MOVERSANDSHAKERS(cid:2) Robert Akerlof and Richard Holden Mostprojects,inmostwalksoflife,requiretheparticipationofmultipleparties. Whileitisdifficulttouniteindividualsinacommonendeavor,somepeople,whowe call ‘‘movers and shakers,’’ seem able to do it. The article specifically examines movingandshakingofaninvestmentproject,whosereturndependsonitsquality and the total capital invested in it. We analyze a model with two types of agents: D o managers and investors. Managers and investors initially form social connections. w n Managersthenbidtobuycontroloftheproject,andthewinningbidderputseffort lo a d intomakinginvestorsawareofit.Finally,asubsetofawareinvestorsaregiventhe e d chancetoinvestandtheydecidewhethertodosoafterreceivingprivatesignalsof fro theproject’squality.Wefirstshowthatconnectionsarevaluablesincetheymakeit m easier for a manager to ‘‘move and shake’’ the project (i.e., obtain capital from http iidnevnestitcoarls)e.xWahnetne,waesienngdleogmenainzeagtehreenmeetrwgoersk,aswmeofisntdcotnhnaetctwedh;ilehemcaonnasgeqerusenatrlye ://qje earnsarent.Inextensions,wemoveawayfromtheassumptionofexanteidentical .o x managerstohighlightforcesthatleadonemanageroranothertobecomeamover fo rd andshaker.Ourtheoryshedslightonarangeoftopics,includingentrepreneurship, jo u venturecapital,andanchorinvestments.JELCodes:D31,D85,G30,L26. rn a ls .o rg I.Introduction a/ t U Most projects—in business, politics, sports, and academia— niv require the participation of multiple parties. In business, they ers usually involve, among other things, raising capital from dispa- ity o rate sources. Many projects fail—or do not ever get off the f N e ground—because of the difficulty of bringing together the rele- w S vant parties. Although it is not easy to unite individuals in a o u common endeavor, some people—often called ‘‘movers and sha- th W kers’’—seem able to do it. This article develops an equilibrium ale s theory regarding who these movers and shakers will be and o n whythey receive outsize compensation for their endeavors. O c Skill,ofcourse,helpsinobtainingparticipationsincepeople to b e aremoreinclinedtoparticipateinskillfullyrunprojects.Another r 3 0 , 2 0 (cid:2)Wearegratefultotheeditors,AndreiShleiferandPolAntras,andthreeanon- 16 ymousrefereesfordetailedandthoughtfulsuggestions. WealsothankGeorge Akerlof, Wouter Dessein, Rosalind Dixon, Bhaskar Dutta, Bob Gibbons, Oliver Hart, Bengt Holmstrom, Johannes Horner, Rachel Kranton, Hongyi Li, Barry Nalebuff,MichaelPowellandespeciallyAndreaPratforhelpfuldiscussions,and seminarparticipantsatANU,theSpring2015NBEROrganizationalEconomics meetings,MIT,LMUMunich,Oxford,Paris2e,QUT,UMassAmherst,UNSW,and Yale.HoldenacknowledgessupportfromtheAustralianResearchCouncil(ARC) FutureFellowshipFT130101159. ! The Author(s) 2016. Published by Oxford University Press, on behalf of President and Fellows of Harvard College. All rights reserved. For Permissions, please email: [email protected] TheQuarterlyJournalofEconomics(2016),1849–1874. doi:10.1093/qje/qjw021. AdvanceAccesspublicationonJune12,2016. 1849 1850 QUARTERLYJOURNALOFECONOMICS attribute—social connectedness—can also make someone a mover and shaker. Someone who is well connected can increase participation not only by making agents aware of a project but, even more important, by making agents aware that others are aware and are considering participating. Expressed differently, connectionshelpinraisingawarenessandinmakingthataware- ness commonknowledge. D Inourbaselinemodel,thereareanumberofpotentialman- ow n agersofaproject—allequallyskilled—andanumberofpotential lo a investors. Initially, there are no connections between managers de d and investors. The model has four stages. In stage 1, investors fro