Transaction Cost Research - Kendall Kim
First Published Tuesday, 10th June 2008 07:59 am from Automated Trader
An excerpt from Kendall Kim's forthcoming book "Electronic and Algorithmic Trading Technology: The Complete Guide" Chapter 10: Transaction Cost Research
Copies of this book can be pre-ordered from the Automated Trader Bookshop by clicking here
New technologies, such as utilizing algorithms and straight-through processing, result from the drive to lower transaction costs, as well as the associated research involved behind each execution. According to the TABB Group, Transaction Cost Research (TCR) is defined as the amount of money spent to open a new position or to close an existing position. Transaction cost analysis started with fulfilling regulatory requirements. It can significantly drag performance, especially for portfolio strategies that include high turnover. All transactions have explicit and implicit costs. Explicit costs are disclosed prior to the trade and include commissions, markups, and other fees.
Implicit costs represent the costs that are not determined until after the execution of a trade or set of trades is completed. TCR can be defined as the movement of the stock price from the time of the investment decision to the expiration or completion of the order. Minimizing implicit cost is a key factor in gauging execution quality. Commissions are generated through trade execution; however, commissions fund multiple services, which include execution, research, conferences, and technology. Transaction costs affect investors, pension plans, money managers, and broker-dealers. These costs are ultimately passed on to the investor. TCR includes the measurement of transaction costs after the trade is executed (post-trade) as well as expected costs before the order is placed (pre-trade). 1 As investment management becomes increasingly competitive, portfolio managers will look for methods of enhancing their returns through lower transaction costs to boost their overall rate of return. On the contrary, broker-dealers and the sell side will try to adapt and continue to service the investment community through lowering commissions and transaction costs by routing executions via electronic venues such as direct market access (DMA) or algorithms (see Exhibit 10.1).
Brokers are under enormous pressure to reduce brokerage commissions. This has caused profit margins to fall, and research costs become increasingly paid for by the broker. The push by the buy side to segment commissions and transaction costs between research and trading led to Fidelity's landmark deal with Lehman Brothers. Fidelity agreed to pay approximately $7 million USD annually for research and approximately 0.02-0.025 cents per share for execution services. Large buy-side investment managers such as Fidelity already have their own research staff, and would most likely put further pressure to segment and lower sell-side research and trading costs with other broker-dealers. Some money managers pay for research out of their own pocket (hard dollars) and receive lower commission costs, translating to higher management fees. Other investment managers may combine research and execution commissions paying higher rates. Firms with limited research needs may use DMA and algorithms. Electronic and algorithmic trading has been one avenue that has helped sell-side firms to retain order flow and lower transaction costs, but they lack a solid method of retaining relationships with the investment community.
Regulatory reporting for executions required by the exchanges and the NASD has led to a wealth of trade data and new software data providers willing to provide this information. Due to these reasons, post-trade analysis was developed prior to pre-trade analysis. The data used to research posttrade analysis include commissions, market data, and the attributes of the order. After the data is collected, the analysis attempts to piece together the transaction costs and determine their origin. The more detailed the information, the more precise the analysis can be. A high-level overview may show how the trade's execution compares to a particular benchmark, or ideal price. A more detailed analysis goes beyond calculating transaction costs and attempts to show when the costs were incurred or why it happened.
Post-trade analytics face many sets of challenges (see Exhibit 10.2). One of the biggest problems the buy side faces is methodology and flawed measurements. Critical data such as historical volatility, liquidity constraints, and adjustments for various market capitalizations are not being included in transaction cost calculations. Post-trade analytics can also be inaccurate, which impacts adoption rates. Another issue is selecting the appropriate benchmark. It is difficult to select a benchmark that can measure different kinds of transaction costs equally. Some benchmarks may not be appropriate for evaluating trades in stocks with widely different levels of liquidity and volatility. The primary benchmark that is used is the Volume-Weighted Average Price (VWAP). VWAP dominates as one of the most useful benchmarks because the concept is already well known and easy to understand. If, for example, an execution price is better than the average price (weighted by volume), then it is a good execution. Another popular benchmark is the arrival price for measuring absolute transaction costs.
