pension fund perspective: commissions and trading costs

CFA Institute

Pension Fund Perspective: Commissions and Trading CostsAuthor(s): William W. Priest, Jr.Source: Financial Analysts Journal, Vol. 41, No. 1 (Jan. - Feb., 1985), pp. 11-12+16+37Published by: CFA InstituteStable URL: http://www.jstor.org/stable/4478801 .

Accessed: 16/06/2014 09:01

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

.

CFA Institute is collaborating with JSTOR to digitize, preserve and extend access to Financial AnalystsJournal.

http://www.jstor.org

This content downloaded from 62.122.72.104 on Mon, 16 Jun 2014 09:01:57 AMAll use subject to JSTOR Terms and Conditions

http://www.jstor.org/action/showPublisher?publisherCode=cfa

http://www.jstor.org/stable/4478801?origin=JSTOR-pdf

http://www.jstor.org/page/info/about/policies/terms.jsp


Sn i n Fun P r tivD Patrick J. Regan Pension Fund Perspective V.P., BEA Associates, Inc.

The control and use of brokerage commissions by investment managers promises to be one of the major issues of 1985. As nmore third-party research and performance measurement services are offered on a "soft dollar" basis, the ability to measure and monitor trading costs becomes crucial. In 1978, Gilbert Beebower, Myron Scholes and William Priest developed a system for measuring trading costs. That system, offered by SEI, currently monitors the activ- ities of 15 major institutions that traded some $85 billion of equities in 1983.

In the following column, Bill Priest addresses the "soft dollar" controversy and offers some specific advice on the measurement of transaction costs. Bill is a Vice President and Director of BEA Associates, Inc. P.J.R.

Commissions and Trading Costs William W. Priest, Jr.

Who should control commissions? Do they constitute the only costs of trading? This column attempts to answer both questions.

Who Controls Commissions? The nature of the relationship between a professional investment ad- visor and a client should be delineated in the contract that exists between them. The contract may stipulate, for instance, that the investment manager will select brokers and the client will place the transaction orders. The ad- visor's service could simply be that of providing advice.

In practice, however, most clients grant their investment managers, not only the discretionary authority to decide the specific security, the time to transact and the price to be paid, but also the authority to select a broker and to place the transaction orders that execute these judgments. Where the client's contract prescribes this cus- tomary, "full service" investment management service, the manager

agrees to provide the full package of services for the sole benefit of the client.

When the client, who in the end pays the commission, grants to the manager the power to select the broker, place the order and negotiate the commission, that power is granted "in trust," which means that it is granted to the manager for the sole purpose of its being used for the exclusive benefit of the grantor, the client.

The law recognizes that the money management business is not the gar- den variety commercial transaction where caveat emptor prevails. On the contrary, the relationship between money manager and client is more like that between physician and patient, or lawyer and client. It is conducted on a professional and fiduciary basis, for that is the character of the business. The manager offers the client a learned skill, and promises to apply it for the benefit of the client.

Given the nature of the relationship between the money manager and the client, two absolutes govern the manager's behavior-loyalty to the client's best interest, and professional com- petence and care. If the manager benefits along with or (worse) in place of the client, the firm is legally in a con- flict of interest position. The manager's compensation comes from the client and consists of the fees that, ac- cording to the client's written contract, the manager agrees to receive as pay- ment for the described services. (The law requires that there be a written contract, and that it contain a descrip- tion of the manager's duties and compensation.) In other words, the manager will provide the services described in the contract for the fees stated, not for the fees plus ancillary benefits from third parties with whom the manager deals. I

This duty of care is related to the duty of professionalism. When an order is placed with a broker, the manager knows the account will be di- rectly affected by the placement of that

order. Because he is exercising power delegated to him in trust, he must ask, "On this trade, at this time, is this the best we can do for our client?"

Who's the Client? In some instances, there may be a

question as to the identity of the client. When a business corporation creates a plan for the benefit of employees, commits assets to the plan, and retains a manager to manage those assets, the manager's client, under fiduciary law (ERISA), is the plan, not the business corporation.

ERISA states that an employee plan cannot exist without a writing, and that this writing must include statements about the purposes, benefits, and administration of the plan, including the people responsible for administering the plan-i.e., the plan's fiduci- aries. When an investment manager signs a contract to supervise the investment of the assets of a plan, the manager becomes a fiduciary to the plan.

