Payment Oriented Brokerage

_-,6

ffisffi

*"ffi

#*ffiffiffiw& fiq wp%g%w*# ;x;; ffis*w*-sp w

Wayne H. Wagner, Partner

Edward C. Story, Partner

SERVICES PROVIDED TO Institutional Investors

Larry J. Cuneo, Partner Mark Edwards, Associate

Investment Managers

Michael Keady, Associate

Institutional Traders Plan Sponsors

-

Rosemary Tribulato, Associate

ALPHA CAPTURETM CASE STUDY: EVALUATING PAYMENT.ORIENTED BROKERAGE COMMENTARY #30

January 1991 Plexr

developed ALPIIA CAPTURE_ to help investment managers q4ptule the intrinsic value of Portfolio Manager decisions into client portfolios. This case study is drawn from a recent report to one of our clients. It illustrates how investment managers use ALPILA CAPTURE to fine-tune their investment management process through careful identification of broker skills.

The he::Problem

tiveness as full-service and execution-oriented brokers.

The investment manager wants answers to these questions:

1.

The graph below is taken

manager's trade

desk believes that brokers selected for nontraditional research may not handle trades with the same skill and effec-

mission and impact costs of

actual trading with specific brokers. Impact is measured from the price/time at which the broker receives the order. It represent the true cost incurred while the broker was in charge of the order. Impact costs clearly dominate commission costs. 0

Is the perception correct? Are there detectable differences in broker skills? Do the differences apply to whole classes of brokers, or only to specific brokers?

-o.2

3.'* .E -o.c t -.t

3.

Can broker skills be objectively measured, considering differences in trade difficulty, order instructions, etc.? Can trade cost measurement pinpoint problems for the desk's monitoring of the brokers?

How can the manager establish

damage control so that problem-broker trades do not

handicap client interests? How are trading deficiencies best communicated to the probIem broker so he can improve trading effec-., tiveness -- to the benefit of manager, sponsor, and broker?

rbI n

t

,q

5q -t g

2.

from the Alpha Capture report. It shows the com-

EVideneC

-''t

s-'*

T

0

-t.E -2

1231587!9

lzho.titet$gmcoffikll@t

+otfldl}

|

The diagonal line across the chart shows a measure of difficulty. Order size, stock liquidity, market direction and the degree of urgency all affect Alpha Capture's difficulty measure. Brokers handling easier trades sort to the right. Note that broker #8, a payment-oriented brokerage firm, incurred the greatest impact. Next

#1, a full broker. (There is no systematic effect by broker type.) The table below compares the largest impact comes from broker service

trading done by these two:

6SSWi!*hireS*r:i*varql,$a.rit*31#

il

$ar:taMonie,a"e#if*rnia**4*3

ll

{313}45t-5*75

il

fAX {213}394-6t3e

#

orders Avg shs Vo

of.vol

Difficulty (1=easy) Days to complete

Broker #8

Broker #1

47

7T

20M

45M

2A% 2.5 0.36

36%

C6ncluSionS

3.1 0.51

The Sharpe-Grossman formula for the Thomas Loeb trade cost data imputes theoretic cost. Broker #L had more difficult trades to handle, yet executed them at lower cost than Broker #8:

cost Actual cost Cost ratio Theoretic

-1.2Vo

'1.87o

-Z.I%

-7.4%

I.jX

0.8x

The desk considered Broker #1's execution skills satisfactory for the more difficult/delicate mix of orders. By any absolute or comparative stan-

dard, however, Broker

#8

appears

to

incur

Broker #8, receives significant commission business to pay for valuable services. Yet they traded worse than any other broker, even though the trades handled were much easier than average.

There is an additional secondary effect: brokers who do a good job are concerned about "liquidity skimming"-- the deliberate feeding of easy orders to less qualified brokers. And they are right! Services purchased through Broker #8 are paid for many times over: with the commission, plus through a performance penalty averaging nearly 150 basis points on every trade made!

excessive transaction impact. Changes

A key factors in transaction costs is the speed of order completion. When time to completion is thoughtto be critical to the success of the trade, the manager must be prepared to "payup" for the liquidity demanded.

Trade instructions communicate trading urgency to the market. Market orders and principal trades, which rely on the specialist or broker to provide liquidity if none is available, convey a sense of urgency. Working or limit orders, in contrast, indicate a willingness to wait for lower cost natural liquidity. The trader thereby signals that he or she is more price-impact sensitive than time sensitive. The table below summarizes the trade orders sent to Broker #8. The figures below Broker #8 data are the averages for all brokers used:

Market Work Limit

Orders

Avg

shs

Conclusion: Vo

vol

Conclusion:

The results are not encouraging. Broker #8 is more costly to use for all types of orders except the trivial (less than 1000 shares!) working orders.

20410 20M 1M (40) (15)

30M (s3)

Poor broker performance Ieads to poor investment performance. Thispenalizes

both the investment manag-

er and the plan sponsor.

There is neither business or fiduciary justification for using a less-than-adequately-skilled broker. Either a broker has to deliver execution services comparable to other brokers. or he should not be used.

A

conscientious broker will respond to these client analyses by working to improve procedures, personnel, and correspondent relationships. The goal is not punishment, but positive change to align the quality of the execution services to the trading needs of the manager.

True "bestexecution"cannot be defined as a simplistic cents/share commission. Sewices purchased through commissions must be justified both on the value of the research and the excellence of the execution. A manager who fails to perform either test compromises bothhis fiduciary responsibility and his investment performance.

-w.w.

Easier

25Vo 0.3% 70% (20) (2e) (2e) Easier

cost -1.4Vo -0.4% -0.9% -2.4% -0.1% -1.2Vo Actual cost Cost Ratio 1.8X 0.3X t.4X (0.s) (0.7) (0.3) Evidencs Much more costlv

(c)1991, Plexus Group,

A General Partnership

Theoretic

You are welcome to reprint quotationsor extracts

from this material with credit given to Plexus Group, 606 Wilshire Boulevard, Suite 310, Santa Monica CA 90401. Tel. (213) 451-5075.