Market Timing Behavior of the Secondary Equity Offerings of REITs

Transcription

1 Marke Timing Behavior of he Secondary Equiy Offerings of REITs Ying Li Naional Universiy of Singapore Deparmen of Real Esae 4 Archiecure Drive, Singapore Tel: (65) ying@nus.edu.sg Seow Eng Ong Naional Universiy of Singapore Deparmen of Real Esae 4 Archiecure Drive, Singapore Tel: (65) rsongse@nus.edu.sg

2 Marke Timing Behavior of he Secondary Equiy Offerings of REITs Absrac In corporae finance lieraure, he concep of marke iming means ha managers would make heir equiy finance decisions according o he condiions of he capial marke. REITs are special invesmen vehicles and are no included in he sudies of general socks. Thus he quesion is: do REITs have he same marke iming behavior in heir secondary equiy offerings like general socks? In his sudy, we would use he secondary equiy issues daa of US equiy REITs o analyze he marke iming behavior of REITs. We are rying o learn abou he shor-run and long-run performance of REITs around equiy issues, o find ou he marke iming paern of REITs SEOs. Our findings could help us undersand he meanings behind he equiy issuing decisions of REITs managers, and help invesors make appropriae reacion o hese signals. Our resuls show some evidence of he marke iming behavior under he asymmeric informaion hypohesis. However, compared o general socks, he iming behavior is less obvious for REITs. Wih regard o he long-run performance, no long-run underperformance is found afer REITs SEOs. These resuls have wo implicaions. Firs, alhough REITs are comparaively ransparen han indusrial firms, he REITs managers sill have he marke iming abiliies o uilize he marke condiions. Asymmeric informaion and adverse selecion sill exis for REITs, alhough he iming behavior is less obvious for REITs compared o general socks. Second, he absence of long-run reurn anomaly afer SEOs of REITs means ha he marke efficiency is no violaed. 1

3 1. Inroducion Real Esae Invesmen Truss, or REITs, are always reaed as special invesmen vehicles in finance research, and are always reaed differenly from general socks when doing empirical finance sudies. Since 1960 s, he REITs indusry has undergone rapid growh, boh in U.S. and oher counries. REITs, especially U.S. REITs, are now he focus of many research sudies. The special characerisics of REITs, including he special insiuional srucure requiremens and he close relaionship wih heavy-financed properies, ec, have been sudied o differeniae REITs from general indusrial socks. Wih regard o he equiy financing paerns, he firs day overpricing of IPO and long-run over-performance afer IPO waves also disinguished REITs from general socks. In his paper, he marke iming abiliies of REIT managers in secondary equiy offerings will be sudied o uncover he validiy of equiy marke iming heory for US equiy REITs. 3. Lieraure Review In corporae finance lieraure, equiy marke iming means ha managers end o make financing decisions for firms according o he marke condiions. In oher words, hey would issue equiy when he sock prices are overvalued, and repurchase hem when sock prices are undervalued. Marke iming firs appears as a dynamic form of Myers and Majluf (1984). In Myers and Majluf (1984), managers are assumed o have informaion ha invesors do no have, and hey ac on behalf of curren invesors insead of fuure invesors. When hey face an invesmen opporuniy, hey have o issue equiy o finance he projec. Under hese assumpions, if he sock price of he company is undervalued, hey would wai unil he real value is realized and forego he curren invesmen. If he curren sock price is overvalued, managers would issue equiy immediaely. Thus equiy issues would appear afer posiive abnormal sock reurns. Also i is believed ha compared o sock shares, deb and inernal capial do no have he problem of being under-priced, hus if here is an invesmen opporuniy, managers would choose inernal capial firs or deb, and he las resor is equiy. This is he pecking order heory. 2

