We think it makes sense to tackle some topics this year that challenge common wisdom and hopefully improve the intuition of buy side traders who want to achieve best execution outcomes. In this article we want to take on an interesting debate: what’s the most significant indicator of best execution – minimization of spread or market impact? Clearly both indicators are important, and how they are measured is critical if the buy side hopes to improve their trading choices. And with the continuing growth of algorithmic execution by buy side institutions, market impact measurement is more important than ever.
Definition and Context
Spread is recognized as the price paid for a simple risk transfer execution, always relevant when a trader employs an RFQ or RFS. Many factors, both fundamental and technical, can influence the relative size of the spread between bid and offer prices. In theory, spread should increase in an arithmetical pattern in relation to transaction size. Trading at a tighter or wider spread can occur when the currency is over-bought or over-sold. Given the opaque nature of the FX marketplace, it is critical for the buy side trader to know the market’s condition before he or she trades and have pre-trade analytics that are independently derived and based on market data that recognizes such market conditions.
Market impact is the change in price that occurs as a result of a trade execution. Because the FX market is so large and diverse, it does not have an easily calculated market beta measurement. So it is a certain challenge to know the impact of any individual trade. However, with the onset of algorithmic trading and its increasing use by the buy side, the measurement of market impact has taken on a very important significance given the alternative execution strategies that the buy side trader can use, including a simple risk transfer for the whole size of the trade that was broken into pieces by using the algorithm. The measurement of the price changes during the algorithmic execution may indicate the adequacy of the algorithmic strategy chosen as well as the quality of the liquidity pool where the individual trades are occurring.
Spread – Sometimes it’s complicated
While tight spreads are universally coveted in the context of best execution outcomes, there are a variety of factors that can both promote tight spreads and cause negative market impact for the buy side trader. It’s worth understanding this paradox. Below we have described a few different scenarios. In these scenarios, the factors that tighten spread promote negative market impact.
High frequency traders are very active in low volatility markets. A tight market price range is the optimal environment for them to capture spread as often as possible. To succeed they have to buzz around the top of book. But their own market making algorithms will certainly react if they get hit too often or too fast. When this happens, they are likely to do two things: step away from the market and simultaneously try to benefit from the information they are consuming which will lead to potential negative market impact for everyone else trading in that market.
Perhaps a worse scenario for negative market impact occurs when tight spreads are the result of last look liquidity.
Market makers are incented to offer tighter spreads if they have the ability to reject a trade. Once a trade is rejected, the information regarding the buy side trader’s intention is known and the rejecting counterparty will adjust its pricing with that knowledge. The counterparty will likely maintain the same tight spread but gap its pricing to reflect the information gleaned from the rejection. Negative market impact occurs as well since price has moved even without an actual trade being executed. That is a terrible scenario for the buy side trader.
Certain venues present tighter spreads by aggregating liquidity. While this approach can provide a tighter spread picture, it also comes with several potential negative outcomes. In such a marketplace, if the buy side trader experiences a suboptimal execution, it will be impossible to determine the cause of the problem. Aggregated liquidity creates scenarios where multiple top of book prices may emanate from the same liquidity provider, appearing on different venues. An algorithm will of course not know that and will therefore hit the same liquidity provider on several different venues in smaller amounts. This activity will generate a signal to that liquidity provider who will adjust pricing accordingly and create adverse market impact for the buy side trader. It is entirely likely that the same buy side trader would have caused less market impact by executing in a single pool with good distribution across multiple liquidity providers.
So what’s the answer?
We are certainly not suggesting that spread evaluation is not a critical element of pre-trade analysis. Of course it is. But the buy side trader should not miss the forest for the trees. As the buy side embraces the benefits of algorithmic execution, they must do their homework to understand the underlying liquidity pools that the algorithm is facing. We cannot overstate the importance of this responsibility. Merely relying on a tight spread as a sign of good liquidity is a simplistic approach that could result in suboptimal trading results.
At Cürex, we believe that understanding market impact and taking all steps to reduce such impact is the most critical, holistic factor in achieving best execution outcomes for the buy side. FX market makers are in the business of making money. The buy side cannot rely on one observation – a tight spread - to conclude that they are executing in a good market.
Our guiding, best execution principles are to trade anonymously and avoid last look liquidity and aggregated liquidity pools. By shunning this advice, the buy side will harm itself through signaling and information leakage. These factors will cause negative market impact without regard to how tight a spread appears. And negative market impact in the context of an algorithmic execution costs the buy side money.