FX Traders Demand Better TCA

Continued adoption of electronic FX trading is increasing demand for performance measurement while providing the necessary execution data to calculate it.

Nothing drives the buyside to adopt transaction-cost analysis faster than perceived malfeasance by their sellside brokers and dealers. When Bloomberg published its June 2013 story on the U.K.s Financial Conduct Authoritys investigation into alleged collusion by traders from the largest foreign-exchange dealers to fix the WM/Reuters 4 p.m. benchmark rate, it could have created a spike in interest in FX TCA.

Since local regulators in Asia, Europe and North America are conducting their own investigations, it has given buyside clients an incentive to analyze how their FX trades are executed. Theyre questioning whether they received best execution, which could be partially measured by FX TCA, said Howard Tai, a senior analyst at industry research firm Aite Group.

One would assume the buyside would have more FX TCA on their minds due to the current market investigation, said Evangelos Maniatopoulos, head of global Advanced Execution Services FX trading and FX products at Credit Suisse. We speak to our customers about TCA on a daily basis, and we can certainly say that there is a significant focus on the topic from the buyside community. Perhaps this is also driven by current market investigations, Maniatopoulos said.

The initial spike in interest actually first appeared in the second half of 2009, when the California Public Employees Retirement System and the California State Teachers Retirement System sued their custodian, State Street, for overcharging the pension funds on standard instruction FX trades. Other custodians like Bank of New York Mellon also found themselves subject to similar suits brought by other pension funds.

The perceived wisdom was that the foreign exchange markets were efficient and cheaply priced, explains James McGeehan, co-founder and CEO of FX Transparency. For most investors, who have third party dealing relationships, price discovery mechanisms and oversight processes, that is the case as trades can be done reasonably efficiently in the OTC market, but for many after the California litigation came out there was weakness that people had not recognized and not uncovered. It spurred the need to look at currency execution since it trades over-the-counter. Banks and custodians are not acting as fiduciary — theyre acting as principal, which makes them economically opposed to their client on every trade.

“We measure currency transaction as if we were managing the risk at a sell-side dealing desk, explains McGeehan. We use price benchmarks in the same way banks measure their own traders, which is the proper way to do it in terms of market risk and transaction costs.”

According to a recent Aite Group white paper written by Tai, FX Transaction Cost Analysis: The Buy Side Weighs In, the buyside trading desk was the prime consumer of FX TCA. After them, the portfolio managers and the compliance staff were just beginning to delve into best-execution policies.

FX Transparency, which opened its doors in the spring of 2009, has seen interest from every type of buyside market participant, such as pension funds, pension fund clients, asset managers and asset owners, according to McGeehan.

The adoption of TCA in FX trading, however, is still in the early stages, but we have seen an accelerated uptake since late 2010, says Tai. The use of TCA in equities has been around for more around two decades, but it really has not been done in earnest until a decade ago, when asset owners began demanding proof of best execution, which drove the asset managers to demand equities TCA offerings from service providers, brokers and others.

FX Is Not Equities

Although analysts have been using TCA in the cash equities and other listed securities markets, applying it to the fragmented over-the-counter FX market brings a separate set of challenges.

Theres no counterpart of the U.S. equities markets national best bid and offer, consolidated market data feeds or standard benchmarks in FX, said Credit Suisses Maniatopoulos. The FX market has been and remains largely a risk maker and risk takers domain. Comprehensive transaction-cost reporting is difficult to achieve when simply measuring a counterparts performance versus a bespoke risk price.

Without any central time and sales reporting platform for the FX market, any FX TCA study will have a handicap where one has to define a reasonable market environment and an appropriate executable venue for the trades currency pair and size, added Aite Groups Tai.

It is possible to implement popular TCA models, such as implementation shortfall, even though a trade or portion of its child trades might be executed on an order-driven or quote-driven venues. Without publicly listed volume metrics to which to refer, other performance metrics like volume-weighted average price (VWAP) are difficult to calculate, he added.

There is plenty of data available, but you need in-depth currency market domain expertise to properly understand the risk profile of the trade being measured from the available market data.

Sourcing the appropriate market data used to generate an aggregated bid-offer spread that acts as an execution benchmark is not for the faint of heart. FX Transparency pulls in data from approximately 10 different sources to calculate the midpoint of its benchmark bid-offer spread.

Our contracts with the data providers prevent us from identify the individual sources, but we source data from interbank dealing systems, voice brokers quotes, FX ECNs and other places, McGeehan said. The only thing we are missing is the actual trading volumes.

Similarly, Credit Suisse AES sources its trade and quote data from FX multi-dealer trading platforms like EBS and Thompson Reuters, and FX ECNs like Currenex and Hotspot FX, as well as exchanges like the Chicago Mercantile Exchange, according to Maniatopoulos.

The most difficult historical data to obtain is from individual dealers that deliver it over the telephone or via their single-dealer platforms, since it is a closed loop of prices, according to Tai. There are very few buyside firms that have the IT budget and sophistication to store, retrieve and analyze that data for FX TCA. Theres probably only a handful of firms that are capable of doing that at this moment.

Yet, without that data, any performance calculations, will be skewed since the voice-broker market is where the FX market truly exists, McGeehan said.

The Next Generation

As FX traders continue to adopt electronic FX trading and its associated use of trading algorithms and direct-market-access capabilities, the resulting execution data will let analysts dissect execution quality even further, according to Tai. Earlier forms of TCA didnt have this sort of benefit, since they could be executed over the phone without a proper time stamp, he explained. These trades could be recorded merely as part of the post-trade confirmation and settle process like a message received from ones custodian. When a client receives this information from its custodian or third-party FX broker, it makes it difficult for the client, with the help of service providers, to retrace the orders step for proper baseline comparison.

Tai sees this increased availability of electronic execution data feeding a new segment of TCA in FX that examines best execution.

Dubbed quality-of-execution analysis, or QEA, it examines an FX trades execution price as well as the state of the market before a trade is entered, the execution method used, where the order is sent, how it is executed, how long it takes for an order to be completed and how each fill performs during that period, as well as the overall performance of the trading strategy used.

If it sounds as QEA is drawing FX TCA analysis closer to TCA measurements used in the equities markets, it shouldnt be surprising, said Victor Lin, a director in the trading strategy group at Credit Suisse AES. There are different issues in gathering data for FX compared to equities, but the general concepts are similar-for example, grouping trades based on factors like size and aggressiveness. Those concepts lay the groundwork for the direction of the product, and it evolves from there based on client needs.

When Credit Suisse AES introduced its FX TCA offering in 2007, it relied heavily on lessons learned with its equities TCA product.

Weve used our equities TCA product as a rough blueprint on where we want to take our FX offering, said Lin. Over the years, we have innovated the product along those lines. If you look at Credit Suisses equities and FX TCA offerings, they are purposely similar in design-often, the people looking at execution quality are looking across multiple asset classes.

The brokers original FX TCA offering simulated the moment when a client theoretically clicked the trade-execution price and compared that to the average price a client would receive at the end of the trade.

We now look at what happens before and after the trade-the spread capture, the reversion, top-of-book spreads, depth-of-book liquidity, as well as interval TWAP, interval VWAP-and look at how the market reacted at the beginning of a trade versus at the end of the trade, Maniatopoulos said. There are a number of additional pieces of information that help build the complete picture.

Credit Suisse AES delivers these results in various custom reports and visualization tools, which brings patterns to life and helps clients understand the multiple drivers for a trades cost like trends, order size, liquidity, trading strategy aggressiveness and underlying volatility, he added.