4 On he basis of Myers and Majluf (1984), Lucas and McDonald (1990), Korajczyk, Lucas and McDonald (1991) and Korajczyk, Lucas and McDonald (1992) sudied he variaions of adverse selecion coss of equiy issues wih he exisence of asymmeric informaion. Since he equiy issues only happen when he sock prices are overvalued, he sock prices would decrease on announcemen of he equiy issues. Ousiders would lower heir evaluaion of he issuing firm s qualiy, and his creaes a lemons marke in new equiy issues. In Korajczyk, Lucas and McDonald (1991) and Korajczyk, Lucas and McDonald (1992), he asymmeric informaion is no fixed over ime, and firms end o issue equiy when he marke is mos informed abou he qualiy of he firm, for example, afer earnings releases. They find ha managers can conrol he informaional disadvanage of he marke by choosing he iming of an equiy issue. In Baker and Wurger (2000), i is found ha he equiy share in oal new equiy and deb issues is a srong predicor of US sock marke reurns beween 1928 and This finding could no be explained under he marke efficiency explanaions. The assumpion of marke efficiency is relaxed and managers or invesors can be irraional. Managers ry o exploi he marke condiions, and hey ime boh heir idiosyncraic reurn and he marke reurn (Baker and Wurgler 2000). The marke inefficiency is challenged. The underperformances afer IPO repored by Rier (1991), he long-run underperformances afer equiy issues recorded in Loughran and Rier (1995), ec. All hese findings seem o indicae ha firm managers have he abiliy o ime he marke when hey make equiy financing decisions. Besides above evidence from US, daa using oher counries daa also find suppor for he marke iming behavior. Besides he marke iming behavior of managers before equiy issues, oher corporae decisions, for example spin-offs, deb offerings, and sock splis all indicae marke iming abiliies of managers from he long-run abnormal reurns (Desai and Jain (1999), Spiess and Affleck-Graves (1999), and Ikenberry, Rankine and Sice (1996)). Above findings documen he long-run abnormal reurns and such findings challenge he marke efficiency hypohesis (Baker and Wurgler (2002)).A he same ime, some doubs are also cased on his version of marke. Before Baker and Wurgler (2000), oher researchers also performed similar ess bu hey do no find srong evidence agains marke efficiency (Fama and French (1988) and Kohari and Shanken (1997)). Besides, on-going debae exiss o give mehodological 3

5 suggesions on he long-run anomaly (Fama (1998), Eckbo, Masulis, and Norli (2000), and Brav, Geczy, and Gompers (2000)). The robusness of he long-run reurns calculaions are quesioned by researchers. Wih ho debae on he marke iming behavior of general socks, he marke iming of REITs is no sudied horoughly. The reasons for us o choose o sudy he marke iming of REITs are as following. Firs, REITs are also considered o be differen from general socks and are always excluded from general socks in empirical corporae finance sudies. The resul is ha above research findings and heories are no necessarily valid for REITs. Second, less asymmeric informaion is expeced from REITs compared o indusrial socks. Such expecaion is because of he higher angibiliy of REITs asses and he predicabiliy of cash flows. Third, he research findings abou he long-run performance of REITs sock reurns afer IPOs and SEOs seem be in a mixure of underperformance and over-performance. Take iniial public offering as an example. Wang, Chan and Gau (1992) found ha he long-erm underperformance exiss for REITs afer IPO. However, Ling and Ryngaer (1997) showed over-performance insead of underperformance of REITs afer IPO. A more recen sudy by Buimer, Hyland and Sanders (2005) documened he long-run performance of equiy REIT IPOs during and afer IPO waves, and found no evidence of he long-run underperformance for REITs. All hese special characerisics of REITs make he marke iming of REITs differen and even more ineresing. In his sudy, we focus on he secondary equiy financing of REITs, because secondary equiy financing has imporan influences on he capial srucure changes of firms, and are imporan reflecions of financing decisions of managers. Thus i is meaningful o find ou wheher REITs managers have he marke iming abiliies when hey issue secondary equiies, and undersand he shor-run and long-run performance of REITs afer SEOs. Firs we would es wheher marke iming exiss for REITs by looking a he relaionship beween sock reurns and equiy issues. Then we would examine he shor-run and longrun performance of REITs afer equiy issues o es some hypoheses of he heory. According o he asymmeric informaion framework, managers have superior informaion abou he sock prices compared o invesors. Under such condiion, managers would choose o issue equiy when he sock prices are overvalued and 4