There are currently a number of firms that provide post-trade TCR (see Exhibit 10.3). These providers include Plexus, Elkins-McSherry, and Abel-Noser, and other newcomers such as Quantative Services Group (QSG) and GTAnalytics. Firms such as QSG are focused on high-touch analyst outreach while NYFIX is focused on delivering real-time TCR. ITG has become a formidable challenger to other TCR providers through its purchase of Plexus.
Pre-trade analytics offers historical and predictive data on price behavior or how a trade position might react to different trading strategies. It can help a buy-side trader justify an execution or help assess performance. The information can provide data on a single stock order or program trade details such as volume, volatility, illiquidity, and other risk characteristics (see Exhibit 10.4). For single stocks, a trader may analyze a number of different parameters such as the share quantity or duration of the order. Historical data or predictive modeling may derive estimates of the impact of the order, or price movements. When the buy side executes illiquid stocks, traders may have to analyze the risk of market impact and opportunity cost. Electronic trades can be analyzed through risk characteristics, which include the overall risk of the electronic trades the trader will pay to the various brokers/dealers competing for the best bid. Pre-trade analytics are a critical component of the bidding process on program trades, as traders often cite the refusal of brokers to bid on a program without supporting cost validation. Pre-trade data can enable the trader to identify whether the estimated costs for a stock are attributable to volatility, spread, or lack of volume.
Pre-Trade Transaction Cost Research Providers
Pre-trade analytics is an important aspect for vendors when providing trading tools for the investment community. Brokers and vendors have a variety of prepackaged products that may educate traders on a particular stock and help avert high transaction costs before they happen. Pre-trade analytics is an important part of dialogue between traders and portfolio managers. One of the leading providers of TCR is ITG (see Exhibit 10.5). ITG was quick to integrate its pre-trade analytics and distribute the product efficiently to the investment community with first mover advantage. The inherent weaknesses of pre-trade analytics are accessibility and accuracy, along with the questionable willingness of investment firms to utilize it.
The Future of Transaction Cost Research
According to the TABB Group, TCR research products will continue to
evolve based on four basic trends:
1. Integration into the trading platform
2. Customization of benchmarks
3. Flexible data formats
4. Pre-trade cost analytics
These basic trends will address the majority of issues investment firms face with transaction cost research. These include easy access, orders that must be judged based on multiple parameters, and the need to make better decisions at the time of order entry. Information being passed between trading platforms and TCR providers is a critical step between investment managers and broker-dealers. Execution platforms such as Portware and Flextrade have successfully integrated TCR into their trading engines. Direct market access platforms have the ability to calculate short-term transaction cost analysis in real-time utilizing trading information.
Pre-trade TCR may see radical improvements in the next few years. Most brokers currently offer two standard equity analytics: one for single stocks and another for baskets. Single-stock analytics may provide various attributes such as industry/sector, potential hedges, intraday volatility, and volume slices, which are simply average volume of stock over specified intraday intervals. Basket analytics can judge the overall risk in a basket, its exposure to different industries, and the potential implicit costs of the basket.
The interest in transaction cost research is widely attributable to increasing competition for lower transaction costs, and regulatory pressure. Investment managers are pushed to measure and manage transaction costs to increase investment returns, retain clients, attract new prospects, and satisfy regulators. When investment managers began to be judged by transaction costs, this began the push for algorithms and other advanced electronic execution tools. One universally known method of rating quality of execution is through achieving or exceeding the Volume-Weighted Average Price (VWAP). Broker-dealers have responded to the growing pressure from regulators and investment firms' desire for lowering transaction costs. Investment managers increasingly want to reduce implicit costs, and broker-dealers must fulfill this demand in order to retain client business. In the end, transaction cost research will be absorbed into the trading process, and soon incorporated into stock charts, annual reports, and employee compensation plans.
|1 Adam Sussman, From Best Ex to Coaching, TABB Group Report, June 2005, http:// www.tabbgroup.com/our_reports.php? tabbaction¼ 4& reportId¼ 105.|