m formconnectionswithmanagers.Forsimplicity,weassumeeach h investorcanlinktoonemanager.Instage2,managersbidtobuy ttp an asset. The asset is necessary for undertaking the project and ://q je entitles the owner to the project’s return. For instance, if the .o x project were the construction of a shopping mall, the asset ford might be the plot of land on which the mall is to be built. In jou stage 3, the winning bidder puts effort into raising awareness rna ls of the project among investors and gives a subset of the aware .o rg investors the chance to invest. In stage 4, investors given the a/ chance to invest decide whether to do so after receiving private t U n signals oftheproject’s quality. iv e Wefirstanalyzethemodeltakingthesocialnetworkbetween rsity managers and investors as exogenous (i.e., we exclude stage 1). o f N Connectionsincreaseamanager’svaluationoftheassetbecause e w theymakeiteasiertoraisecapitalfortheproject.Consequently, S o in equilibrium, the manager—or one of the managers—who is uth most connected wins the auction and puts effort into moving W a and shaking the project. Furthermore, provided the auction les o winner is strictly more connected than other managers, he re- n O ceives a higherexpected payoff. c to Whenweendogenizethesocialnetwork(i.e.,addstage1),we be find that all investors link to one particular manager, whom we r 30 mayrefertoasM.Therefore,eventhoughmanagersareidentical , 20 1 exante,onemanager(M)emergesasmostconnected.Mwinsthe 6 auction,movesandshakestheproject,andearnsahigherpayoff thanothermanagers.Investorslinktothesamemanagerinequi- libriumbecausetheyhaveapreferencetolinktowhicheverman- ager is most connected. The most connected manager ends up controlling the project; unless an investor connects to the man- agerwhocontrolstheproject,hewillnothaveanopportunityto invest. MOVERSANDSHAKERS 1851 We later extend the model by making managers heteroge- neousalongseveraldimensions:(i)theirskillatrunningtheproj- ect; (ii) their talent at communicating with investors; and (iii) how much capital they have personally. We assume that man- agerscanusetheirpersonalcapitalasseedmoneyfortheproject. Takingthesocialnetworkasexogenous,wefindthatthesechar- acteristicsaffecthowmuchmanagersvaluetheprojectandwhich D o oneofthembecomesmoverandshaker.Whenweendogenizethe w n network, wefindthatthese characteristics are alsopredictive of lo a d who emergesas mostconnected. ed Inthinkingaboutmoversandshakers,itisusefultohavea fro m concrete example in mind. To that end, consider William h Zeckendorf, who was, in the 1950s and 1960s, the preeminent ttp://q realestatedeveloperintheUnitedStates.Heundertookavariety je .o of ambitious projects, including Mile High Center in downtown x fo Denver, Place Ville-Marie in Montreal, and L’Enfant Plaza in rd jo Washington, D.C. He was also famous for his role in bringing u rn the United Nations to New York.1 Key to Zeckendorf’s success als .o (and his ability to move and shake) were his social connections, rg asherecognizedhimself:‘‘thegreaterthenumberof...groups... at U/ onecouldinterconnect...thegreatertheprofit’’(Zeckendorfand n iv McCreary 1970, p. 42). He knew all the important real estate ers brokers, bankers, and insurance agents; he served on numerous ity o corporate boards; and he was a fixture of New York society. f N e Zeckendorf also owned a nightclub, the Monte Carlo, where he w S wouldholdcourtseveralnightsaweek,entertainingfriendsand o u business acquaintances. th W HisMontrealproject,PlaceVille-Marie,providesanexcellent ale s exampleofhistalentsasamoverandshaker.Sincethe1920s,the o n CanadianNationalRailway(CNR)hadbeenattempting,without O c success,todevelopa22-acresiteindowntownMontreal,adjacent tob e to the main train station: ‘‘a great, soot-stained, angry-looking, r 3 0 open cut where railway tracks ran out of a three-mile tunnel’’ , 2 0 (Zeckendorf and McCreary 1970, p. 167). Although the site had 16 enormouspotential,Canadiandevelopersshiedaway,considering the challenges too daunting. Desperate, CNR approached 1.UponlearningoftheUnitedNations’difficultyfindingasuitableNewYork site—andtheirintentiontolocateinPhiladelphia—herealizedhecouldhelp.He offeredthemasitehehadassembledontheEastRiverforalargedevelopment. 