A manager asked to direct trades should receive these instructions in writing from the person or committee responsible for administering the plan's assets and should ascertain that the plan beneficiaries will benefit from the service to be performed by the se- lected broker. The manager has an ob- ligation to discuss with the plan committee the probable costs to the plan account of following these directions, if he feels that such costs are likely to hurt the plan. The plan committee or designated person can decide whether the service to be rendered is worth the cost.

The principles discussed so far are outlined in the accompanying exhibit.

Commissions and Executions What does it cost to trade? Is the charge reasonable?

A broker acts as an agent, facilitat- ing the purchase or sale of stock by one party to another. The broker charges, for this service, a commission cost that 1. Footnotes appear at end of article

FINANCIAL ANALYSTS JOURNAL / JANUARY-FEBRUARY 1985 C 11



will vary depending upon the diffi- culty of the trade. An agency trade of modest size, involving a large, liquid stock, is often referred to as a "no brainer." "No brainers" are cheap to execute-they usually entail just clearing costs, which vary from about two to three cents per share at the larg- est brokerage houses.

Not all trades are so simple, however. They may involve large blocks of stock in smaller, less liquid, compa- nies. If liquidity cannot be found at a reasonable cost, the manager is faced with unexecuted orders. In such cases, the broker may step in as the party on the other side of the trade. These principal trades have no explicit commission costs. But even though costs may not appear on a broker confirmation, there are costs.

In short, trading costs constitute more than commissions. There are principal costs as well as commissions.

Measuring Trading Costs Several years ago, BEA Associates

developed a way to measure trading costs. Gil Beebower, Myron Scholes (Professor of Finance at Stanford Uni- versity) and I looked at every BEA trade for one and a half years before deriv- ing a process to evaluate and measure costs.

Two problems confronted us-defining trading costs and measuring them. We started by defining trading costs as commissions plus principal costs. In order to estimate the latter, we sought to measure the market impact of a trade-i.e., the wealth impact a given trade has on the market in that stock. This requires estimating what the stock would have done had it not been traded; this is somewhat like trying to trace footsteps on paths not taken.

The problem may become clearer if we look at a more familiar analogy. When you get into a bath full of water, the water level rises; when you get out, the water level falls. Similarly, when placing an order for a stock in the mar- ketplace, you might expect the bid-ask spread to rise when the order is placed and then decline when the transaction is completed. We wanted to measure the costs of both steps and add them to commissions in order to approxi-

mate trading costs. The problem is basically one of

measuring very short-term performance. And it is complicated by the fact that market impact reflects the combination of execution costs (which are the product of the interactions between the manager, the trader and the broker) and any costs arising from very short-term information effects (for our purposes, opportunity costs). Execu- tion costs are our primary concern, for they diminish real wealth. Opportu- nity costs-arising, for example, from an inability to trade on the basis of information before that information has been discounted by the market-are more hypothetical; they are so called because they represent the cost asso- ciated with missing out on a poten- tially profitable opportunity, rather than reduction in real wealth.2

We developed software to measure both parameters of principal costs- execution costs and opportunity costs.3 We used the sum of the execution costs plus commissions to measure trading costs.

The Cost of Trading BEA has employed the measure-

ment process for more than six years, and 15 major institutions (including Morgan Bank, Mellon, Citicorp, Trav- elers, Equitable, T. Rowe Price and Wellington) now use the methodol- ogy. These institutions collectively traded $85 billion worth of equities in 1983 and $35 billion over the first six months of 1984.

On the basis of several years of data, the study concluded that trading costs (execution costs plus commissions) currently average about 50 basis points round-trip. Commissions, at their cur- rent average of slightly more than 16 cents per share round-trip, account for virtually the entire 50 basis points. On average, market impact costs are neg- ligible.

The averages, however, disguise some important differences between investment organizations. Some organizations incur trading costs amount- ing to 200 basis points. Others actually have negative total trading costs. Fur- thermore, these differences are sys- tematically related to the investment manager's requirements and trader

(conltinued 01l page 16)

I 3I

-~~~ S .