6 repurchase hem when sock prices are undervalued. Thus he posiive abnormal reurns preceding he equiy issues would be used as one mehod o es he marke iming abiliies. If managers could ime he marke when hey issue he equiy, he sock prices should be increasing before he equiy issues. As he sock prices have reached is peak when new equiy is issues, he sock prices would decline a he announcemen or issues of he new shares. Anoher hypohesis is managers would choose o issues equiy when he marke is bes informed of he qualiy of he firm o reduce he adverse selecion problem. The long-run performance afer secondary equiy issues of REITs would also be esed. Previous research findings abou he long-run performance of REITs SEOs include Howon, Howon and Friday (2000), which found long-run underperformance of REITs SEOs when comparing he holding period reurn of equiy issuing REITs wih non-equiy offering REITs. The mehod applied by Howon, Howon and Friday (2000) is quesionable because i compares equiy issuing REITs wih non-equiy offering REITs, we would use he Fama and French four facor model o calculae he long-erm reurns. One poin mus be made clear is ha he second version of marke iming does no require he marke o be inefficien. The long-run underperformance indicaes managers are successful in iming he marke. While failure o find he long-run underperformance only means he iming behavior is no successful. Previous research works abou he marke iming behavior of REITs secondary equiy offerings include Li and Ooi (2004). They found ha US REITs ime boh he equiy and deb marke condiions using a sample of equiy REITs from 1986 o The P/B raio and P/E raio is considered as he proxy of marke misevaluaions and equiy/deb issues/repurchases are analyzed. They conclude ha REITs ime he marke in boh equiy and deb offerings. However, hey did no sudy he performance of RETIs afer equiy issues. Before our empirical ess, le s firs look a he following graph o have a brief undersanding of he equiy issues and sock price rends for REITs. In Figure 1, he NAREIT equiy REIT price index and he number of secondary equiy offerings in he corresponding year is ploed. Alhough his only gives us a brief undersanding, i is 5

7 obvious ha hese wo variables are correlaed, and he marke iming is observed from he char. Figure 1: NAREIT index and number of REITs SEOs NAREIT index and number of REITs SEOs Number of SEO Year Number of SEOs in he year NAREIT index The sudy proceeds as follows. Secion 4 inroduces he daa. Secion 5 sudies wheher marke iming exiss in he ne equiy issues of REITs. Secion 6 ess he shor-run performance of REITs afer equiy issues. Secion 7 ess he long-run performance of REITs afer equiy issues. Secion 8 gives some discussion and in Secion 9 he expeced fuure work is described Index Value 4. Daa All he US equiy REITs sudied in his paper are repored by Naional Associaion of Real Esae Invesmen Trus. US REITs are of relaively long hisory, and his allows us o know he marke iming behavior in more developed REITs markes. There are 123 equiy REITs capured in his sudy, which comprise a grea majoriy of he equiy REITs repored by NAREIT. The lis of REITs names ha are capured is lised in he Appendix. The quarerly daa of hese equiy REITs cash flows comes from Compusa daabase. The daily price daa comes from CRSP. Daes of firs public announcemen of quarerly earnings are from Compusa daabase. The record of all equiy issues of REITs 6