1852 QUARTERLYJOURNALOFECONOMICS Zeckendorf in 1955. He was immediately enthusiastic, appreciat- ing that‘‘a sortofRockefellerCenter-cum-GrandCentral Station could create a new center of gravity and focal point for the city’’ (ZeckendorfandMcCreary1970,p.170).Makingthisvisionare- ality would require the participation of two constituencies. First, hewouldneedtoraiselargesumsfrominvestors:$100millionfor the tower he proposed to build as the site’s centerpiece. Second, D and even more vexing, was the challenge of leasing office space. ow n EverymajorcompanyhaditsofficesonSt.JamesStreet.‘‘Thevery lo a idea of a shift to center-town offices struck many as dangerously de d radical’’ (Zeckendorf and McCreary 1970, p. 174). Zeckendorf ini- fro m tiallyfacedafreeze,unabletogetanyonetoleasespace.Asheput h it,‘‘nobody...believedwewouldeverputupaprojectasbigaswe ttp said we would’’ (Zeckendorf and McCreary 1970, p. 174). But ://q je through his tireless efforts, the freeze began to thaw. The first .o x crack came when he convinced the Royal Bank of Canada to ford moveintothenewbuildingandbecomeitsprime tenant.Hehad jou been introduced to the CEO, James Muir, by his friend John rna ls McCloy, chairman of Chase; Zeckendorf set out to woo Muir, .o rg making him his Canadian banker. With RBC lined up, he man- a/ aged,withconsiderablepressing,toobtaina$50millionloanfrom t U n MetLife—halfofwhatwasneeded.Alsowithconsiderablepress- iv e ing,helinedupasecondbigtenant:AluminiumLimited.Atthat rsity point, it became clear that the project would indeed become a re- o f N ality.Othercompanies—whichhadpreviouslyturnedhimdown— e w agreed to take space, and he was able to obtain the additional S o capitalheneeded. uth Our theory sheds light on a range of topics, one of which is W a entrepreneurship.Foundingabusinessoftenrequiresmovingand les o shaking. One can think of real estate developers such as n O Zeckendorf as a type of entrepreneur. A number of ideas have c to been advanced regarding entrepreneurs’ function. Schumpeter be (1934), for instance, stresses their role as innovators involved in r 30 ‘‘creativedestruction’’;Knight(1921)seesthemprimarilyasrisk- , 20 1 takers;RajanandZingales(1998)highlighttheirroleinregulating 6 access to resources. Others, such as Baumol (2010), bemoan that despite economists’ long-standing interest, ‘‘[entrepreneurs] are almost entirely excluded from our standard theoretical models’’ (Baumol2010,p.2).Ourtheoryoffersanewperspectiveontheir role.Theaspectofentrepreneurshipcapturedbyourmodelisnew toeconomics,butitisrelatedtotheoreticalperspectivesinsociol- ogy.RonaldBurt,for instance, argues thatentrepreneursexploit MOVERSANDSHAKERS 1853 network position. In his terminology, they bridge ‘‘structural holes.’’Hewritesthat‘‘bringingtogetherseparatepieces[ofanet- work]istheessenceofentrepreneurship’’(Burt2001,p.210). Our theory also speaks to the role of venture capitalists. According to Kaplan and Schoar (2005), venture capital (VC) funds, on average, yield roughly the same return, net of fees, as the S&P 500; however, certain fund managers consistently out- D o performthemarket,achievinghigherrisk-adjustedreturns.The w n standard interpretation of this finding is