* 6 6 , w 5~~~~~*' '

0 6

56

Eg5

5-1S1_SO

;5 p

m~~~~~I

TH GNOMES-'-

5 5 == *

.Freddie

Mac5

Fedra

*

FINANCIAL ANALYSTS JOURNAL / JANUARY-FEBRUARY 1985 O 12



U.S. POSTAL SERVICE

STATEMENT OF OWNERSHIP, MANAGEMENT AND CIRCULATION (required by 39 U.S.C. 3685)

1. Title of Publication: Financial Analysts Journal. 1A. Publication No. 571200. 2. Date of Filing: October 1, 1984. 3. Frequency of issue: Bimonthly. 3A.Annual Subscription Price: $36.00. 4. Location of Known Office of Publication (Steet, city, county, state, zip code) (Not printers)

1633 Broadway, New York, N.Y. 10019. 5. Location of the Headquarters of General Business Office of the Publishers (Not printers)

1633 Broadway, New York, N.Y. 10019. 6. Names and Addresses of Publisher, Editor, and Managing Editor

Publisher (Name and address) The Financial Analysts Federation, 1633 Broadway, New York, N.Y. 10019.

Editor (Name and address) Charles A. D'Ambrosio, 3604 42 Avenue N.E., Seattle, Washington 98105.

Managing Editor (Name and address) Judith F Kimball, 1633 Broadway, New York, N.Y. 10019.

7. Owner (If owned by a corporation, its name and address must bc stated and also immediately thereunder the names and addresses of stockholders owning or holding 1 percent of more of total amount of stock. If not owned by a corporation, the names and addresses of the individual owners must be given. If owned by a partnership or other unincorporated firm, its name and address, as well as that of each individual must be given.)

Name Address The Financial Analysts Federation, 1633 Broadway, New York, N.Y. 10019 (a non profit corporation devoted to the advancement of investment management and security analysis)

8. Known bondholders, mortgagees, and other security holders owning or holding 1 percent or more of total amount of bonds, mortgages or other securities (If there are none, so state)

None 9. For Completion by Nonprofit Organziations Authorized to Mail at Special Rates

(Section 132.122, Postal Service Manual) The purpose, function, and nonprofit status of this organization and the exempt status for Federal income tax purposes.

(Check one) X Have not changed during Have changed during (If changed, publisher must

preceding 12 months preceding 12 months submit explanation of change with this statement.)

10. Extent and nature of Circulation Average No. Copies Actual Number of Each Issue During Copies of Single Issue

Preceding 12 Months Published Nearest to Filing Date

A. Total No. Copies Printed (Net Press Run) 20,569 20,660 B. Paid Circulation

1. Sales through dealers and carriers, street vendors and counter sales 0 0

2. Mail Subscriptions 19,250 19.343 C. Total Paid circulation (Sum of BI and B2) 19.250 19.343 D. Free Distribution by Mail, Carrier or

Other Means, Samples, complimentary, alid other free copies 202 140

E. Total Distribution (Sum of C and D) 19,452 19,483 F Copies Not Distributed

1. Office Use, Left-Over, Unaccounted, Spoiled After Printing 1,117 1,177

2. Returns From News Agents 0 0 G. Total (Sum of E & F-should equal

net press run shown in A) 20,569 20,660 11. I certify that the statements by me above are correct and complete.

(Signature of editor, publisher, business manager, or owner.) Judith F Kimball, Managing Editor

12. For completion by publishers mailing at the regular rates (Sectoion 132, 121, PostaL Service Manual) 39 U.S.C. 3626 provides in pertinent part: "No person who would have been entitled to mail matter under former section 4359 of this title shall mail such matter at the rates provided under this subsection unless he files annually with the Postal Service a written request for permission to mail matter at such rates." In accordance with the provisions of this statute, I hereby request permission to mail the publication named in Item 1 at the phased postage rates presently authorized by 39 U.S.C. 3626. (Signature and title of editor, publisher, business manager, or owner)

Judith F Kimball, Managing Editor

expertise in broker selection. Limiting trader discretion in cases where trades are not "no brainers" may wind up costing more than the few cents saved

or recovered on commissions. The commission, like the tip of the

iceberg, is merely the most visible part of trading costs. It now generally

Principles of the Relationship between Investment Manager and Client

The manager operates under professional and fiduciary law, not caveat emptor.

The manager, as a result, has the twin duties of professional "care" and "loyalty."

The manager must act solely in the client's interest, avoiding all conflicts of interest that have not been explicitly agreed to by the client in the account contract.

The nature of the manager-client relationship is detailed in a written agreement between the two parties.

In the case of an ERISA client, the client to which the fiduciary duty is owed is the plan, not the plan sponsor.