8 (excluding IPOs) comes from SDC daabase. Long-erm ineres rae is found from he websie of U.S. Federal Reserve Board and he daa of NAREIT equiy REITs price index comes from NAREIT websie. The daa of equiy REITs ha are capured spread from January 1980 o March 2004, 279 monhs in oal. The ime series plos of some imporan variables are presened in he appendix. 5. The marke iming of REITs In his par, he exisence of marke iming behavior would be examined by looking a he relaionship beween equiy issues and sock reurns. Marke iming heory implies ha managers would choose o issue equiy when heir sock prices are (believed o be) overvalued and repurchase he socks when hey are (believed o be) undervalued. The equiy issues always occur afer posiive sock reurns. For he individual REITs capured in his par, he following hypoheses would be esed: would ne equiy issues change according o he capial marke condiions? The daa here is unsrucured panel daa insead of ime series daa, as he daa of ne equiy issues and sock reurns of individual REITs are unsrucured panel daa. In his par, he VAR model and panel regression would be employed o es he relaionship beween relevan variables. The auoregressive model is y = µ + τ 1 y 1 + Λ + τ p y p + ε (1) Where ε is a vecor of non-auocorrelaed disurbances wih zero means and conemporaneous covariance marix E[ ε ε ] = Ω '. This equaion sysem is a vecor auoregression, or VAR. Vecor means ha we are dealing wih a vecor of variables while auoregressive means he appearance of he lagged value of he dependen variable. Akaike or Schwarz Crierion can be used o decide how many lags o ake. One of he virues of he VAR is ha i obviaes a decision as o wha conemporaneous variables are exogenous; i has only lagged (predeermined) variables on he righ-hand side, and all he variables are endogenous (alhough in some circumsances exogenous variables can also be included). VARs are special, because in simulaneous or srucural equaion models, he equaions in he sysem should be deermined before he esimaion of he model. This kind of idenificaion is ofen subjecive. Using VAR his subjeciviy 7

9 could be avoided. Employing VAR model allows us o es which variables are more relevan in deermining he equiy issues or sock reurns, and we do no need o specify he relaionships before he esimaion. The VAR es for individual REIT is run according o regression equaion 2 shown below. ( Equiy, Deb, M / B, Sockreurn ) A ( Equiy n + C Size 1 1 n + D SP 1 1, Deb n D, M / B n SP n n = µ + A ( Equiy, Sockreurn + ε n The meanings of he variables are explained as following: 1 1, Deb 1 ) + B Ineres 1, M / B Equiy : The ne changes of equiy of individual REITs in quarer ; B, Sockreun n (2) Ineres Deb : The ne changes of book value of long-erm deb of individual REITs in quarer ; M / B : The average M/B raio of individual REITs in quarer ; Sockreur n : The sock reurn of individual REITs in quarer ; Ineres : Change in he long-erm ineres rae of U.S. in quarer ; Size : The oal asses of individual REITs in quarer ; SP : S&P 500 composie index reurn in quarer. The deailed explanaions of hese variables are: Specifically, he calculaions of he variables from he daabase are: (a)marke value of equiy= price closed a he end of he hird monh of each quarer * shares ousanding (splis adjused) (b) Long-erm deb= long-erm deb oal (c)marke-o-book raio= marke value of asses/ oal asses= (marke value of equiy + long-erm deb+ preferred sock + deb in curren liabiliies)/oal asses (d)long-erm ineres rae= 10 year ineres rae The saisical summary of variables differen from indusry analysis is lised in he following able. 1 n ) 8

10 Table 1: Saisic summary Variable Sock reurn Average m/b Size Equiy issues Deb issues Average Sandard Deviaion The resul of he uni roo es is shown in Table 2. Table 2: Panel Uni Roo Tes Sock reurn Average m/b Size Equiy issues Deb Issues I(0) I(0) I(1) I(0) I(0) The regression resuls are lised in Table 3 Table 3: VAR-1 Included observaions: 5200 afer adjusmens, Sample (adjused): 1981Q3 2004Q1 T-saisics in [ ] EQUITY DEBT M_B STOCK_RETURN EQUITY(-1) E E-05 [ ] [ ] [ ] [ ] EQUIY(-2) E E-05 [ ] [ ] [ ] [ ] EQUITY(-3) E E-05 [ ] [ ] [ ] [ ] EQUITY(-4) E E-05 [ ] [ ] [ ] [ ] 9