that these fund man- lo a d agers are particularly skilled at originating investment ideas. ed Although this is a possibility, the model suggests a novel expla- fro m nation. Such fund managers may instead earn high returns by h moving and shaking. Such VC firms take an equity stake in a ttp://q startup; then they move and shake on the company’s behalf (in je .o particular,helpingthestartupfindadditionalinvestors).Forex- x fo ample, Andreessen Horowitz, one of the preeminent Silicon rd jo Valley VC firms, ‘‘maintains a network of twenty thousand con- u rn tactsandbringstwothousandestablishedcompaniesayeartoits als .o executivebriefingcentertomeetitsstartups.’’AccordingtoMarc rg Andreessen,‘‘wegiveourfounders...networkingsuperpower.’’2 at U/ Additionally,seedingofprojects—or‘‘anchorinvestments’’— n iv seemstobeempiricallyimportant.Moversandshakersareoften ers independentlywealthyandusetheirownfundstoseedprojects. ity o In other instances, a mover and shaker might obtain help in f N e seeding a project from a large investor. Our model speaks to w S this topic aswell,and wediscuss this brieflyin theconclusion. o u Our article relates to a number of different literatures. At a th W formal level, the problem we analyze is a global game and thus a le s relatestothenowlargeliteraturepioneeredbyCarlssonandvan o n Damme(1993) and Morris andShin (1998). O c The model we analyze also relates to large theoretical and to b e empirical literatures in finance. A natural benchmark for r 3 0 , 2 0 1 6 2.TadFriend,‘‘Tomorrow’sAdvanceMan,’’NewYorker,May18,2015,re- trieved from http://www.newyorker.com. In line with this view, Hochberg, Ljungqvist,andLu(2007)findthatventurecapitalistswithsuperiornetworkpo- sitionsearnhigherreturns.Itisnotamatterofindifferencetoastartup,ofcourse, whichVCfirminvests.AstartupwouldrathertakemoneyfromaVCthatisbetter atmovingandshaking.Lower-rankedVCs,inconsequence,findithardtocompete. Andreessenputsitthisway:‘‘Dealflowiseverything...Ifyou’reasecond-tierfirm, you never get a chance at that great company’’ (Friend, ‘‘Tomorrow’s Advance Man’’). 1854 QUARTERLYJOURNALOFECONOMICS thinking about investmentsand returns is, ofcourse, Q-theory.3 InQ-theory,investorsearnthesamerateofreturnwhetherthey invest $1 or $1 million. By contrast, investment is lumpy in our model. Agents invest in projects; projects yield a poor rate of return unless they are well capitalized. An important conse- quence is that the rate of return to a project/asset depends on the social network that exists among agents. We predict, more- D o over, that agents with a privileged position in the network will w n earnoutsizereturns,becausetheycanmoveandshakecontribu- lo a tionsfromothers.4Ourmodelis,tothebestofourknowledge,the de d first to emphasize the importance of network structure for fro m investment. h Our article connects to the economic literature on net- ttp works—particularly work on network formation. In the endoge- ://qje nousnetworkversionofourmodel,investorshaveapreferenceto .ox fo linktothemostimportantmanager.Thisfeatureofourmodelis rd referred to as ‘‘preferential attachment.’’ Two classic papers, jou rn Jackson and Wolinksy (1996) and Bala and Goyal (2000), show a ls that this force will generally lead to the emergence of a star .o rg network.5 a/ Although our focus is an investment setting, our model also t U n relatestoaliteratureonattentionwithinorganizations—seees- ive rs peciallyDessein(2002),DesseinandSantos(2006,2014),Alonso, ity Dessein, and Matouschek (2008), Rantakari (2008), Calvo- of N Armengol, de Marti, and Prat (2015), and Dessein, Galeotti, e w S o u 3.A host of papers have documented departures from Q-theory and high- th lightedtheimplicationsofsuchdepartures.Liquidityconstraintsareimportant