Building a Best Execution Dashboard for Fund Directors
Written by C. Meyrick Payne & Jay Keeshan   

Monitoring best execution (BEX) is a difficult and oft-misunderstood component of mutual fund governance and certainly one where the impact on shareholders is substantial. Measuring best execution can be mind-numbingly complicated because experts often spend much time on the complexities of trading and the intricacies of order flow. While there are plenty of experts to pursue the details, just a handful of directors are responsible for having the skill to oversee trade execution. What fund directors need is a simple dashboard with measurement dials to indicate problems. This bulletin summarizes the basic design of a dashboard which contains the critical measurements for monitoring best execution. A companion bulletin contains a similar dashboard for the control of soft dollars. Not surprisingly the dashboard will differ for each complex depending on its structure and fund configuration. However the principles should generally follow those discussed in the paragraphs below.

Best Execution Dashboard 

1. Quarterly Alpha Loss

Whenever investment positions are traded there are inevitably frictional costs which can be expressed as a loss of alpha (margin in excess of market return). The science, although some would refer to it as art, of transactional cost analysis (TCA) measures this loss of alpha. What is quite remarkable is how large this can be. For example in the fourth quarter of 2004 the average loss of alpha for large cap growth stocks was 68 basis points. When compared to the average expense ratio of these funds (124bp) during the same quarter, the alpha loss from trading represented an additional 55% drag. This is the first dial on the dashboard.

2. Alpha Loss by Investment Style

For every quarter, the fund board should know the average loss of alpha for each investment style within the complex. Unfortunately the science of TCA is not well developed for fixed income products so these will typically be omitted from the dashboard. Alpha loss can simply be arrayed in the Morningstar style box.

3. Good/Bad Brokers

The actionable component of TCA is broker selection. As a result one of the most important dials on the dashboard is the identification of the best and worst performing brokers. In the science of TCA, good or bad is measured in relation to a database of brokers which conduct the same sort of trades in the same time period. The TCA vendor has such a database. The yardsticks differ by TCA vendor but may include PAEG/L (by Plexus), ACE (by ITG) or VWAP for several others. Whether a particular broker is good or bad in any specific quarter is not as important as the trend – a broker which constantly ranks in the lowest quartile is probably not delivering best execution.

4. Good/Bad Trades

Over the years MPI has been building best execution dashboards, we have found that fund trustees benefit greatly from seeing the best and worst three trades in each quarter. It is not because the individual trades are particularly meaningful, but rather that the ensuing discussion greatly helps to understand the decision process which the trader has to go through. Fund directors have a tendency to think of portfolio decisions as beginning and ending with the portfolio manager, when they are in fact a product of a much larger organization including at a minimum investment research, portfolio management and the trading desk.

5. “Execution Only” Broker Percentage

A critical part of the evaluation of best execution is the selection of broker type; particularly “execution only”, “third party” and “proprietary” brokers. In recent years Electronic Crossing Networks (ECNs) have become a major force within the brokerage community and now ECNs can handle a great many of the routine trades. The benefit of using ECNs is almost always a lower commission cost which should also be shown for each of the three classes of broker. The dashboard should contain a percentage breakdown of all the trades made in a quarter by these three categories. Over time fund trustees should look for an ever greater percentage of “execution only” trades with brokers.

6. Good/Bad Broker Classes

Within each category of broker the dashboard should show the comparative rating against the vendor’s benchmark. The board discussion around this measure helps to ensure that fund directors are aware that commission is but one small part of the total cost of trading. Different vendors have slightly different ways to compute the total cost of trading, but all take into account the price change that a trade causes. This is typically called “market impact”. Another source of alpha loss is the “cost of delay”. The trading desk will often divide a trade (as specified by the PM) into several tranches. Each tranche may be executed separately and as a result the delay will either gain or lose alpha.

7. Good/Bad Trade Classes

Fund directors will want to know if there are patterns of alpha loss. For example, do trades of equity positions with market capitalization of less than $1 billion routinely rank poorly when compared to the TCA vendor’s benchmark. Or do trades of more than 10,000 shares rank poorly. Such measures are helpful for locating patterns where corrective action on the trading desk might be appropriate.

8. Portfolio Turnover Percentage

A tried and true measure of best execution is to analyze the comparative portfolio turnover of the funds under the Board’s purview. Portfolio turnover is best measured in comparison to other funds in the complex and to that of competitive funds organized in the Morningstar style box.

****

The concepts discussed in this bulletin will be extrapolated and refined in the series of regional workshops which MPI will conduct with the Mutual Fund Directors Forum during 2009.