What Goes In Determines What Comes Out

plexusgroup B u i l d i n g

B e t t e r

P e r f o r m a n c e

C O M M E N TA RY 7 9

JANUAR Y 2004

WHAT GOES IN DETERMINES WHAT COMES OUT Evaluating Research Inputs

Insightful and accurate Equity Research is the basis of informed investment decisions. Yet despite the central role of research, few managers can point to a disciplined process that quantifies the broker’s inputs and tracks them over time. In this Commentary we discuss Research Evaluation with Kei Kianpoor of Investars. Research evaluation is one of the key components of Plexus’ Total Cost Management™ along with trading evaluation and transaction processing evaluation.

The AIMR Trade Management Guidelines put transaction costs into a cost:benefit context: do the returns from trading justify the cost of trading? This is why Plexus tracks the decision value of the trades as well as the cost. The bottom-line goal of the Plexus relationship is, of course, improvement in the individual firm’s investing process. We have often used the concept of “the investment factory” to illustrate the approach, with ideas being the raw materials and the resulting investment performance being the value of the ideas less the sum of the visible and hidden costs of transacting.

In their ceaseless quest to deepen the relationship to the money manager, the Great Wall Street Selling Machine offers an astonishing array of useful and frequently critical services. As the factory analogy suggests, one of the most critical inputs is investment ideas, a.k.a. research. In today’s post-FD, post-Eliot Spitzer settlement, and post-internet bubble markets, it seems fashionable for managers to pooh-pooh the value of broker research as more hype than content and thus distance themselves from the naïve retail investors. Nonetheless, active managers do not ignore sellside estimates, nor industry or company inputs. These inputs often lead the buyside to its own conclusions. Consequently, the buyside continues to pay up to receive research (and other services from brokers) in soft dollars or a high-dollar relationship with key fullservice brokers. The commission pays for a lot more than “the seat in the front of the bus” from the execution perspective.

You don’t know what to pay for execution unless you know the expected worth of what you’re buying. In this Commentary, we widen our perspective from the transaction view to the inputs to the investment process, the investment ideas. Where do ideas come from? Who is the source of the ideas? For Quants, where do the model inputs come from? Questions like these quickly lead to the realization that many of the inputs to the investment process come from brokerage firms. The relationship of the typical investment management firm with the servicing brokerage firms is quite complex.

How is the broker paid for the services delivered? Through commissions or order flow that can be converted into more commissions and/or trading profits. In trader talk, that amounts to commissions and impact, two components of the cost iceberg. Again, we are looped back to the cost:benefit analysis: what did we pay, and what did we get for it? We can identify very few investment managers who have

attempted to determine the objective value of research. Without a systematic review, commissions get allocated to the favorite brokers. Needless to say, it is often very unclear why certain brokers are favored.

clients.

Evaluating the quality of the research input leads us to the Supply Chain Management concept so familiar to industry. Can you imagine Toyota or General Electric buying parts from suppliers without evaluating the quality and price of those parts? Closer to home, could you imagine a portfolio manager investing in such a careless company?

Strong talk! Clearly underlying the desire to cast a light into these dark shadows is a suspicion that the research might not be worth what it costs. We’ve certainly seen the same phenomena in the US with the SEC’s soft dollar sweeps, Spitzer ’s settlement, the Sarbanes-Oxley legislation and the new regulations on mutual fund disclosures recommended by Congressman Baker. It also seems that Bill Donaldson is going to focus the attention of the SEC in this area in order to help restore the confidence of the investing public.

In order to bring research evaluation to the investment management table, Plexus has partnered with Investars, one of the leading firms specializing in research evaluation. Investars is run by Kei Kianpoor, and Plexus asked Kei to explain his methodology for the benefit of Plexus clients.

So more transparency and accountability is coming to the fund management companies. Managers will be required to be “more answerable to their clients.” Proactive managers will seize this as an opportunity. Demonstrating both effectiveness and care on how they handle clients’ money, both principal funds and

Plexus: Why the big current interest in evaluating research?