11 DEBT(-1) E E-05 [ ] [ ] [ ] [ ] DEBT(-2) E E-05 [ ] [ ] [ ] [ ] DEBT(-3) E E-05 [ ] [ ] [ ] [ ] DEBT(-4) E E-06 [ ] [ ] [ ] [ ] M_B(-1) [ ] [ ] [ ] [ ] M_B(-2) [ ] [ ] [ ] [ ] M_B(-3) [ ] [ ] [ ] [ ] M_B(-4) [ ] [ ] [ ] [ ] STOCK_RETURN(-1) [ ] [ ] [ ] [ ] STOCK_RETURN(-2) [ ] [ ] [ ] [ ] STOCK_RETURN(-3) [ ] [ ] [ ] [ ] STOCK_RETURN(-4) [ ] [ ] [ ] [ ] C [ ] [ ] [ ] [ ] SIZE E E-06 [ ] [ ] [ ] [ ] S & P(-1) [ ] [ ] [ ] [ ] S & P(-2) [ ] [ ] [ ] [ ] S & P(-3) [ ] [ ] [ ] [ ] S & P(-4) [ ] [ ] [ ] [ ] INTEREST(-1)

12 [ ] [ ] [ ] [ ] INTEREST(-2) [ ] [ ] [ ] [ ] INTEREST(-3) [ ] [ ] [ ] [ ] INTEREST(-4) [ ] [ ] [ ] [ ] R-squared Adj. R-squared In Table 3, we can find ha he sock reurns in he previous wo quarers have posiive effecs on he equiy issues. This is consisen wih he marke iming heory. However, his consisency is only limied o wo quarers horizon. When we race back o hree quarers before, hen he coefficien could no be explained by marke iming. Wih regard o he effecs of S & P 500 index on equiy issues, hey also have posiive effecs and he posiive effecs are also confined o wo quarers. However, heir effecs are no significan. Anoher finding is ha he equiy issues have negaive effecs on he sock reurn in he following hree quarers. This could be observed from he las equaion in Table 3, where he sock reurn is he dependen variable. If we look a he effecs of he ineres raes on he issues of deb, we could find ha increasing ineres rae would lead o decrease of deb issues. This is consisen wih Li and Ooi (2001) ha REITs also ime he deb marke. 6. Shor-run Performance of REITs afer SEOs In his par, some hypoheses under he asymmeric informaion heory would be esed. Since he equiy issues are considered as signals o he marke, he price would drop on he announcemen day and he issue day (Asquih and Mullins (1986), Masulis and Korwar (1986), and Barclay and Lizenberger (1988)). Some companies acually would announce he equiy issues on he issue day, while for ohers here are several weeks beween hese wo daes. As he announcemen daes of he secondary equiy offerings 11

13 are unavailable in our daa, we would use he issue day as he focus of sudy. This is he firs hypohesis ha would be esed. Also, in order o reduce he price drop a he issuing day, managers would issue equiy when he marke is mos informed, such as afer quarerly earnings announcemen. If managers delay he equiy issues, wih ime goes by, hey have more asymmeric informaion, and he price drop a he issuing would be larger. We would es hese hypoheses separaely. Hypohesis 1: The sock prices drop a he equiy issuing day. Here we calculae he gross sock reurn and he abnormal sock reurn on he equiy issuing day. The abnormal reurn is he gross sock reurn minus he CRSP equally weighed reurn excluding dividend. From Table 4, we can find ha here is significan negaive reurn on he issuing day. Table 4: Even day reurn Average T-saisic N Gross reurn on he issuing day Abnormal reurn on he issuing day Hypohesis 2: Equiy issues would cluser afer earnings releases. This hypohesis is based on he ime-varying asymmeric informaion in equiy issues. Wih ime-varying asymmeric informaion, managers wan o issue equiy when he marke is mos informed of he qualiy of he firm and reduce he cos of adverse selecion. One of he mos imporan sources of informaion is he quarerly earnings announcemen. I is expeced o see ha equiy issues would cluser shorly afer earnings releases, and i is suppored by he resul in Korajczyk, Lucas and McDonald (1991). The lag beween he equiy issues and he laes quarerly earnings announcemen is calculaed for each secondary equiy issues of REITs. I is found ha, abou 50% of he equiy issues occur in 40 days afer he quarerly earnings announcemen, compared o 75% for general socks. The disribuion of he ime lags is shown in Figure 2. Figure 2: Disribuion of ime lags beween equiy issues and he laes quarerly earnings announcemen 12

