Data is the lifeblood of trading and with complexity in equity market structure, analyzing the cost of trading has become an exercise in big data. “When I’m doing pattern analysis I’m looking at half a billion trades,” says Henry Yegerman, director, Quantitative Specialist, Data Analytics & Research at Markit.
Markit, known for trade processing, reference data and pricing data on OTC derivatives, has been expanding into a number of areas including TCA on the equities and FX side. In November of 2011, Markit acquired Quantitative Services Group LLC (QSG), an independent equity research firm known for providing trade analytics to buy-side institutions. Markit TCA is an independent global analysis platform that combines execution, algorithmic, venue and smart order router evaluation analytics.
Buy side firms often rely on sell-side firms to provide them with transaction cost analysis or TCA to measure their trading costs and justify their choice of brokers. While some of the main players in TCA are agency brokers like Abel Noser, Instinet, ITG and SJ Levinsohn & Sons, Markit focuses on data and doesn’t execute trades.
“Unlike just about everybody else in the space, we are not a broker,” said Yegerman in an interview in the firm’s New York office. “The real difference here is the information richness of the metrics and the analytics we use,” says Yegerman.
“Markit is essentially a data provider. Everything we do is somewhat enriched by analytical and data,” he emphasizes. Take the case of measuring the trading costs of exchange traded funds or ETFs. “Everybody measure the cost costs of the ETF against the strike price, the last price in that particular ETF security, says Yegerman. For example, “If you are trading the SPY (spider), other TCA firms measure it against the last price of the SPY.
Markit goes a step further, relates Yegerman. It created the Markit ETF Encyclopedia that provides the constituent information to all the ETF trades and issuers. “Everyday we are sending out the data on the underlying stocks, what are their weights and [factoring in] corporate actions. “We actually snap the price for all the underlying stocks that make up the ETFs and create a benchmark price for the ETF TCA,” he says, adding “Then we answer the underlying question which is should I have purchase d the ETF or the underlying basket of stocks?”
Now Markit is selling TCA to both the buy side and the sell side. It's offering “actionable items” for the buy side trader, which they can point out to the sell side, says Yegerman. This is important today since the sell side is stretched and may not have the resources to focus on all their clients, he adds.
Sell Side Has Limited Resources: Buy Side Needs TCA With the lower equity volumes and commission pressures, the sell side has its own issues and needs to focus on their core competency which is about the markets,” says Yegerman. The big issue on the TCA side is the lack of resources on the sell side. TCA can help the buy side manage their sell-side relationships in a period of limited resources by providing it with information, says Yegerman. “They certainly go out to their top 20 clients, but with the exception of a Blackrock or a Fidelity, they can’t all be a top client. Small clients are not going to get the attention, and it’s up to them to point issues out tot the sell side, he asserts.
“Right now it’s incumbent upon the buy side to become more proactive in managing their brokers,” emphasizes Yegerman. “We provide this level of granular information with which a buy-side trader can go out to the sell side trader.” For example, a buy side trader can do the analysis and find that he/she is being too aggressive on a trade. Or, the buy-side trader can point out that when using the broker’s algo, it’s deviating from the participation level the trader set. “If the buy side goes to the sell side and identifies an issue, it will get addressed,” says Yegerman.
However, a large part of its TCA business is white labeling its products to the sell side. “We can actually create a level playing field for analysis among the sell side firms so they can compete on their own particular business strengths and not who has the best TCA report,” says Yegerman.
Paying attention to data quality is critical to TCA, illustrates Yegerman. When Markit calculates TCA it doesn’t use the average trade basis. While many people that take the order allocation decisions out of the order management system (OMS), Markit built out its implementation shortfall (IS) from a trade-by-trade basis. It measures the cost of each execution against the prevailing quote. This lets Markit break out the IS into a new layer of granularity, says Yegerman. It divides IS into how much is due to the trader’s market impact, plus the momentum or timing cost of what everybody else is doing in the market. This can enable the buy side to adjust the levers or actionable times they use to check their trading strategy or to analyze different brokers.
However, Yegerman showed how best execution isn’t only about numbers. In one case, he showed that if a trader was willing to pay a higher price for a stock, due to the impact of momentum and timing, it resulted in reduced market impact. Another aspect of TCA and where this meets big data, Markit is taking in all the FIX messaging tags to figure out the broker’s routing logic. This helps buy side work with their sell-side partners on what’s the ultimate routing logic for their orders. It also supports the sell side with their routing logic in terms of the performance of the venues.
Venue analysis can be eye opening. It allows the buy side to check whether a broker is pending too much time in the dark and what is their rebate relationship with that venue, observes Yegerman.