The Changing Role Of The Buyside Trader

Building

Belter

9rorrlt Perf

orrneinae

AUGUST 2OO4

COMMENTARY 81

THE CHANGING ROLE OF THE BUYSIDE TRADER The role of the buyside trader has changed dramatically over the last decade. Buyside traders now have an arsenal of execution tools at their fingertips and are taking greater control of their orders. Delegating the responsibilities for order execution to a broker is now supplanted by electronic trading.

-r, '.' ' ',s MaNe a Good-lrader? was tne subject of the January 2003 P exus Commentaryl which focused o^ the use of Orde' Varage-ent Systems and Straight Througl^ Processir^g (STP) trrougr FIX connectivity. We questroned whether the fast connections of

Hs irades shift away from a central market io scattered venues, soliciting liquidity became difficult. Without intermediation, both buyer and seller must show up

electronic trading were getting traders where they needed to go.

Just a few quarters later, a proliferation of providers is exploiting network connectivity to provide better, faster and less expensive service in areas such as liquidity aggregation and algorithmic trading. In this Commentary we review recent market structural changes and discuss how these changes motivate and empower the buyside traders to better control their orders.

How Much Have Things Changed? In'Consolidation, Fragmentation, Segmentation

and

Rerr-;laf;^n2 " | Haliis discusseri th= re:scns wh;, "rry markets fragment, the forces leading to re-consolidation, and, a srrhqenrent sit;ation that Harris def nes as a

senmnntcd merkct Clrrite nrescientlv What

independently.

While traders do not want sellside

intermediation on every trade, thev need access to deep liquidity pools. Thus isolated "buyside only" pools

needed to reconsolidate in order to raise the rate of successful matching. The changes mandated by the '1997 SEC Equity OrderHandling Rules combined with technological improvements such as OMS systems and FIX protocols to accelerate the development of electronic communication networks (ECNs). Very quickly, the ECNs saw thai connectivity could improve liquidity, and began

inier-connecting with each other and wrth NASDAO. Growing popularity and volume begat more popularity ano vorume.

haS

happened is very close to what Harris describes. Markets fragment because traders have different needs and trading problems. Pre-1990 market structures were

floor-centric, with broker/dealers intermediating most trades. A "fourth market" emerged in the 1970's as a private phone-based crossing network among buyside. traders to fulfill the budding demand for direct buyside to buyside crossing. In the 1980's the fourth market went electronic via Instinet, followed by both the Crossing Network and ITG's POSIT. Even so, the demand continued for more alternatives to invitino broker intermediation into the orocess.

The 1997 Order-Handling Rules also

mandated in decimalization, which impacted markets three ways: . As the increase in the number of price points thinned

out liquidity at any given price, it became difficult for buyside traders to find size. Traders were forced to break orders into small pieces to access thinned-out

.

liquidity.

"Pennying," jumping in front of bids/offers by penny increments to force better prices or faster execution, became a significant deterrent to buyside traders for

I Available on www.plexusgroup.com '!Consolidation, Fragmentation, Segmentation and Regulation Harrs, Lawrence, Financial l\y'arkets, Institutions & Instfuments v. 2, no. 5, December 1993, 1-28.

posting limit orders to attract hidden liquidity. The book became even thinner. . The reduction of spreads make it difficult for brokers to profit from their traditional market-making business.

.

traded has fallen 1B%. Interestingly, the number of manager clients requesting evaluation of individual traders is up 67% Individual trader skills are increasingly under scrutin,.

This forced them away from the market-maker wholesaling business model toward an agency-based trading service model. This shift positions brokers/dealers as facilitators for both their clients and the facilities offered bv ECNs.

Meanwhile, low returns expectations and increasing surveillance caused buyside firms and fiduciaries to focus on the costs of portfolio turnover. The increased sensitivity motivated them to monitor transaction costs and adjust behavior where appropriate. While it appears that we still have fragmented markets today, the reality is quite different. Rather than being consolidated at the destination, institutional markets are now de-fragmented at the source, the buyside trade desks. New trading systems allow buysiders to efficiently scan all market fragments for liquidity. Larry Harris' "segmented market" seems close to today's reality.