“The combined quantitative/qualitative ability to analyze and understand the implications of news and ideas is a fund’s ultimate competitive edge.”

Kei: The interest comes from three sources: regulators watching out for the retail investor, plan sponsor and mutual fund boards watching out for their fiduciary interests, and a renewed questioning of the value of sellside research by investment managers.

The combined quantitative/qualitative ability to analyze and understand the implications of news and ideas is a fund’s ultimate competitive edge. Let’s begin with the positive side. The primary value proposition of Investars is to help an actively managed fund assess the analytical resources that create the ideas that drive active performance. Ideas include such inputs as estimates, judgments and insights, industry knowledge, updated information, and access to management. These are all part of the research picture. Add to that the regulatory drivers. The clients want transparency, and the regulators are responding as you would expect. The Financial Services Authority, the UK’s equivalent of the SEC, put it very bluntly: Up to 40% of total commission spend is used to acquire services in addition to dealing [trading] so it is important that investors are clear on how their money is spent. These proposals [to require express agreement from the customers for the payment for services in soft dollars] are designed to do away with distortions in the market and make fund managers more answerable to their

payments of commissions from the funds is a competitive performance and marketing advantage. Plexus: Surely some managers have a process in place to review research quality – or don’t they? Kei: Many say they do, but our surveys have failed to turn up even one manager that has a disciplined process to quantify and track broker inputs over time. The resulting process tends to emphasize selective recall of the home runs, especially the recent home runs, as well as those brokers who put the most emphasis on client relationships without necessarily having the backing of strong research. We think we have a better approach. Plexus: How does it work? Kei: As usual, it starts with a database of ideas; who said what, when, and what resulting actions were taken. This must be merged with a database of security returns plus the usual categorization information. Then research quality is measured. All this, of course, is available in a web-enabled, graphical environment. Plexus: Can you give me an example? Kei: Contrast the record of predictions on Garmin by

announcement. By the time that the analyst or PM can react, prices are already moving, and trading costs may consume all of the adjustment's value. More timely changes lead to lower cost executions at a desirable price and time.

•GRMN (GARMIN LTD) CSFB ratings 55 45 35 25

Plexus: Surely it’s not that simple! What about the qualitative aspects of broker relations?

15 5

Oct

02

Apr

Jul

Oct

03

Apr

Jul

02

Apr

Jul

Oct

03

Apr

Jul

•GRMN 642 322 2

Oct

CSFB and W.R. Hambrecht. Let me explain the symbols used in the charts: • An up arrow is a raised expectation; a down arrow is a downgrade. • The diamonds with a light triangle in the top represent a positive recommendation reiterated; the diamonds with a dash across the middle are neutral recommendations reiterated.

Kei: Ah yes, that’s critical. The measurement process should include the qualitative evaluations of the inputs received from the broker: Our Qualitative Performance System weighs and analyzes previously unquantified aspects of unstructured, inbound research in four categories: Thought Leadership, Service Quality, Information/Management Access and Overall Brokerage Capabilities. •GRMN (GARMIN LTD) WR Hambrecht ratings 55 45 35

The first observation is that Hambrecht pays attention to this stock. Hambrecht sent out sixteen messages; CSFB only three in the same time span. Hambrecht made a timely buy recommendation before the stock went from 20 to 50, confirming it eight times along the way. Then they downgraded to neutral, reconfirming the rating six times as it declined to 35. In contrast CSFB – and I’m not picking on them, but this is their record in this stock – downgraded the stock at 25 and stuck with it as it went to 50. Clearly any portfolio manager would prefer to have listened to Hambrecht on this stock. Plexus: We don’t want to read too much into one example, as spectacular as it may be. Let’s talk about the bigger picture. How do you measure quality of research? Kei: We have developed three proprietary algorithms, which we call • Actual - Relative, • Consensus - Relative, and • Time to Estimated Accuracy. (See http://www.investars.com/methodology.asp) The last measure is particularly interesting. It asks “When is this research valuable to the investment process?” There is limited value from an analyst who adjusts earnings estimates just before the earnings

