Market microstructure can be defined simply as how rules and mechanics of trading affect outcomes and, as such, market microstructure researchers have studied the ways in which the processes of a market affect transaction costs, prices, quotes, volume and trading behavior. As to why this matters, in todays complex, fast moving, highly regulated environment there have never been more inputs to study. However, traditional, high-level transaction cost analysis (TCA) doesnt go far enough in truly evaluating execution and rating counterparties.
Traditional TCA is a study of the spread between two possible prices. The difference between those prices is often called slippage or price improvement. Current TCA in its most basic form is an end-of-day scorecard used to measure the efficiency of execution on a set of orders against metrics, usually VWAP, arrival price and algo participation. When trading equities, options and futures, its important to not just look at the high-level end result but microstructure metrics that led to them.
Looking at algo routing and how MMA could provide needed insight is useful. One way to assess basic smart-order routing (SOR) efficiency is to compare the available liquidity in the market as a whole against the volume executed at the time of the trade. Typically in TCA, this is done by only comparing against the volume of trades completed over that time. An MMA analysis goes much further, studying each fill against the complete order books at each venue before, at and after each fill, providing insight into how effectively brokers were capturing liquidity, responding to market changes and causing information leakage.
One ranking often overlooked within post-trade data analytics is latency statistics, often viewed as an HFT problem, but ignoring it can give you an incomplete picture of the market and execution, which is why understanding how the data feed affected the execution needs to be a measure that is included to ensure best execution.
The Securities Information Processor (SIP) has become an industry standard and on a whole does a reasonably good job. However, the variations between the SIP and direct market feeds are much larger than many consumers would like. A relevant academic study of SIP vs. depth of book feeds can be found at http://faculty.haas.berkeley.edu/hender/nbbo.pdf.
The performance of the SIP varies by venue, trades, quote and time of day. Brokers and exchanges utilizing direct depth of book feeds instead of the SIP can make debatably better but definitely quicker routing decisions.
Its important to remember that when discussing best execution, traders must account for latencies within data sources and the corresponding time the executions take place.
Looking at benchmark strategies like IS and VWAP will give you a general rating against the given benchmarks. But as you drill down and look at the more granular aspects of trading related to fills, liquidity, venue selection, latency, data, and even effectiveness of algo strategies in relation to HFT gaming at the millisecond level, you get a clearer and more comprehensive picture.
Data is changing the way we look at everything. In trading, MMAs are the next-gen of TCA. This degree of clarity can bring needed answers to traders, shedding light on questions to ask trading partners. After all, buy-side firms are unlikely to have direct responsibilities or control over many of the micro-strategies used to execute their orders.
Drilling down into MMA provides another advantage: separating luck from good performance. The fact that one brokers VWAP beat anothers for a segment of orders in a certain timeframe may be random and potentially reverse the next given period of time. However, drilling down into data around the order, seeing how the order manifested itself in terms of venue selection and spread capture and/or reacted to toxic venues will provide a better analysis of the algos performance and a better overall picture of how the algo works in different situations.
This level of detail allows trading firms to have a higher comfort level that their counterparties are acting strictly in their interests. This is why MMA included in TCA can shed greater light on performance, overall market impact and suggest where to look for actionable improvements.
Thats no small issue to be overlooked.
David Streltsoff is head of business development for MayStreet