The Buyside Traders Workload In addition to allowing buyside traders to take control of their trading, the technological advancements of the last decade facilitated large increases in productivity on the trade desks.

OMS Sysfems And Connectivity Order Management Systems (OMS) are the keystones of the buyside's ability to manage workflow. These systems collect orders from portfolio managers, aggregate them into trading blocks, manage executions, collect fills and perform allocations. Nearly all of Plexus's manager

clients use an OMS system, although there are big differences in system capabilities. OMS systems are becoming the technological backbone of the buyside firm 3, Automating the 1!z,.rinal oart of the job is not enough; firms demand integration of portfolio management functions such as optimization models and risk and compliance modules.

The key advance for OMS systems in the last five years has been enhanced connectivity to brokers, trading venues, direct market access brokers and algorithmic trading systems. The Tabb Group found that the buyside now handles 52% of its order flow electronically and only 49o/o manually by phoning a broker. The larger the manaqer. the lower the level of manual order handlino.

Aggregation Tech nology : Creati ng A Segmented Market Place

Back-of-the-envelope calculations, using lists of clients

for whom Plexus evaluates individual traders. show

Reconsolidation, the natural evolution of fragmentation,

interesting trends:

has been accomplished via aggregation technology.

2. ,..,; ;

,:,;;;

t

markets. Aggregation technology

gathers data f rom the various electronic marketplaces, thus

!.

. The number of orders per trader processed by Plexus's manager clients is up 20% since 2000. This is probably the best indicator of the increase in trader workload. . Despite increased workload, half of the desks have the same number of traders today as they did in 2000. . The number of managers reducing trader population on their desks exceeds the number of managers. increasing desk population by a factor of 2:1.Individual desk population data indicates that the trader population has decreased B% since the turn of the millennium.

'

Both buyside and sellside now see one consolidated montage of the

While workload is up in terms of number of orders, the shares per order has dropped 24o/o and the total dollars

integrating otherwise isolated pools of liquidity. These systems also provide execution facilities that can trade seamlessly across all trading venues.

ECNs were early providers of aggregation across NASDAQ dealers. Clients would use Bloomberg or Archipelago to access multiple liquidity sources. Even if the ECN did not have the other side, it still got a piece of the action as the point of market entry. The result was

dramatically improved liquidity and quick user acceptance. But the trading was labor intensive; the electronic markets still required manual intefaces. Users 3

See Commentary #74, Januaty 2003. Also see the repod by the Tabb Group. Both afe available on www.plexusgroup.com.

spent a lot of time punching keyboards to work the systems.

Smart order routing technology represents the latest advances in aggregation systems. The technology analyzes order flow for particular patterns of size and liquidity, ident!ir:s attractive venues and automatically routes orders for execution. By re-combining liquidity pools and allowing the user to set custom trading rules, aggregation providers hope to pre-empt the need to link to four or five competing vendors. Advanced aggregation systems can simulate sophisticated order types across market venues even though some ECNs may not support them.

Sn'an o'der rout'rg is still in its infancy: according to the leading providers of this technology, few trade desks have made use of anything beyond simple VWAP trading.

buyside trading desks are fairly new, but penetration has been broad. Both the Tabb Group and the Plexus survey

found that 60% of the desks use algorithmic trading systems, with the percentage soaring for large firms to 80o/o.

The basic strategy of these systems is still crude: "slice and dice" the orders down to the size being shown in the order books. The widespread use of automated trading stratagems such as probe trades and other devices to test the tenor of the market still lies in the future. Modeling trading strategies, even a basic VWAP strategy, is not an easy task. The best algorithmic trading systems have developed much broader capabilities than a basic VWAP model, but the vast majority of users still use only the basic model. The main use of these systems is for the fairly liquid, easy to trade orders where little value can be added by more human attention.

Broker-provlcied aggregation systems are the most rnridelv rrco.i h r hrrrici.lo firme fnllniriod hri inrlononrioni

oroviders such as Lava and FlexTrade, FlexTrade is also

As was the case for aggregator systems, broker provided algorithmic systems dominate the market place. CSFB, Goldman Sachs and ITG are the current leaders. CSFB has a big presence among the large and medium size

offered through some broKers, Smaller aggregators are also aggressive competitors, and together command a larger market share than FlexTrade.

firms and ITG mainly among medium size firms.

