A dist inct ion t hat is fundament al for firm behavior in inefficient m arket s is bet w een exploit ing mispricing, defined as an act ion t aken in response t o a preexist ing level of mispricing, and incit ing, an act ion designed t o shift t he level of mispricing (Hirshleifer (2001)). Incit ing t akes advant age of t he funct ion describing t he relat ion bet w een m arket price and t he firm’s act ion.1
1Incit ing encom passes act ions t aken t o shift mispricing eit her upw ard or dow nw ard. In cont rast ,
“ cat ering” (Baker & Wurgler (2012)) is def ined as an act ion t aken t o increase price above fundament al value.
Also, it is comm on t o dist inguish incit ing or cat ering from t iming, wherein t he f irm is sure t o undert ake t he act ion, but uses discret ion as t o w hen. How ever, t his is not an exhaust ive part it ion of cases; a firm can exploit in it s choice of w het her rat her t han w hen t o t ake an act ion.
Post -event ret urn drift is oft en int erpret ed as t im ing w it hout considerat ion of t his very plausible possibilit y. M ore import ant ly, t he possibilit y of incit em ent of misvaluat ion is oft en ignored in favor of t iming in response t o preexist ing misvaluat ion.
39
To illust rat e t his dist inct ion, consider a firm t hat issues equit y t o exploit preexist ing overvaluat ion. Ow ing t o t he negat ive average react ion t o t he announcement , t here t ends t o be a reduct ion in overvaluat ion, but t his w ill normally be an unavoidable adverse side-effect f rom t he firm ’s view point , in w hich case t his is not incit ement . In cont rast , a repurchase can be incit ement if it s purpose is t o induce higher valuat ion (rat her t han merely dist ribut ing cash, or profit ing from purchasing underpriced shares).
Upw ard earnings management designed t o induce overvaluat ion (or eliminat e
undervaluat ion) is also incit ement . M ost financial execut ives in one survey report ed t hat t hey would sacrifice economic value in order t o avoid missing quart erly earnings forecast s (Graham et al. (2005)). Sim ilarly, managing earnings dow nward w it h t he purpose of reducing t he st ock price (e.g., t o persuade pot ent ial com pet it ors t hat t he business is unprofit able, or t o reduce t he cost of share repurchase), is dow nw ard incit em ent . Verbal communicat ion can also be used t o incit e m isvaluat ion, as wit h m isleading disclosures, and discussions w it h m edia and analyst s (t ypically upward “ hype” ).
a. Theories of exploit ive advisors and firms
Sect ion 5 point s out t hat neglect of public signals result s in ret urn predict abilit y based upon t he account ing informat ion, and t herefore t hat m anipulat ion of disclosures can incit e over- or undervaluat ion (Hirshleifer & Teoh (2003); Hirshleifer et al. (2011)).
St ein (1996) models t he exploit at ion of exogenous st ock m arket mispricing by f irms in t heir financing and invest ment decisions. In St ein’s model, misvaluat ion affect s real invest m ent decisions m ore w hen managers have short t ime horizons, and firms should somet imes
40
pat ernalist ically discount using bet a even w hen bet a is not a ret urn predict or. In Daniel et al.
(1998), new issues and repurchase am ount s are select ed by a firm as a funct ion of m ispricing t o exploit invest or overconfidence. This im plies posit ive abnorm al ret urns aft er repurchase and negat ive aft er new issues.
Ljungqvist et al. (2006) model t he exploit at ion of individual invest or opt im ism in init ial public offerings. Cornelli et al. (2006) provide evidence t hat inst it ut ional invest ors and
underwrit ers exploit m isvaluat ion of IPOs by individual invest ors.
Invest ors wit h limit ed at t ent ion w ill somet im es overlook opport unism. One way t o exploit cust om ers is t o add com plexit y; in t he m odel of Carlin (2009), int ent ionally added complexit y of financial product s result s in equilibrium price dispersion am ong compet ing providers.
Exploit at ion and incit ement can have adverse macroeconomic effect s as w ell. In t he t heory of Gennaioli et al. (2012a), int ermediaries design securit ies t hat seem nearly riskfree t o t ake advant age of invest or neglect of nonsalient risks. This result s in booms and crashes.
b. Evidence on exploit ive advisors and firms
Evidence suggest s t hat invest ors are overly credulous about t he st rat egic incent ives of informat ion sources, leaving t hem vulnerable t o manipulat ion by f irms, advisors, and
int ermediaries (such as analyst s, brokers, and m oney managers). Daniel et al. (2002) argue t hat credulit y derives from lim it ed at t ent ion and overconfidence, and t hat it explains a w ide range
41
of financial behaviors and pricing anomalies. Jensen (2005) argues, for example, t hat firm overvaluat ion promot es exploit ive behavior on t he part of managers.
