that stand against historical risk transfer prices? The higher the risk transfer price is, the more sense it makes to use an algo for your execution.
- What are the liquidity and volatility conditions? The lower the liquidity, and the lower the volatility, the longer time you might want to select for your algo execution.
- Are there any upcoming economic events that we need to take into consideration?
A lot of information needs to be considered when placing an algo. As we use the same algorithms for our own trading as the ones we offer to our clients, we do have a great history of executions to draw conclusions from. This, together with the expertise from our sales giving color on the current market conditions, is one of the success factors of the SEB execution algos.