W a (see,amongothers,Fazzari,Hubbard,andPetersen1988;Hoshi,Kashyap,and le s Scharfstein 1991; Blanchard, Lopez de Silanes, and Shleifer 1994; Kashyap, o n Lamont, and Stein 1994; Sharpe 1994; Chevalier 1995; Kaplan and Zingales O c 1997; Lamont 1997; Peek and Rosengren 1997; Almeida, Campello, and to b Weisbach2004;BertrandandSchoar2006)asareshort-termbiases(Stein1988, e 1989).Moreover,thereiscompellingevidencethattherearerealconsequencesof r 30 suchinefficiencies(see,forinstance,Morck,Shleifer,andVishny1988andthe , 2 0 largeensuingliteratureontheequitychannelofinvestment). 16 4.Another,quitedistinctformof‘‘lumpiness’’hasbeenwellstudied:adjust- mentcosts(seeUzawa1969;LucasandPrescott1971;Hayashi1982;andfora recentdynamicanalysis,MiaoandWang2014).Itiswellknownthatsuchlump- inesscanhavesignificantmacroeconomicimplications(see,forinstance,Lucas 1967;Prescott1986;Caballero,Engel,andHaltiwanger1995). 5.OtherpapersthatpredicttheemergenceofstarnetworksincludeGaleotti, Goyal, and Kamphorst (2006), Goyal and Vega-Redondo (2007), Feri (2007), HojmanandSzeidl(2007),BlochandDutta(2009)andGaleottiandGoyal(2010). ArecentpaperthatisparticularlyrelevantisHerskovicandRamos(2015). MOVERSANDSHAKERS 1855 andSantos(2014).Agentsinthesemodels,asinourown,wishto coordinate their actions. In Calvo-Armengol, de Marti, and Prat (2015), agents decide whom to pay attention to; attention is dis- persedinequilibrium,incontrasttoourmodelinwhichattention is concentrated on a mover and shaker.6 Dessein and Santos (2014)andDessein,Galeotti,andSantos(2014)considerasetting inwhichaprincipaldecidestheallocationofattention.Theyfind D o thatitisoptimalfortheretobesomeconcentrationofattention, w n becauseitaidscoordination.Attentionisalsoconcentratedinour lo a d model,butitisnotnecessarilyoptimallyplaced.Inparticular,we ed obtain equilibria in which the mover and shaker is more or less fro m skilled,resultingrespectivelyinamoreorlessefficientoutcome.7 h Another related paper, Hellwig and Veldkamp (2009), examines ttp://q attention in a trading rather than an organizational setting. je .o Somewhat analogous to the coordination of attention in our set- x fo ting, they find traders may coordinate attention on one piece of rd jo information oranother. u rn Our model relates to the economic literature on leadership als .o sinceamoverandshakerisarguablyatypeofleader.Itparticu- rg larly relates to work examining how leaders persuade followers. at U/ Several publications consider signaling by leaders as a means of n iv persuasion (see, for instance, Prendergast and Stole 1996; ers Hermalin 1998; Majumdar and Mukand 2004). There is also ity o work on leaders creating cascades to influence followers (see f N e Caillaud and Tirole 2007). In our article, the mover and shaker w S persuadesinvestorsbypublicizingtheproject.Thisfeatureofour o u modelbearssomerelationtoDewanandMyatt(2007,2008),who th W have explored how public speeches by politicians can influence ale s followers. Chwe (2001) emphasizes the role of public announce- o n ments in acting as coordination devices in a variety of settings, O c such as advertising. In addition, there is work on the use of tob e r 3 0 6.InCalvo-Armengol,deMarti,andPrat(2015),agents’attentionisdispersed , 2 0 inequilibriumbecausethereareneitherincreasingcostsnordecreasingbenefitsof 16 listeningtomultipleagents. 7.Intuitively,investorscoordinateonlinkingtoaparticularmanager,whose skillmaybehigherorlower.Thisfindingsuggeststhepossibilityofconstructinga mover-and-shakermodel,similartoourown,inwhichtherearepersistentperfor- mancedifferencesacrossfirms.Somefirmsgetstuckpayingattentiontothewrong people.Persistentperformancedifferenceshavebeenshowntobeubiquitous(see GibbonsandHenderson2012).Thereisconsiderableinterestinunderstanding what drives these productivity differences (see Gibbons 2006; Chassang 2010; EllisonandHolden2014). 