14 Number of Issues ~6 7~13 14~20 21~27 28~34 35~41 42~48 49~55 56~62 63~69 70~76 77~83 84~90 >90 Days afer quarerly earnings announced In Figure 2, here is roughly a rend of decreasing number of issues afer he quarerly earnings announcemen. However, his rend is less obvious compared o general socks. For example, for general socks, he number of issues in more han 90 days is abou 20% of he number of issues in 7 days (Korajczyk, Lucas and McDonald (1991), while for REITs, i is abou 40%. For general socks, he larges number of issues happens abou 28 days afer earnings announcemen. While for REITs, i happens in 49~55 days, which is much laer han general socks. Hypohesis 3: The closer he issue follows an earning release, he smaller he price drop a he issue day. The closer he issue follows an earning release, he marke is more informed, and he price drop a he issue day is expeced o be smaller. A simple regression is run here o es his relaionship. The resul is shown in Table 5. Table 5: Regression Resul (wih saisic in parenhesis) Consan Gross reurn ( )*** Abnormal reurn ( ) Days beween announcemen and -7E E-05 13

15 equiy issues ( ) ( ) No. of observaion R-square Table 5 shows ha here is no significan relaionship beween he lag of days and he price drop a he issuing day. This is also differen from he resul of general socks. In his par, some hypoheses of he marke iming under he asymmeric informaion heory are esed. These resuls give only weak evidence of he exisence of he asymmeric informaion in he marke iming of REITs. 7. Long-run Performance of REITs afer SEOs Generally here are wo mehodologies when researchers calculae he long-run performance. Firs mehod, is he mehod used by Rier (1991), comparing he reurns wih ou-of-sample asses. The second mehod is he Fama and French (1993) hree facor model. Since mos of he REITs have offerings during he sudy period, he second mehod is applied o es for abnormal reurns. The momenum variable in Carhar (1997) would also be included in he regression equaion. Thus he model we applied here is he four-facor model. The porfolios are formed by he REITs ha have equiy offerings during he previous five years. The sudy period is 1980 January o 2004 March. The average monhly reurn of REITs ha have equiy issues in he pas five years is calculaed for each monh. Boh value-weighed and equally-weighed monhly reurns are calculaed. Then he abnormal monhly reurn for each porfolio is regressed agains he hree Fama and French (1993) equiy risk facors, Marke, SMB, and HML, Carhar s (1997) momenum variable, and he reurn on he NAREIT index minus he risk-free rae. Firs we use he above four facors in he regression, hen he NAREIT index is included. The NAREIT index reurn is added o mimic he risk in reurns specifically relaed o he real esae indusry. The regression equaion is REIT Rf = α + β 1 ( ) 2 ( ) + β SMB + β HML + β MOMENTUM + ε 3 4 NAREIT Rf 5 + β Marke Rf 14

16 Where, REIT : The value-weighed (equally-weighed) reurn in quarer of REITs wih equiy offering in he las 5 years; Rf : Risk-free rae in quarer ; NAREIT : Reurn of NAREIT equiy REIT index in quarer ; Marke : Value-weighed CRSP reurns in quarer ; SMB : The difference beween he reurns on small and big sock porfolios wih abou he same book o marke equiy in quarer ; HML : The difference beween he reurns on high book-marke firms and low bookmarke firms in quarer ; Momenum : The high momenum sock reurn minus low momenum sock reurn where momenum is measured based on pas one-year reurn. The regression resul is shown in Table 6. Table 6: Long-run performance calculaion (wih saisic in parenhesis) Inercep Equally Weighed Sample ( )*** Marke-Rf ( )*** SMB ( )*** HML ( )*** Momenum ( ) ( ) (2.659)*** ( ) ( ) -8.6E-05 ( ) NAREIT-Rf ( )*** (5) Value Weighed Sample ( ) ( )*** ( )*** ( )*** ( ) ( ) ( ) ( ) ( ) -9.5E-05 ( ) ( )*** Adjused R-square In Table 6, significan long-run underperformance is found only in equally weighed four facor regression. When he NAREIT is included in he regression, he explanaory power of he equaions are increased, boh in equally and value weighed reurn. If we compare equally weighed regression wih value weighed regression, we can find ha he 15