25 15 5

Oct

02

Apr

Jul

Oct

03

Apr

Jul

02

Apr

Jul

Oct

03

Apr

Jul

•GRMN 642 322 2

Oct

Our philosophy is that the combined quantitative/ qualitative ability to analyze and understand the implications of information and ideas is a fund’s ultimate competitive edge. This level of higher understanding is fueled by the interaction between internal and external research. We recognize that the individualistic styles of investment managers dictate what kind of information will be most valuable to them. Investars allows them to tailor the use of the system to their individual needs. Plexus: Thus avoiding the situation where all managers following the most accurate ideas leads to even greater herding effects? Kei: Precisely. We do not pre-judge a manager’s research supply. Our tools enable their research staff to understand better the research they receive and assess how it best fits their needs. Plexus: Give us your thoughts on earnings surprises.

Kei: Consensus estimates don’t lead to compelling investments. The best ideas come from spotting a change in earnings prospects and getting ahead of the curve. The best returns come from (correct) bets that the consensus is mis-estimating earnings. We provide measures of both how accurate estimates are and how quickly analysts get to those correct estimates. Plexus: What effect will this have on research providers? Kei: Accountability to a higher standard is the best way to get better. Look what happened to the quality of American-built cars when the Japanese turned quality management into a competitive advantage. Either the brokers deliver value or they should pursue more meaningful careers. Plexus: Peter Bernstein, in a recent Financial Analysts Journal article, opines that since May Day in 1975, brokerage revenue has been too low to finance investment research without the added revenues of investment banking. As a result, Peter believes that soft dollar research is not objective, and is likely to remain so despite the administrative solutions proposed. He anticipates a bigger payoff in a hard dollar world focused on truly independent research. Kei: Our point exactly. No free lunch; why would anybody expect something that brokers give away to have value? This begs the question of whether there is a shortage of research commission dollars or simply a misallocation. With truly accountable and verified research value added, payments flow to the quality providers.

Reprint any portion with credit given to:

plexusgroup 11150 W. Olympic Blvd., #900 Los Angeles, CA 90064 PH: 310.312.5505 FAX: 310.312.5506 www.plexusgroup.com Plexus Group is a wholly owned subsidiary of JPMorgan Investor Services Company, a division of JPMorgan Chase. © 2003 Plexus Group, Inc.

Plexus: How does this merge with trading? Plexus has been noodling about the relationship between transaction costs and research value. Think of it like this. If a manager creates a trade in a stock which he thinks will go from $20 to $25, he is likely to keep trading that stock until it sells at a price close to $25. Up to that point there is no reason to stop trading; above $25, there is no reason to continue. This leads to the thought that the maximum trading cost that managers are willing to absorb is bounded by the value of research. This ultimately leads to a concept of “capturable research value” that factors in round trip cost estimates and tax efficiency. Kei: That brings us full circle. You don’t know what to pay for execution unless you know the expected worth of what you’re buying. You can’t address the AIMR’s Trade Management Guidelines framework of evaluating trading costs without the context of idea value. The obvious solution is to measure and justify the valueadded by your research sources. If certain broker relationships prove to be of little value, and the execution is not superior, that soft-dollar allocation can be put to better use. Plexus: Thank you, Kei. We look forward to seeing how this works in a real-world, real-time environment. When can we expect to see it? Kei: We have a demonstration available now for those who would like to see how it works. We are eager to get feedback, so we can better hone this offering to client needs.

Plexus News Look for the conference summary for the 9th Plexus Client Conference at http://www.plexusgroup.com. The conference was held at Silverado Country Club, Napa Valley, September 21-24, 2003. ************************************** Release 2.0 of our IceBreaker ‘drill-down’ tool, and a new on-line application for reviewing daily trading activity was made available in late 2003. More detailed descriptions and instructions will be communicated very soon, or contact your consultant to get the latest information.