BY CHRISTOPHER WRIGHT

t has come to be known as “thesalty snack report.” Now leg-endary among Morningstaranalysts as a prime example of

what not to do, the report aboutPepsiCo devoted six to seven pagesto an arcane discussion of the saltysnack market in Argentina, a topicthat was only relevant to a singledivision of the company (Frito-Lay)and could not possibly move the dialon Pepsi’s share price.

Although Pat Dorsey, CFA,research director at Chicago-basedMorningstar, doesn’t recall whichinvestment firm produced the analy-sis, he still considers it the worstreport he has ever seen. In fact, “Ohno, not another salty snack report”has become code at Morningstar formissing the big picture and gettinglost in the details, according toDorsey, author of The Five Rules forSuccessful Stock Investing and TheLittle Book that Builds Wealth.

Susan Mach, a partner at com-munication strategy firm MachCreative Services, predicts that theanalysts left standing after the globalrecession will all have one thing incommon—strong writing skills.“No matter how brilliant a researcheryou are, you will not get the careeryou deserve unless you are a strongwriter,” says the adjunct professor ofcommunication at the New YorkUniversity Stern School of Business.In addition to her other roles, Machhas twice run a hands-on, three-hourreport-writing workshop for the NewYork Society of Securities Analystsknown as the “Top Tips for WritingGreat Research Reports.”

Mach divides the attributes of agreat report into structure, style, andsubstance, but in her view, no report

can be great without one criticalingredient—insight.

Insight. Inexperienced analystshave a tendency to bury readersunder a blizzard of details, but moreinformation is not necessarily better.“Information is cheap; insight ispriceless,” says Dorsey. “The job ofthe analyst is to separate one fromthe other.”

Insight is the foundation uponwhich great research reports arebuilt. Mach remembers a group ofher business school students whowrote a glowing report on Enronshortly before the firm crashed andburned. “I should really go back andchange their grades,” she says.“They obviously got snookered.”

“You will never find a mispricedsecurity by having the same opinionas the market,” says Dorsey. “If yourexpectations are in line with themarket, you have—congratula-tions!—just found a fairly valuedstock.”

A five-paragraph companydescription is worthless. The job ofthe report writer is to identify whichof the company’s divisions will createthe most value in coming years andwhich will face headwinds. “You canwrite a great research report in fivepages if you understand the busi-ness, distill things down, and saywhether the stock price fairly reflectsyour expectations,” says Dorsey.

Structure. The structure of thereport dictates how quickly readerscan grasp the key points. After youfinish your research and sit down towrite, begin with your conclusion.Build the investment case the way anattorney prepares for a trial by start-ing with the closing argument andworking back from there. Put athesis paragraph at the top of thefront page that states your conclu-

I

PR

OF

ES

SIO

NA

LP

RA

CT

ICE

C F A M A G A Z I N E J A N – F E B 2 0 1 026

sion and rationale and gives thereader a map of where the discussionis headed. Also, make the report easyto skim, with sections clearlymarked and key points bulleted, inboldface type, or otherwise high-lighted for instant accessibility. Don’tleave key points buried in the text ormake the reader hunt for them later.

Style. Not only is insight the keyto reaching the right conclusion, itis the mother of clarity. Clear writingdepends on clear thinking. Yourconclusions may be simple and read-ers may effortlessly glide throughyour prose, but that’s only becauseyou put the time and effort intodisentangling the disparate strandsof thought you encountered alongthe way and getting to the heart ofthings. “If it was easy to read, it washard to write,” Mach says.

Also, with regard to style, thequickest way to destroy your credi-bility is to hand in a report repletewith misspelled words, grammaticalerrors, and poor diction. Keep youraudience in mind. The report youwrite for journalists will be differentfrom the one you write for investors,analysts, or executives to deliver ona conference call. Be simple and direct(the title of a great book on writingby Jacques Barzun). Avoid jargonand big words when commonexpressions will do. Use the activevoice (“the company earned a mil-lion dollars,” not “a million dollarswas earned by the company”). Turnphrases into adjectives (“an above-consensus earnings estimate,” not“an earnings estimate that is above

How to Write a Great Research Report“Information is cheap; insight is priceless,” says one expert

KEY POI NTS

• Insight is the foundation upon whichgreat research reports are built, butstructure, style, and substance areneeded to communicate that insight.

• Great reports have an easy-to-skimstructure, a simple and direct style, andsubstance readers can’t find anywhereelse.

• An analyst’s fieldwork with noncom-pany sources is key to determiningwhether the security is fairly valued.