Drilling down on firm size shows a different picture. Large

systems with the largest percentage of users.

Proprietary models and FlexTrade were the non-broker

firms main y rely on the Lava aggregation system, followed by FlexTrade. Small buyside firms, on the other hand, rely heavily on brokers to provide this technology. According to the Tabb Group, breadth of functionality ^hAru uurL ^^^+ ^^^^,^+^^ c. the aggregation sysiem from each - rEpdro,vr other. Although sub-second speed - the time it takes to get your order to the pornt of liquidrty - was considered critrcal, most buyside desks felt that most systems were comoetitive on soeed.

The goal of traders using these systems is to either reduce costs or improve efficiency. A few have started to analyze the results from their algorithmic system and are often not pleased with the results. Plexus has evaluated the results of naive VWAP engines versus PAEG/L and also found poorer than expected results.

Conclusions

A growing number of brr,l:ers sell their expertise in usino

The relationship between buyside trade desks and

ECNs and aggregation systems to provide buyside

brokerage firms is in the process of redefinition. Indeed, today's markets operate radically different than they did three or four years ago.

traders agency executions without invoking "shopping" of the order Brrvside desks use these firms to become active users of aggregation and access technology

without investing time and money. Firms like Pulse Trading, White Cap Trading, Firefly, UNX and others have

experienced good volume growth in the Plexus Broker Universe over the last few years. Other firms in this space are EGS, Electronic Specialist, EquityStation, Terra Nova lnstitutional. Vie Financial and Paravane Partners.

Algorithmic Trading Algorithmic trading systems model the trading rules and logic of different trading strategies and implement the models via FIX connectivitv. The svstems offered to the

The idea of "relationships" being the gravitational force holding together buyside trader and broker is weakening. Changes in market structure and order-handling rules allow the buyside desks to take more control of their orders. The days when brokers controlled market access and market information appear gone forever. This will not be good news for brokers who rely on little more than relationships for the securing of order flow. With more trading being done electronically or through direct floor access, the remainder of the pie is getting thinner. We predict that many brokerage firms will seek

to merge or be

acquired

to beef up their

value

Algorithmic trading systems model the trading rules and

logic

proposition.

For a while, we worried that the '1997 Equity OrderHandling Rules would fragment the market and make trading more difficult. No longer; recent developments in trading system technology now provides buyside traders with the ability to efficiently scan all fragments of the market place for liquidity.

Aggregation systems

are

re-consolidating the

fragmented market for the buyside desk. These systems

orovide one consolidated montage across

of

trader tactics and apply

communication

technology to implement the model. All current models use VWAP as a basis for their execution strategy, with the best systems offering broader models. Clients are still struggling with how best to use the systems.

In the future, we can expect advances in modeling technology to result in more sophisticated automated order handling. Increasingly, the models will add value directly by reducing trading costs or indirectly through imorovements in efficiencv. The role of the trader will

'-

all market

i...|:;i:: ,,,i.:,.;

;,,:,,,r1,

fragments, and now incorporate

smart order routing technology ,::,:'. 1.. ,:;., to automatically route orders to soecific venues for execution.

Algorithmic trading systems provide the buyside desk with more advanced methods of handlino small orders.

evolve more towards trading strategy as a component of investment strategy, and away from the clerical roles of phone handler, commission cashier, and lion tamer.

Plexus News

Put this on your calendar! The Plexus Group 9th Client Conference is scheduled for September Fairmont Turnberry ls/e Resorf & CIub (http://www.fairmont.com/turnberryisle/)

Reprint any portion with credit given to:

g forrp 11150 W. Olympic Blvd., #900 Los Angeles, CA 90064

PH:

31 0.31

2.5505 FAX:

31 0.31

2.5506 www.plexusgroup.com

Plexus Group is a wholly owned subsidiary of JPMorgan lnvestor Services Company, a division of JPMorgan Chase. @

2004 Plexus Group, lnc.