Key decisions to help your implementation of FX algorithms

The number of FX algo providers has grown apace in recent years, with the amount and variety of strategy types now at an all-time high. Christian Gressel, Head of FX Algo Execution, UBS outlines the most important factors for clients to weigh up when selecting providers.

The FX market today offers an almost unlimited choice of execution strategies from a growing number of providers. Yet the increased availability of algos has made it even more essential than ever before that clients approach their implementation with a clear idea of what to look for and how to make the right selection for their individual execution needs.

One of the most important considerations any client needs to make is deciding which providers to use and for what scenarios. As with any commercial decision, trading firms should first run through a number of key questions. Not considering any of the non-algo factors, like the broader relationship, the main objective a client should have is to make a clear assessment of their type of flow and what their execution objectives overall and in particular for algos are. Only once the trading firm is clear on how they want to use strategies available to them can they start making informed decision on the
providers that will help them achieve their goal. This will of course vary depending on the nature
of the client and their type of flow.

Some clients may simply want to use FX algos as a tool to break down their larger trades into smaller child orders that are more easily executable and less dependent on deep liquidity. If however one is looking to minimise market impact and improving the overall execution cost as much as possible then an assessment of the quality of the platform, the quantitative work going into it and the liquidity available through a particular offering becomes the next important step. There are big differences between simple time slicers, which logic might even be outsourced, and a fully in house build, complex algorithmic trading platform capable of making decisions based on live and historic data.

Trading firms should look out for indications such as the complexity and intelligence of the smart order router and how it evaluates different types of liquidity. For example, the liquidity and price on the primary market might be something very different than a price shown on an ECN from a liquidity provider with last look. Either option offers valuable liquidity but being able to correctly value these differences and handle the child executions accordingly is not something all smart order routers are able to do. In this context one should also consider any type of unique liquidity, like the ability to access offsetting franchise flow, only available to a certain provider.

The second factor to look at is the quality of the scheduling part of an algo strategy. In general, the more quantitative effort that goes into how these decisions are calculated, the better the results
will be for the client over time. The outcome of any trading strategy will show a high degree of randomness if only used once. However clients looking to apply a systematic approach to the
execution of their flow will find that the quality of the algos and the smart order router used will make
a significant difference to the overall outcome.

Less can be more

The next consideration is the breadth of the offering. This should include a detailed review of how strategies available can complement each other, rather than just looking at the number of strategies available. Most clients will want to use more than one algo provider and their combined offering, if we consider each one has several tactics to choose from, can easily get overwhelming and sometimes even confusing. It is not uncommon for a client who is partnering with four or five providers to end up with over 30 different algos on their disposal. Understanding that many different strategies and their specific use cases is not an easy feat and in reality clients often find it is more practical to establish a set of core strategies, which best align with their execution style, to focus on. This approach enables a user to become much more familiar with each algo and ultimately enables them to better identify the best strategy to use in any given situation.

When it comes to UBS, for example, we have a streamlined offering for exactly this reason – to make it easier for clients to understand which algo to use in what situation. The key scenarios we aim to cater for with our offering are for clients to target an execution schedule based on volume going through the market (Tap), avoiding to pay the full bid/ask spread by posting passively or at mid (Float), timing the exact period over which to spread their executions (TWAP and VWAP) and lastly to sweep the market within a specified limit (ORCA Direct). By having kept it so simple, it is much clearer for our clients to know exactly what they will be getting when they select one of our algos.

Workflow should be a further important consideration for clients. This goes beyond the flexibility of making limit adjustments or changing the speed or aggressiveness of the executions, which is mostly the standard repertoire but ideally gives clients the option to automatically rolling a trade forward, changing from one strategy into another or sweeping the balance towards the end without having to cancel the parent order and potentially breaking their workflow.

Evaluate and review

Once this is understood, clients need to decide how to connect to their provider. For example, some prefer to have a number of single dealer platforms on their desktop, while other clients might opt to have one connection, often through a third party vendor that connects them with all their counterparties. However, not all providers will necessarily be able to offer a client their preferred
distribution channel. UBS is certified with all major vendors and clients can generally pick their
preferred option.

Clients need to further consider analytics and how these can be used to further improve outcomes.
It is now essential to look at the whole posttrade reporting and analytics suite, which can vary considerably between providers. At UBS, for example, we can connect to third party, post-trade
providers or send data directly. Yet over and above that, we offer post-trade consultancy which can
be particularly valuable to clients who are looking to fine tune their flow and their execution style
with our offering. We offer considered suggestions about what they could change or tweak, often
resulting in significant improvements to their execution outcomes.

Even once clients have gone through a comprehensive review process and selected their panel of providers and FX algo strategies, the client should ensure this selection is monitored and reviewed if necessary. Clients should be assessing the outcomes of their executions between those various providers and making full use of the post-trade analytics and TCA available. However, clients do need to have a degree of minimum order flow and a large enough sample to conduct proper analysis. This move towards a more quantitative assessment will ultimately benefit those providers that put a lot of effort into their models and into the platform and will likely foster the continued growth and development of the FX algo market as a whole.