Eugene A. Noser, Jr.
February 23, 2004
Mr. Jonathan G. Katz
Re: Concept Release: Request for Comments on Measures to Improve Disclosure of Mutual Fund Transaction Costs; File No. S7-29-03
Dear Mr. Katz:
Abel/Noser Corp., a member of the New York Stock Exchange and other leading exchanges, has long been respected as a leader in the campaign to lower the costs associated with trading. To that end, we provide a range of effective tools and services for institutional sponsors and investment managers. Abel/Noser provides Transaction Cost Analysis in conjunction with other services that help our clients achieve savings.
Abel/Noser Corp. provides Trade Cost Analysis services to over 400 institutional investors, including many of the largest mutual fund companies in the world. We evaluate over $4 trillion of institutional trading data annually. Our universe of institutional trading statistics includes information from over 500 separate investment advisors representing over 750 portfolio manager managers, including some of the largest global money managers. Using a combination of strike price and volume weighted average price measures we have developed investment style trading cost benchmarks that allow institutional plan sponsors and mutual fund directors to make relevant cost comparisons between investment advisors. We welcome the opportunity to comment on proposals to publish trading cost information in mutual fund disclosure documents.
Every officer of a defined benefit plan or mutual fund has an obligation to ensure that all actions taken by the plan are in the best interest of the beneficiaries. Our plan sponsor clients first concern is to ensure that their investment managers make hard-nosed decisions regarding the disposal of plan beneficiaries' commissions. To that end, they monitor explicit commission expenses by market and compare them to benchmarks derived from our universe. This enables plan sponsors to ask their managers meaningful questions when results deviate significantly from expectations.
Mutual fund investors should be able to look at similar statistics for the funds where they invest. Each quarter, a fund could publish total commission expenses by exchange in dollars, cents per share and basis points. The funds could also state these costs as a percentage of assets under management on an average basis. This latter statistic would sum up turnover as an input to commission costs.
The section 28(e) safe-harbor enables investment advisors to acquire independent investment analysis through the use of "soft-dollars." Soft-dollar practices are not in and of themselves wrong, but they are subject to abuse by corrupt practitioners. Many of our clients demand that their investment advisors provide them with information regarding the uses of commission payments that exceed "execution-only" rates, that is, the rate that would pay for trade execution alone. Execution-only rates are easy to determine, they are rates that ECNs like Island, Archipelago, and Instinet offer for access to electronic markets.
The commission should also require that Mutual funds disclose the commission payments in dollars, cents per share and basis points for different types of commission payments. These types of payments should include, proprietary research, independent research, sales of fund shares by the brokerage, and services such as market data or statistical services. Furthermore, fund advisors should disclose the extent to which they fund investment operations out of advisory fees.
Transaction Costs Matter, Too
The conversation becomes more recondite when our clients turn their attention to implicit transaction costs. We usually provide pension sponsors with results of three measures: the market opening price, the daily volume weighted average price and a volume weighted average price for a five-day time period centered on trade-day. Our mutual fund clients also use combinations of strike prices and volume weighted average prices to evaluate trades that extend over more than one day.
Strike prices are prices that represent the prevailing market price at a specific instant. The strike price reveals the direction of the prevailing market trend during the time the manager executes a trade. Strike price measurement results are influenced largely by the mechanical workings of the market. Strike price measures reveal mechanical characteristics of the measurement method.
The open price serves as a reasonably useful proxy for the price that might have prevailed when the investment advisor decided to enter an order to trade. The difference between the open and the actual execution price represents a gain or loss enjoyed by the investor. The result versus the open price reveals information about the market conditions under which investment advisors complete their trades. Investors who prefer to buy rising stocks and sell falling stocks will incur higher costs than investment advisors who do the reverse. Neither strategy can be proven to offer an advantage to investors.
Published spreads fluctuate constantly. Published spreads offer only a hint of available liquidity at given prices. Market participants seldom publish their trade intentions to trade to the world. Often, traders can submit bids or offers, provide liquidity, and capture the theoretical spread as a trading profit. Funds might publish an average spread for the stocks they trade, but we do not know of a generally accepted source of such information.
Volume Weighted Average PriceThe daily volume weighted average price measure reveals a manager's ability to achieve reasonable executions compared to other simultaneous investors. We believe this measure offers a useful indication of a manager's ability to achieve "best execution." The difference between the trader's execution price and the volume weighted average price removes the market movement from the trading cost calculation. Many practitioners accept that volume weighted average price measures provide insight into execution quality. We suggest that volume weighted average price measurement results can be used as a reasonable estimate of spread costs. Documents could express costs in dollars, cents per share, basis points, and as a percentage of assets under management.
Unfortunately, the volume weighted average price measure is unpopular with many investment advisors because of a perception that their traders will fail to execute unfavorable trades in an attempt to "game" the measure. This argument assumes that traders will sabotage an investment idea in the pursuit of a better trade measurement result. Our experience tells us that no trader could play these games for very long. We believe that one can level the same "gaming" criticism at any measurement scheme that relies on reference prices that traders can deduce at the time of the trade. But these reference prices provide the most meaningful information about an investor's ability to implement their strategies and compete with other investors.
Opportunity cost is defined as "the road not taken" in investment. An order that the manager does not fulfill to his satisfaction contains an element of opportunity cost: the gain that the investor had to forgo by failing to achieve an ideal investment. One could calculate opportunity cost as the subsequent performance of shares not bought. Several problems arise in capturing the data Portfolio managers may not reveal orders in a way that the firm can record. Portfolio managers with competing orders in the same fund family may enter larger orders than they require with some expectation that they won't get all the shares they request. Little mention is made of losses avoided, or the performance of dollars put to work elsewhere, but these would have to be taken into account in an accurate opportunity cost computation.
The trade effect measure discussed in the concept release relies on strike prices after the trade to evaluate trading effectiveness. Like pre-trade strike prices, measurement results will overwhelmingly depend on market movement beyond the scope of the trader's skills. Little information is offered by looking at trades one or two days hence. An alternative trade effect measure would compare the active portfolio performance to the performance of the original portfolio held at the beginning of the measurement period. This would illustrate the effect of decisions to buy and sell securities versus a simple "buy and hold" strategy. We believe this simple analysis could be accomplished with fewer burdens on the investment manager and is easier for the investor to understand.
No single measure can reveal all the issues that surround the notion of "trading costs." However, we believe that the measure mentioned above, which reveals performance versus the investment advisors' original static portfolio, captures most of the value of decisions to buy and sell securities over the investment period.
Managers will find it very difficult to estimate trading costs. They will also face the inevitable conflict that will result from wanting to appear to minimize those costs in disclosure documents. Investment managers would have to adopt a uniform set of data acquisition techniques in order to provide the level of specific trade cost data that this concept release contemplates. Stringent data acquisition requirements could be onerous to small investment advisors, putting them at a disadvantage to large investment advisors. Also, uniform transaction cost reporting requires a universally accepted methodology for measuring costs. So far, this methodology does not exist in practice.
We thank the commission for the opportunity to comment on the proposed rules and hope you find our comments useful. If you would like to learn more about our experience providing transaction cost analysis to institutions, please contact me at 212-344-2610.
Eugene A. Noser, Jr.