1856 QUARTERLYJOURNALOFECONOMICS authority by leaders in settings where agents, as in our model, haveadesiretocoordinate.Forinstance,Bolton,Brunnermeier, and Veldkamp (2013) argue that resoluteness is an important quality in a leader because a leader who is overly responsive to new information canundermine coordination. Thearticleproceedsasfollows.SectionIIcontainsthesetup of our model and the analysis of equilibrium. We first take the D o networkstructureasexogenous;subsequently,weendogenizeit. w n In Section III we consider a number of extensions of the basic lo a d model analyzed in Section II. Section IV concludes. Proofs of all e d formalresults are containedin theOnline Appendix. fro m h ttp ://q II.The Model je.o x fo II.A. Statement oftheProblem rd jo u Considerasettingwithaninvestmentprojectandtwotypes rn a ofagents:managersandinvestors.Managershaveskillsneeded ls.o toruntheproject;investorseachhaveoneunitofcapitaltheycan arg/ contributetotheproject.Weassumethereareatleasttwoman- t U n agers; the total number of managers and investors is finite. Let iv e N denotethesetofmanagersandN denotethesetofinvestors. rs M I ity jareAconnentewcotrekd;ggex=is0tsobtheetwrweeisne.aFgeonrtnso.wgi,jw=e1tiafkaegethnetinaentwdoargkeanst of N ij e w exogenous; wewill endogenize it inSection II.D. S o The model has four periods. All choices made by agents are u th observable. In the first period, managers bid in a second-price W a auctionforanassetA.Theassetisneededtoundertaketheproj- le s ect and entitles the owner to the project’s return. For instance, on O the asset might be a parcel of land; the project might be the c to construction of a building on that parcel. The project yields a b e returnRattheendofthegamethatdependsonboththeproject’s r 3 0 underlying quality ((cid:2)) and the amount of capital raised for the , 2 0 1 project(K).Morespecifically,R¼(cid:2)þv(cid:3)K,wherev>1parame- 6 terizesthereturntoraisingcapital(i.e.,thereturntomovingand shaking). The agents have a common prior that (cid:2) is distributed Nð(cid:3);(cid:4)2Þ, with (cid:3);(cid:4) >0. Let b denote manager i’s bid in the auc- i tion, let b denote the second highest bid, and let M denote the ð2Þ winningbidder.Intheeventofatieintheauction,themanager oflowestindexwins.Weassumemanagersdonotfollowbidding strategies thatareweakly dominated. MOVERSANDSHAKERS 1857 In the second period, the auction winner, M, decides how much effort e 2½0;1(cid:4) to exert to make investors aware of the M project. An investor’s chance of becoming aware of the project depends on e and on his degree of separation from M. M Specifically, investor j becomes aware with probability (cid:5)ðlengthpathj!MÞ(cid:5)1(cid:3)e , where ðlengthpath j!MÞ denotes the M length of the shortest path between investor j and M that does D not include other managers, ðlengthpath j!MÞ¼1 when no ow n suchpathexists,and(cid:5)2ð0;1Þ.Weassumemutualindependence lo a of investors’ awareness of the project. The cost to M of exerting de d effortiscðeMÞ,wherec0ð0Þ¼0andc0ðeÞ>0fore>0.Letndenote fro m the numberof investors who becomeaware oftheproject. h Inthethirdperiod,Mcanofferawareinvestorsequityinthe ttp projectinexchangeforcontributingtheircapital.Mchooseshow ://q je muchequity,(cid:6) ,toofferandthenumber,m(cid:6)n,ofequityoffers .o M x hewillmake.Theminvestorswhoreceiveequityoffersareran- ford domly drawn from the pool of aware investors. Let S denote the jou set ofinvestors who receive equityoffers. Oncedrawn, S iscom- rna ls monly knowntoinvestors in set S. .o rg InvestorsinsetSthenreceiveprivatesignalsoftheproject’s