Concerned about how to generate-and preserve-above-market returns, fund companies are now arming their portfolio managers and trading desk combatants with a new tool to boost performance.
It's called transaction cost analysis, or TCA, for short. While it sounds like it focuses on cost, it really tracks results against benchmarks. Of particular importance: How far away from a target price a transaction was achieved.
"To make better decisions on their trade execution-such as which brokers to use, which algorithms to use and even determine just how effective communications are between the portfolio manager and the trading desk- buy-side firms are beginning to rely more heavily on transaction cost analysis," said Jon Fatica, managing director of Investment Technology Group (ITG), a brokerage and financial technology firms servicing asset managers.
TCA comes in three basic flavors-pre-trade, post-trade and real-time, also known as at-trade. Specialized software suppliers such as ITG and Ancerno as well as leading execution brokerages offer one or more of the three. So do brokerage firms such as Credit Suisse, Goldman Sachs, Knight, Pipeline, SJ Levinson and UBS, to name a few.
The TCA process tracks explicit trading costs, such as commissions and spreads. But also monitored are implicit costs such as the market impact of a trade, its timing, the gain or loss in delay, liquidity risk and price volatility, all of which have been growing as the result of the influence of high-frequency trading firms. Those firms now account for roughly two-thirds of daily trading in equities in the United States.
That spells trouble for most institutional fund managers-namely mutual funds and pension plans-given the increased risk of information leakages and competition for liquidity on any given venue.
Pre-trade analysis is used by the trader to calculate expected costs and which tactics to use including the algorithm and setting the pace and timing of execution.
Real-time or at-trade TCA also allows the trader to monitor the selected strategy for a particular stock on any given day; that means how many shares to buy at what pace at what time of day.
Post-trade TCA is typically used by the portfolio manager to evaluate how well the trading desk or broker-dealer performed against a theoretical benchmark known as implementation shortfall arrival price.
The best scenario in using TCA tools, said Matthew Celebuski, head of the quantitative research and analytics team at SJ Levinson, is for traders to run the pre-trade transaction cost analysis on the unexecuted portion of a trade, the post-trade on the previously executed order and the real-time or at-trade analysis on working orders. Doing so allows traders to monitor the results of their strategy and act quickly to change it, if necessary.
Regardless of the type of transaction cost analysis used, fund managers should ask their providers some basic questions, recommends James Noser, president of Ancerno in New York. Among these: Does the service provider evaluate results from portfolio managers with different investment styles? Are reports customized? Is peer-analysis offered? And, what additional consulting services are provided?
Even skilled managers may want help in understanding the numbers.
TCA providers are eager to offer user-friendly reports. ITG's Fatica says his firm's Interactive Dashboards provide high-resolution graphics to reduce the time needed to understand reports. Delivered to the fund manager via e-mail, the reports also add context to the analysis such as when the order was sent by the portfolio manager to the trader and executed by the trader or the broker-dealer.
Some firms will go as far as to develop proprietary transaction cost models and compare their results to those provided by third-party transaction cost analysis specialists using a range of factors such as the percentage of volume provided or taken from the market, average spread, sector tracking bias, reaction to volatility and the number of times a print occurs in a day.
"Transaction cost analysis specialists have inherent strengths and weaknesses, as well as natural biases that need to be recognized and understood," said Jason Lenzo, head of equity and fixed-income trading for Russell Investments, a broker-dealer and investment adviser with $140 billion in assets. "We have found that no single provider satisfies all of our requirements."
But doing the work in-house doesn't come cheap. While Lenzo declined to specify the price tag, he did say the work is data intensive. Russell's multiple TCA models are fed by a proprietary real-time tick database, spread data and risk tables.
As real-time analytics gains popularity, one particular tack known as predictive analytics may be its wave of the future, according to Celent, a New York-based research firm. Pipeline Trading Systems' Algorithmic Switching Engine can predict the performance of algorithms under real-time conditions and electronically switch the order from one algorithm to another.
"Traditional post-trade TCA doesn't give buy-side traders at traditional fund management firms enough information to beat out high-frequency traders," said Fred Federspiel, chief executive officer of Pipeline.
"Because high-frequency trading strategies attempt to extract profits on a huge number of small trades, institutional traders who want to take more control over executing their order flow, need to dynamically anticipate and capture liquidity opportunities across all markets," Federspiel says.
SJ Levinson is also opting for a real-time approach, but for post-trade analytics. Rather than sending the fund manager a report the day after the trade has been executed, it can do so almost immediately after the trade is executed on its desktop. The traders and portfolio managers can, within minutes, compare execution results against predefined benchmarks, taking into account the past performance of trading for a particular manager or for a particular security or industry sector.
Practical Trading Tool
The better the coordination between the trader and the portfolio manager when incorporating TCA results in the investment process the less potential for accidental inefficiencies. Joseph Gawronski, president of Rosenblatt Securities, cites the following examples: If TCA is used by the trader to conclude that the optimum time horizon to buy a stock is within the next two days, he might be right from a simple market impact perspective but could harm the investment process.
Conversely, he says, a portfolio manager may routinely insist that a trader complete orders quickly when there is really no short-term above-market return to be had and unnecessarily push the price of the stock , only to see the stock revert to its prior price over the course of hours or days.