AnalystAGENDA

R E C O M M E N D E D R E S O U R C E S

“Analyzing Investment Research Can HelpInvestors Weather a Turbulent Market”CFA Institute Press Release (6 February 2008)(www.cfainstitute.org)

“Research Objectivity Standards for Sell-SideInvestment Research Firms”CFA Institute Centre for Financial Market Integrity(www.cfainstitute.org/centre)

C F A M A G A Z I N E J A N – F E B 2 0 1 0 27

consensus”).In short, be concise. Your report

may be competing with e-mail andother reports while your boss is atthe gate waiting for a plane. “Writeshort paragraphs of no more than 8–10 lines,” advises Debra Fiakas, CFA,managing member of Crystal EquityResearch in New York City. “Peoplewant to read reports fast, fast, fast.”

Substance. The key to share-holder value will be different foreach company. It might be a deepbench, a distinctive culture, or theability to grow the top line in arecession. But for Tim Baker, CFA,no report is complete without anearnings forecast, a risk section, anda valuation. Currently director ofnew business development for NewYork-based Integrity Research Associ-ates, which helps institutions findresearch providers, Baker reviewed alot of reports in his former positionas a head of research at UBS.

Earnings forecasts should bepresented in a range. “An analyst’ssingle-point estimate is almost bydefinition going to be wrong,” saysBaker. Even if the analyst is eventu-ally proven right, he points out, “theclient will almost certainly have adifferent view of the earnings out-come.” Include a sensitivity analysisthat shows the probable effect onearnings of swings in commodityprices, prices the company can getfrom customers, or other importantvariables. For this purpose, accord-ing to Baker, Porter’s five-forcesmodel* still provides a good frame-work. (The five forces are threatsfrom substitutes, rivals, and newentrants plus the bargaining powerof suppliers and customers.) Lay outscenarios that show, for example,what a 10 percent miss in expectedsales volume would do to earnings.Present the results in a table so read-ers can grasp the relevant factorsand range of potential outcomes—pessimistic, base case, and opti-mistic—at a glance.

A risk section will discuss suchthings as governance problems,related-party transactions that maydegrade the quality of earnings, theladdering of debt maturities and howrefinancing will affect financial state-ments, and other company-specificrisks. Sector-specific risks, such ascommodity prices or country politi-cal risks, should also be included.

As for valuation, explain whythe stock is cheap or expensive rela-tive to historical multiples and thecompany’s peers. Identify catalysts(i.e., what will propel the share pricegoing forward), whether they involvea new product launch, a pendingacquisition, or other event. A lot of

“The good analysts are out there doingtheir own sleuthing,” Baker says.

Another stop on the road tounique money-making insight—or what Michael Steinhardt calls“variant perception” in his book NoBull—is to check whether all thecomponents of your analysis pointto the same conclusion. If your dis-counted cash flow (DCF) analysis isat odds with your earnings multipleanalysis, watch out. Suppose yourDCF valuation puts the stock at 30times earnings, but the multiple hasnever gone above 12 in reality.Choosing the naive DCF result asyour target share price would notnecessarily be wrong, but it wouldcertainly be suspect and wouldrequire further justification. Explainwhy you think the market willrespond to growth in the businessthe way you anticipate.

If Dorsey’s subordinates merelystate conclusions, he sends theirreports back to them. If an analystbelieves sales will grow or marginsexpand, Dorsey will want the reportto contain the supporting reasoningand describe the dynamics that willdrive the results the analyst expects.

“Get to the point quickly, makeyour opinion clear, back it up, andtell readers something they don’tknow,” Dorsey says. If you can dothat, the greatest research report everwritten will be the one you’re writingnext.

Christopher Wright, an award-winningwriter in Arlington, Virginia, publisheshis own investment newsletter andcountry risk reports.

* A framework for analyzing industries anddeveloping business strategies, the five-forcesmodel was developed by Michael Porter ofHarvard Business School in 1979.

“You will never find amispriced security by

having the sameopinion as the market.If your expectations arein line with the market,you have—congratula-

tions!—just found afairly valued stock.”

PAT DORSEY, CFA

stocks are cheap, Baker says, butsome deserve to be. So, discusswhether a low assigned multiplemeans the company is truly a bar-gain or just a value trap.

For David Taube, CFA, founderand president of Kalorama WealthStrategies in Washington, D.C., agood report will “include as muchinformation as can be obtainedindependently outside the companybeing researched.” Baker agrees.Instead of relying solely on informa-tion that management hands out, dosome primary research and fieldworkof your own to support your invest-ment thesis, Baker advises. Forexample, citing evidence from thechannel that a new product launchwill be well received will support anabove-consensus earnings estimate.