For example, evidence suggest s t hat invest ors are naïve about st rat egic behavior by firms in t heir financial report ing. Issuers manage earnings upw ard at t he t ime of IPO and seasoned issue; great er upw ard management is associat ed w it h w orse post -event average abnorm al ret urns (Teoh et al. (1998a,b)). This suggest s t hat firms successfully incit e
overvaluat ion prior t o issue, rat her t han just exploit ing preexist ing m isvaluat ion.
As ment ioned earlier, analyst forecast s do not discount adequat ely for earnings
management . Furt hermore, evidence suggest s t hat invest ors are naïve about analyst incent ives t o bias forecast s (Richardson et al. (2004)) and recomm endat ions (M almendier & Shant hikum ar (2007)). Invest ors seem t o be credulous about t he st rat egic mot ives of managers in various ot her cont ext s as w ell, such as t rust ing t hat name changes are indicat ive of firm and fund policies (Cooper et al. (2005)), t hat f und market ing expenses are unimport ant (Barber et al.
(2005)), and t hat broker recommended funds are superior (Guercio & Reut er (2013)).
The t heoret ical m odels of financing in inefficient market s discussed above predict abnorm al ret urns aft er new issues and repurchase ow ing t o firms selling t heir shares when overpriced and buying back w hen t hey are underpriced. Consist ent w it h securit y issuance being associat ed w it h overvaluat ion, t here is ret ur n cont inuat ion aft er new issues and repurchase (Sect ion 4). In general, t he occurrence of an event can predict subsequent abnormal ret urns eit her because of exploit at ion of exist ing m ispricing, or because it incit es m ispricing. So post -event abnormal ret urn evidence does not , in it self, est ablish w het her overvaluat ion causes
42
issuance, w het her issuance causes overvaluat ion, or w het her ot her act ions associat ed w it h issuance cause overvaluat ion (e.g., earnings management incit ing overvaluat ion at t he t im e of issue). These dist inct ions are oft en overlooked.
c. M isvaluat ion, new issues and repurchase, and post -event ret urns
Several st udies point m ore specifically t o exploit at ion of preexist ing overpricing as part of t he explanat ion. Surveys of U.S. CFOs find t hat misvaluat ion of t heir firms’ st ocks is an import ant fact or in deciding w het her t o issue equit y, and t hat CFOs t ry t o t ime int erest rat es in issuing debt (Graham & Harvey (2001)). Furt herm ore, measures of prior m isvaluat ion based upon t he deviat ion of price from cont em poraneous fundam ent als are associat ed w it h subsequent new issuance of debt and especially equit y, especially among overvalued firms (Dong et al. (2012)).
Invest ment and grow t h-relat ed measures are negat ive predict ors of abnormal st ock ret urns (Tit m an et al. (2004); Cooper et al. (2008); Polk & Sapienza (2009)). Such evidence does not resolve w het her invest ment induces overvaluat ion (eit her as incit ement , or as an
unint ended side-effect ), or w het her invest m ent choices exploit preexist ing misvaluat ion.
Evidence t hat higher discret ionary accruals is associat ed w it h great er invest m ent is consist ent wit h incit ement . How ever, consist ent wit h exploit at ion also playing a role, proxies for prior misvaluat ion predict invest m ent (Gilchrist et al. (2005)).
M isvaluat ion can also affect t akeover behavior. In t he model of Shleifer & Vishny (2003), overvalued bidders use equit y and undervalued bidders pay cash. Pot ent ially consist ent w it h (but not proof of) m isvaluat ion affect ing t akeover behavior, Loughran & Vijh (1997) find
43
negat ive post -event abnorm al ret urns t o st ock acquirers. Proxies for misvaluat ion are also associat ed w it h t he use of equit y as payment , t ransact ion charact erist ics, and m arket react ions t o announcem ent in w ays largely consist ent w it h t he Shleifer & Vishny (2003) model (Ang &
Cheng (2006); Dong et al. (2006)); Rhodes-Kropf et al. (2005) also provide evidence of valuat ion (t hough not necessarily m ispricing) effect s.