a/ quality:x ¼(cid:2)þe,wherethee’saredistributedi.i.d.Nð0;(cid:7)2Þ.We t U j j j n focus on the case where (cid:7)!0 because this results in closed-form iv e solutions. rsity In the final period, investors who received equity offers o f N decide whether to take them. Let aj 2f0;1g denote investor j’s ew decision. Observe that the total capital raised for the project is S K ¼TPhje2Sparjo.ject is then undertaken, yields return R¼(cid:2)þv(cid:3)K, outh W a andplayersreceivethesharesofthereturnduetothem.Wecan les o writeplayers’payoffsattheendofthegameasfollows.Investors n O receive a payoff of (cid:6) R if they invest in the project and 1 other- c M to w(cid:5)ibsðe2I.ÞtTainhsdeuaosutehfcuetirlotmnoawsnuinamgnemerrasrrreiezcceeeivtivheees0ati.mpaiynogf:f(oi)fðm1a(cid:5)na(cid:6)gMeKrsÞRsi(cid:5)mcuðletMaÞ- ber 30, 20 1 neously place bids (b) for asset A and the winning bidder (M) 6 i acquires the asset; (ii) the auction winner (M) decides how much effort to exert (e ) to make investors aware of the project; M (iii) M offers equity shares ((cid:6) ) to m(cid:6)n of the aware investors; M (iv) investors who receive equity offers then acquire private sig- nalsoftheproject’squalityandsimultaneouslydecidewhetherto invest (a), after which the project is undertaken, its return R is i realized,andplayersreceivetheshareofthereturnduetothem. 1858 QUARTERLYJOURNALOFECONOMICS II.B. Discussion ofthe Model Webrieflydiscussanumberofthemodelingchoiceswehave made. First, our game has four periods and, at first inspection, might seem complicated in this respect. In fact, this is the sim- plest formulation that captures all the economics we wish to convey. It is important to us to highlight that in equilibrium, D o more connected players value asset A more than less connected w n players. The simplest way to demonstrate this is through the loa d auction we consider at time 1. Similarly, the effort choice is in- ed dispensabletoourstorybecausethisiswhatmovingandshaking fro m is—hence,time2.Finally,weneedtwoperiodstoaddressinvest- h ttp mentsinceitnecessarilyinvolvestheequityofferandthechoice ://q ofwhether to invest. je .o Second, we model the project’s return as increasing in the x fo amountofcapitalinvested.Thisresults instrategiccomplemen- rd jo taritiesandcapturesourbasicstoryabouttheimportanceofpar- u rn ticipation.NotethatitisimportantthatRisincreasinginKover als .o some range; it is not important that R is increasing in K indefi- rg nitely; wehaveonly made this assumption forsimplicity. a/ t U Third,thesetSiscommonlyknowntoinvestorsinsetS.This n iv assumptionreflects theideathatthemover andshakernotonly ers raises awareness of the project, he also makes the existence of a ity o pool ofpotential investors commonknowledge. f N Fourth,weassumethemarginalcostofeffortisequalto0at ew eM = 0 (c0ð0Þ¼0). This ensures that it is optimal for M to exert Sou positiveeffortwhenhehassocialconnectionsandthereturnsto th W movingandshaking,v,arelarge.Moreimportant,itmeansthat a le more connected managers value asset A strictly—rather than s o n weakly—more than lessconnected managers when vislarge. O c Fifth,weconsideraparticularformoffinancialcontracting: to b equity. The benefit of focusing on equity contracting is that it er 3 resultsinclosed-formsolutions;thisisnotthemostgeneralcon- 0, 2 tracting space one could consider, to be sure. However, we con- 01 6 jecture that ourmain results hold in a more general contracting space.8 Sixth, we adopt a common assumption regarding how infor- mationdiffuseswithinthenetwork(JacksonandWolinsky1996 makeasimilarassumption).Underthisassumption,amanager’s 8.Forinstance,webelieveourresultsholdwhentheprojectisdebtfinanced ratherthanequityfinanced.