The value of directed brokerage commissions, whether instructions come from the manager or the plan sponsor, must accrue to the benefit of the plan; instructions for directed trades should be spelled out in writing between the sponsor and the manager.

amounts to less than the "eighth" (121/2 cents per dollar) that the trader may bargain over. The "eighth" remains unseen, but it is not unfelt. Execution costs will appear in the overall rate of return experienced by the fund. They may be attributed to, say, security selection effects, but they are a cost of trading.

By focusing on commissions, and defining them as the costs of trading, plan sponsors and asset managers may be acting penny wise and pound fool- ish. Their "trading costs" may look great, but their total return numbers will suffer from what they might otherwise have been. The portfolio may avoid the visible part of the iceberg, but still founder on the submerged part.

Conclusions A great deal of theory and evidence

(concluded on paze 37)

FINANCIAL ANALYSTS JOURNAL / JANUARY-FEBRUARY 1985 0 16



11. The model has an R-squared of 0.38, which is significant beyond the 0.001 level. The examina- tion of residual plots and computed serial correla- tion statistics also supports the standard regression assumptions. Earlier large-sample studies report R-squared values (often substantially) be- low 0.20.

12. In more technical language, the predictive accuracy is measured using the root mean squared error (MSE). The predictive accuracy of the model was also examined through the calculation of the mean absolute deviations. Use of the MAD implies a linear loss function, whereas use of the MSE implies a quadratic loss function. The com- parative rankings of the MADs are identical to those presented in this section. These results are

available upon request. 13. The apparent stability of the first several months

is largely due to the scarcity of offers during the 1971-72 period.

14. The technique used here is multiple classification analysis. This procedure is the equivalent of multiple regression with indicator variables but provides a convenient display of average bid premiums for various types of offers. The coeffi- cients from the analysis are essentially the same as for the estimation sample and, consequently, the sums of squares and their corresponding statistics are not reproduced here.

15. Gershon Mandelker "Risk and Return: The Case of Merging Firms," Journal of Financial Economics, March 1974, pp. 303-335.

Pension Fund Continued from page 16

now demonstrate that trading costs are not just commissions. A trade incurs an economic cost, which we define as the sum of the commission and the execution costs. Squeezing, or minimiz- ing, one component may simply cause the other to balloon. Many of the brokerage firms offering rebates of commissions to plan sponsors reflect this phenomenon. The commission is low, but the hidden execution costs may be high.

Finally, under ERISA, the benefici- ary of a security trade is the plan, not the plan sponsor or the plan's asset manager. The only legally sanctioned exception pertains to "research" (Sec. 28E), and this aspect should be cov- ered in the writing that exists between the sponsor and the manager when- ever the former delegates the authority to trade to the latter.

If they are to meet the tasks as- signed them by their client-the plan-both sponsors and asset managers must understand trading costs, measure them, and recognize the ben- eficiary on whose behalf they are in- curred.

Footnotes

1. There is one established exception to this: If the contract so states, the manager may receive "research" from the brokers that execute trans- actions for the account. The client must be told, and must agree in writing, that the manager can pay up for research. Otherwise-the standard is clear-the manager cannot pay up. The manager cannot knowingly trade in the "wrong" place, risking the quality of the execution, just to get the research. This is a highly technical area that de- mands good legal advice.

2. The manager can also force opportunity costs to be favorable-i.e., zero or negative for purchases and zero or positive for sales-by refus- ing to trade unless these results are obtained. It is only after the trade that the determination of fair or un- fair value can be ascertained.

3. A detailed discussion of the meth- odology may be found in Gil Bee- bower and William Priest, "The Tricks of the Trade," The Journal of Portfolio Management, Winter 1980.

UTILITIES COMPANY

COMMON STOCK DIVIDEND

The Board of Directors has de- clared a regular quarterly dividend of 59 cents per share on the common stock of the Company, payable January 2, 1985, to share- holders of record at the close of business December 3, 1984.

PETER B. TINKHAM Secretary

Dallas, Texas November 16, 1984

Principal Subsidiaries:

TEXAS UTILITIES ELECTRIC COMPANY

Divisions Dallas Power & Light Company Texos Electric Service Company Texas Power & Light Company

Texas Utilities Generating Company

TEXAS UTILITIES FUEL COMPANY

TEXAS UTILITIES MINING COMPANY

TEXAS UTILITIES SERVICES INC.

FINANCIAL ANALYSTS JOURNAL / JANUARY-FEBRUARY 1985 O 37



pension fund perspective: commissions and trading costs

Documents