Best Execution. A widely used phrase within the financial industry, but one that has many different interpretations. Clearly, the regulatory focus on execution has brought about a multitude of new interpretations, requirements and definitions. We’ll explore the regulatory implications in a separate article but for now let’s just focus on exploring the definition of best execution from first principles.
Kicking off contentiously…
Best Execution is not necessarily synonymous with best price
What? How can that possibly make sense? Surely my best possible execution is where I have purchased something for the lowest possible price, or sold something for the highest price? Well, not necessarily. What if you were buying a new bespoke suit. You do your due diligence and visit 5 highly respected tailors, that you’ve asked for prices from before, and tell them that you are looking for a new suit, explaining that you are asking for a number of quotes. One tailor manages to provide a price that is marginally cheaper than the others and you decide to purchase from him. However, these tailors know your buying habits from previous experience and anticipate that when you buy one suit you very often come back for another shortly thereafter. True to form, you decide to buy another and once again ask for quotes from the same tailors. The prices have all gone up, and the premium you now have to pay far exceeds the ‘saving’ you made on your first purchase.
Best Execution needs to be viewed as a process
A longer-term perspective is also important, i.e. view average performance over longer time horizons as well as individual transactions. Getting a ‘great’ price on one trade may not be ideal if by doing so you push the market significantly higher, making future purchases much more expensive. Spraying orders across the market, potentially creating significant ‘signalling risk’ (i.e. providing indications to market participants of your trading behaviour and intentions) can be very damaging to your overall execution performance on a given day. We will return to signalling risk in a later article, but suffice to say, some execution methods and products can create much more risk than others, thereby imposing implicit costs. Quantifying these implicit costs and associated risks is a key current research theme.
Firms need to judge implicit costs pre-trade and take all reasonable steps to manage them
The FCA, in its recent Thematic Review of best execution, has explicitly warned on this point, noting that “Firms should measure implicit costs as part of their arrangements to monitor execution performance and review the execution quality of entities or execution venues. A trade may appear more expensive in terms of explicit costs but may be less expensive when implicit costs are considered. For example, a firm that works a large order over time, preserving the client’s confidentiality and minimising market impact, may achieve the lowest total costs (and the best net price). Unlike explicit costs, the impact of implicit costs can only be precisely assessed after a trade is completed and even then, implicit costs are difficult to quantify. As a result, ahead of a trade, a judgement needs to be made by firms about the likely implicit costs of an execution strategy and firms are required to take all reasonable steps to manage them” (TR 14/13 at p.11). The FCA further emphasized that this guidance was relevant to all “firms which execute, receive and transmit or place orders for execution, including portfolio managers.” (TR 14/13 at p.6)
Trading should allow efficient absorption of risk within the market
Further to the FCA’s point, its important to highlight that trading in a way that does not allow for the efficient absorption of risk within the market can also result in sub-optimal execution performance. For example, say you have 1bn AUDUSD to trade over the course of a morning. You decide to divide the parent order up into individual clips of 100m and start executing via risk transfer over the phone. Pleased with the price for the first clip, you immediately call again and ask for prices for another clip. The counterparty with whom you traded the first clip no longer has the ‘best’ price, so you trade with a different counterparty. The first counterparty is still trying to work out of the risk, and now another is doing the same. This creates difficulties for both counterparties, incurring losses and subsequently widening their prices to you for later clips. Waiting for the first clip to be fully ‘digested’ would have been preferable, allowing the liquidity to refresh and enabling your counterparties to minimise losses from hedging activities. Again, quantification would be helpful here, for example, on average at that time of day, how long does it take to trade 100m AUDUSD without significantly pushing the market? This would provide an indication of recommended time intervals for each clip in order to achieve best execution for the entire order.
So, what are the key features for a robust best execution process ?
Best Execution process should be Repeatable, Measurable, Flexible and Justifiable.
Why repeatable? Execution decision making needs to be based on some structure. For a similar trade, in similar market conditions, with the same objective and benchmark, there should be preferred execution methods that, based on past performance, achieve the best possible result. It is difficult to defend a random decision.
Why measurable? To be able to explain, manage and improve the process. Best execution is an evolving process, that needs to adapt to changing market conditions e.g. structural changes/new products/evolving liquidity conditions and performance. What was appropriate or delivered the best results last year, may not be the same this year. It is difficult to make such informed decisions without independent, objective measurement that compares performance on a level playing field. Such comparative analysis needs to be performed cross-sectionally, i.e. across counterparties, venues, products etc, and also temporally, i.e. how performance evolves over time.
Why flexible ? Execution is a complex process. Any given trade can have different objectives or benchmarks depending on the purpose for the transaction. For example, was the FX trade transactional, for hedging purposes or to generate alpha ? Depending on the purpose and mandate of the portfolio and execution desk, this may dictate different execution benchmarks and trading objectives. In addition, market conditions are obviously very variable, and there are indications that depth of liquidity has become more volatile over recent years. One execution method may typically work well when volatility is low, and the available liquidity is deep, but may deliver sub-optimal performance in less liquid conditions. It is important to have the flexibility within the process to allow the selection of the right tool for the job, and it is therefore key to have the full range of execution options available. What method provides best execution on one day for a given trade may be very different on another.
Why justifiable ? The focus on FX over recent years has made this aspect critical. Asset owners, audit and compliance, regulators are all paying significantly more attention to FX execution than ever before. To be able to justify a given execution, in terms of chosen method (e.g. voice risk transfer vs hitting a streaming price on a multi-dealer platform vs using an algo etc), product (e.g. which algo), venue/platform and counterparty has become an essential component of the best execution process. Clearly, justification is only possible if the other features have been incorporated. It is difficult to justify a random decision making process with no supporting data.
Firms need to ensure they are seeking out the assistance they need to achieve the best possible results for their clients.
With a looming deadline of 3 January 2018, the Markets in Financial Instruments Directive (MiFID II) requires investment firms to “take all sufficient steps to obtain the best possible result for their clients taking into account price, costs, speed, likelihood of execution and settlement, size, nature or any other consideration relevant” (Art. 27). However, neither MiFID II, nor any other related regulatory legislation, specifically defines how to achieve best execution. The result is an incredibly introspective process, wherein firms need to have a hard look at the products and markets in which they operate, and make certain they are seeking out the assistance they need to achieve the best possible results for their clients. We will explore the regulatory requirements for best execution in a subsequent article but suffice to say it is clearly an area where better information results in better execution, and where repeatability, measurability, flexibility and justifiability remain key.