17 inercep in value weighed regression is larger han in equally weighed. The reason may be ha larger REITs have beer performance compared o small REITs. In he las column of Table 6, he inercep becomes posiive. Our resul is consisen wih Buimer, Hyland and Sanders (2005), in which sudy on long-run underperformance is found using REIT s IPO daa. Since boh posiive and negaive conceps exis in our resul, he argumen in Fama (1998) is valid, ha he resul of long-run underperformance is no robus. Thus he marke efficiency is no violaed. 8. Conclusion REITs have always been considered as a special invesmen vehicle because of is differen insiuional srucures from general socks, and are always excluded from general socks in empirical sudies. Besides he close relaionship wih he real esae properies, which makes REITs an indusry wih heavy dependence on capial, REITs also have special requiremens on dividend payou and gearings compared o general socks. For example, in US, REITs are required o payou a leas 90% of he disribuable dividends o heir invesors. A he same ime, equiy REITs have higher predicabiliy in heir fuure cash flow as he main source of capial is he ren, hus less asymmeric informaion is expeced in REITs. Also, mos of he asses of REITs are angible asses. Wih all hese anomalies, would he marke iming found in general socks also be valid for US REITs? This paper ries o answer he above quesions. Firs, wheher marke iming exiss for REITs is esed. Using a VAR model, i is found ha REITs secondary equiy offerings always occur when he sock prices are high. This means ha marke iming behavior exis for REITs in SEOs. For he shor-run performance of REITs around SEOs, we find parial suppor o he marke iming heory under he framework of asymmeric informaion. The sock price decreases on he equiy issuing day, bu he iming of he equiy issues is less obvious compared o general socks. There is no relaionship beween he issue day price drop and he iming of he issues. Wih regard o he long-run performance sudy of REITs afer SEOs, our resul shows no long-run underperformance of REITs afer SEOs. All hese evidences sugges ha, alhough REITs also exhibi marke iming behavior, he iming paern is differen from general socks boh in he long-run and shor-run. 16

18 References Asquih, P., and Mullins, D. W., (1986), Equiy Issues and Offering Diluion, Journal of Financial Economics, 15: Baker, M., Sein J. C. and Wurgler, J., (2003), When Does he Marke Maer? Sock Prices and he Invesmen of Equiy-dependen Firms, The Quarerly Journal of Economics Baker, M., Taliaferro, R. and Wurgler, J., (2004), Pseudo Marke Timing and Predicive Regressions, Working Paper Baker, M., and Wurgler, J., (2000), he Equiy Share in New Issues and Aggregae Sock Reurns, Journal of Finance, 55: Baker, M., and Wurgler, J., (2002), Marke Timing and Capial Srucure, Journal of Finance, 57:1-32 Barclay, M. J., and Lizenberger R. H., (1988), Announcemen Effecs of New Equiy Issues and he Use of Inraday Price Daa, Journal of Financial Economics, 21:71-99 Brav, A., Geczy C., and Gompers, P. A., (2000), Is he Abnormal Reurn Following Equiy Issuances Anomalous? Journal of Financial Economics, 56: Buler, A. W., Gusavo, G., and Weson, J. P., (2005), Can Managers Forecas Aggregae Marke Reurns? Journal of Finance, 60: Buimer, R. J., Hyland, D. C., and Sanders, A. B., (2005), REITs, IPO Waves, and Long Run Performance, Journal of Real Esae Economics, 33(1):51-87 Carhar, M. M., (1997), On Persisence in Muual Fund Performance, Journal of Finance, 52: Chan, K., Ikenberry, D. and Lee I., (2004), Do Managers Time he Marke?Evidence from Open-marke Share Repurchases. Working Paper. Desai, H., and Jain, Prem C., (1999), Firm Performance and Focus: Long-run Sock Marke Performance Following Spinoffs- The Case of Volunary Spinoffs, Journal of Financial Economics,54: Eckbo, B. E., Masulis, R.W. and Norli, O., (2000), Seasoned Equiy Offerings: Resoluion of he New Issues Puzzle, Journal of Financial Economics, 56: Fama, E. F., (1998), Marke Efficiency, Long-erm Reurns, and Behavioral Finance, Journal of Financial Economics, 49:

