The evolution of transaction cost analysis (TCA) is reaching a tipping point. Historically, TCA has been about forensics -- mining large data sets for conclusions after the trades have already occurred. As TCA moves forward, and into asset classes which lack transparency, the main issue becomes obtaining actionable information. The challenge for TCA is how to achieve this in the context of an exercise that historically has been about post-trade analysis.
There are several aspects to this: real-time TCA, post-trade analysis, big data, and transparency in over-the-counter markets. Real-time TCA for most is a simply a P&L chart with a TCA benchmark substituted for the value of the package. The problem is that interpreting deviations from that benchmark is virtually impossible. Real-time TCA needs to include data visualizations that allow you to see not only how you deviate, but also why you deviate. In other words, how are real-time market conditions affecting your process?
Pre-trade analytics present a different challenge. Decisions such as choice of strategy, choice of broker, and choice of venue have historically been made based on publicly available market data. But market data alone gives you no view on performance. Simply put, you cannot evaluate whether one strategy is better than another based purely on market data.
TCA should bring together the data used for forensic analysis with real-time market data, and ultimately provide a decision support tool that allows you to react in real-time. A good TCA product should allow you to answer questions such as, what is the best strategy in terms of cost? What is best strategy in terms of certainty of outcome? What broker would provide the best outcome for you given the brokers that you usually deal with?
As you move into using TCA to make real-time decisions, the challenges of big data come into play, particularly its volume, variety, and velocity. Big data is not one-dimensional. You can think about historical performance data as being an extremely large data set in the big-data sense, but with zero velocity. The issue is largely, how do you extract information? What analytics can be derived from that data to get away from the blizzard of numbers and reduce it to concepts people will understand? The velocity of data arrival is getting faster and faster all the time, and the variety of data is expanding in terms of orders being placed, strategies attempted, and order types.
One way we look at this is to combine actionable TCA analysis with a managed financial network. Networks are built, designed, and operated in ways to handle the velocity problem and the variety problem. We then marry what we’ve learned from the volume of historical performance information to derive the analytics in real-time and present them back to decision-makers.
Best execution in other asset classes is not dissimilar to the way we think about equities. Best execution is the ability to obtain the best price for institutional-sized orders, as opposed to retail-sized orders, in a complex world. A simple “best price metric” cannot be relied upon. In areas such as FX or fixed income, the same principle applies. The difference is simply transparency, and this is where TCA can add value.
For example, the FX world is virtually opaque. The ability to aggregate actionable data, such as quotes at which one might trade at an institutional level, is severely limited, because there are no centralized data feeds or actionable quote feeds for the public.
The whole issue is the size of institutional orders. For retail you can trade at a touch, but for institutions, the orders are too large. So what is the price? How do you achieve that price? You need to be able to see something -- and to access that data in timeframes useful for decision-making.
The emphasis at ITG has been to make arrangements with a variety of dealers and new electronic communications networks to bring these multiple data sources together, construct an order book for FX, and from there provide transparency with respect to the price of trading real size. As we move away from equities, we move further and further away from transparent markets, and our goal is to bring just that -- transparency -- to those markets to help the decision process.