Before we get to the Holy Grail of FX trade analytics – the ability to analyze execution performance while actually trading – it’s worth taking a pause to assess the current state of FX TCA as it has developed over the last five years. I’m reminded of perhaps the most famous passage in English literature, from Dickens’ A Tale of Two Cities. “It was the best of times, it was the worst of times, it was the age of wisdom, it was the age of foolishness….it was the spring of hope, it was the winter of despair, we had everything before us, we had nothing before us…” I think you get the idea.
In our opinion, most currently available FX TCA is of little use. Despite the best intentions of entrepreneurs and institutions, the products currently available that detail transaction cost analysis are at times misleading, often times self-serving, generally difficult to use practically, and almost guaranteed not to be prescriptive to improve trading decisions. In spite of FX TCA’s shortcomings, more and more institutions are contracting for TCA services because of the need to comply with a broad regulatory mandate that requires TCA to demonstrate best execution adherence.
Frankly, it is ironic that MiFID II, while promoting TCA, has up to this point spawned many new products and service providers that do almost nothing to actually promote best execution. I can spend time explaining that irony but it will not change the status quo. For a start, good TCA requires access to quality data and a reliable benchmark. If it’s that easy, why isn’t FX TCA better? A pretty simple and obvious answer won’t surprise you: data in the FX market is generally inconsistent – last look/no last look/aggregated/phantom liquidity – and there is no consensus benchmark.
Therefore, different companies providing analytics use different approaches that do just enough for their customers to satisfy the regulatory mandate that executions be “analyzed” after the fact. In fairness, there are TCA providers and tools that use better data and better benchmarks, but they represent a distinct minority of providers. And of course, the providers of FX TCA must be independent to be credible. Counterparties or algo sponsors who provide FX TCA to their clients are akin to the barber who is asked by his customer to opine on the haircut he or she just completed.
But given this broad frustration over the quality of FX TCA, it’s reasonable to concede that the foreign exchange market had to start somewhere when it came to the rollout of TCA, and as the realists like to say, “We are where we are.” So if TCA is a critical component of best execution protocol, perhaps it is useful to take a minute to cover the simple misconception of best execution.
Best execution is not best price
Even today, with all the publicity and pressure around MiFID II compliance, I am not sure that every FX market participant has gotten the message that best execution is not the same as best price. For many, that is too tall an order no matter what the regulators say. The more sophisticated portion of the FX population realizes that other elements are critical to achieving best execution. One important reason why best execution is not the same as best price is the notion of market impact. Market impact specifically measures the change in price that occurs as a result of an executed trade. We think that market impact is probably the most important post trade indicator of best execution – especially in the context of the growth of algorithmic execution. So therefore it makes sense that the measurement of market impact should be a critical component of any valuable FX TCA product. But most available FX TCA products do not include market impact information. Why not? The answer is the same answer to the previously asked question: FX TCA providers do not have reliable access to real time market information that is derived from a consistent set of data.
MiFID II’s prescription for best execution also requires market participants to perform a pre-trade market evaluation. Clearly, the quality of pre-trade market analytics has the same challenge previously mentioned regarding FX TCA and market impact measurement – access to reliable and consistent market data. FX TCA can only become an effective tool to improve the trading outcomes of FX participants if there is a direct linkage between pre-trade analytics and post-trade TCA. And that linkage can only be valuable if the data used in both the pre and post-trade environment is of high quality and entirely consistent. Assuming that can be achieved, perhaps the current inferior status of FX TCA can be elevated to a new, more relevant process that actually measures the effectiveness of trade executions against the market conditions that exist both before, at the time of and during the actual trading activity.
Why don’t we consider a user case to illustrate the value of an analytics approach that actually provides real benefits to the FX market participant. A typical buy side FX trader must execute a variety of trades on a daily basis which may comprise multiple currencies. That trader will be advantaged if he or she can observe, measure and understand the existing market conditions for each of those currency pairs. There are a number of market conditions that would be valuable to consider in that context: quality and depth of liquidity, spread development over trade size, volatility and RSI, to name a few. That information would hopefully inform the trader’s choice of a trading style: simple risk transfer, RFQ/RFS, or algorithm – passive or aggressive.
Assuming that the trader can interpret the pre-trade market conditions, he or she is in a position to make a judgment regarding execution choice. If the trader chooses, for instance, to launch an algorithmic execution, he or she will discover the success of that choice in the future through a post trade TCA report. If that report is robust, it will show the performance of both the parent order and every individual child execution, including the measurement of market impact. If the TCA report does not provide that basic granularity, it is of no use to begin with. One would hope that a robust post-trade analysis would inform the trader if their choice of execution style was positive – if their judgment was correct. Even better post-trade analysis would inform them how to improve their performance and judgment if the execution turned out to be suboptimal.
Unfortunately, we do not believe that many FX market participants have access to such robust FX TCA. Some however do, and they receive the benefit of having analytics that carry them from the pre-trade to post-trade environment seamlessly. Those FX market participants are maximizing their trading performance.
Even though most available analytics packages cannot deliver such valuable outcomes today, we are really not that far away from the delivery of in-flight analytics and the ability to empower FX traders with greater control over their trading decisions. But to be clear, the value of in-flight analytics is completely reliant on the quality of pre-trade market data and ultimately on post-trade analytics which evolves into an intelligent analyzer of trade performance based on the market conditions that existed before, at the time of and during the trade execution.
The value of in-flight analytics, when readily available, should be uncontestable. Buy side institutions are directional traders, typically executing FX orders linked to portfolio manager trading decisions. Those decisions often generate larger sized trades. While algorithmic strategies and executions offer many favorable elements, the time associated with such executions introduces market risk during the life of the trade. If an FX trader understood the market conditions immediately prior to the launch of an algorithmic trade, he or she could project an expected performance based on a statistical analysis. If that same trader could measure market conditions continually during the algorithmic execution, he or she would be able to follow whether the actual child trade performance matched the pre-trade statistical prediction. If market conditions changed during the execution, the trader could pause, accelerate or change algorithmic strategies. Such in-flight analytics would allow the FX trader to improve his or her trading outcome immediately by employing tactics that would lower market impact or take advantage of a change in market conditions and suspend the algorithmic execution to execute the remaining amount, for example, in an immediate risk transfer trade if optimal.
Data creates opportunities
The data that supports in-flight analytics and is generated from it will create opportunities to build intelligent execution capabilities. If this is a vision of the future, when will this future arrive? The answer is sooner than the market expects. In-flight analytics requires access to reliable market data, the ability to process that data in real time, and to apply that analysis in the context of actual trade executions. The natural provider of such information may be a banking institution, but remember my barber comment from earlier in this article. The best provider of such information is an independent third-party who has access to the elements listed above. The good news is that a few parties are already well on their way to delivering the promise of end-to-end trading analytics. The beneficiaries will be buy side institutions and their ultimate customers because achieving lower market impact will save money and support best execution. That is a future that will challenge legacy execution practices and elevate FX TCA to a meaningful tool, instead of a report that generally finds its way to the circular file.