19 Fama, E. F., and French, K. R., (1988), Dividend Yields and Expeced Sock Reurns, Journal of Financial Economics, 22:3-25 Fama, E. F., and French, K. R., (1993), Common Risk Facors in he Reurns on Socks and Bonds, Journal of Financial Economics, 33: 3-56 Fama, E. F., and French, K. R., (2005), Financing Decisions: Who Issue Sock? Journal of Financial Economics, 76: Frank, M. Z., and Goyal, V. K., (2003), The Effec of Marke Condiions on Capial Srucure Adjusmen, working paper Graham, J. R. and Harvey C. R., (2001), he Theory and Pracice of Corporae Finance: Evidence from he Field, Journal of Financial Economics, 60, Howon, S. D., Howon S. W., and Friday, H. S., (2000), Long Run Underperformance in REITs Following Seasoned Equiy Offerings, Journal of Real Esae Porfolio Managemen, 6(4): Ikenberyy, D., Rankine, G. And Sice, E., (1996), Wha do Sock Splis Really Signal? Journal of Financial and Quaniaive Analysis, 31: Korajczyk, R. A., Lucas, D. J. and McDonald, R. L., (1991), The Effec of Informaion Releases on he Pricing and Timing of Equiy Issues, The Review of Financial Sudies, 4(4): Korajczyk, R. A., Lucas, D. J., and McDonald, R. L., (1992), Equiy Issues wih Time Varying Asymmeric Informaion, Journal of Financial and Quaniaive Analysis, 27: Kohari, S. P., and Shanken, J., (1997), Book-o-marke, Dividend Yield, and Expeced Marke Reurns: A Time-series Analysis, Journal of Financial Economics, 44: Li, L., and Ooi, J., (2004), Financing Decisions of U.S. REITs: A Capial Marke Perspecive, ARES meeing 2004 Ling, D., and Ryngaer, M., (1997), Valuaion Uncerainy, Insiuional Involvemen, and he underpricing of IPOs: The Case of REITs. Journal of Financial Economics 43: Loughran, T., and Rier, J. R., (1995), The New Issues Puzzle, Journal of Finance, 50:

20 Lucas, D. J., and McDonald, R. L., (1990), Equiy Issues and Sock Price Dynamics, Journal of Finance, 45: Masulis, R.W., and Korwar, A.N., (1986), Seasoned Equiy Offerings: An Empirical Invesigaion, Journal of Financial Economics, 14: Myers, S. C., and Majluf, N. S., (1984), Corporae Financing and Invesmen Decisions When Firms Have Informaion ha Invesors Do No Have, Journal of Financial Economics 13: Pagano, M., Panea, F., and Zingales, L., (1998), Why Do Companies Go Public? An Empirical Analysis, Journal of Finance, 53: Rier, J. R., (1991), The Long-run Performance of Iniial Public Offerings, Journal of Finance, 46: 3-27 Rier, J. R., (20030, Inroducion o Recen Developmen in Corporae Finance, Edward Elgar Publishers Schulz, P., (2003), Pseudo Marke Timing and he Long-Run Underperformance of IPOs, Journal of Finance, 2: Spiess, K., and John, A., (1999), The Long-run Performance of Sock Reurns Following Deb Offerings, Journal of Financial Economics, 54: Wang, K., Chan, S. H., and Gau, G. W., (1992), Iniial Public Offerings of Equiy Securiies: Anomalous Evidence Using REITs, Journal of Financial Economics, 31: Welch, I, (2003), Capial Srucure and Sock Reurns, Yale ICF Working Paper No

21 Appendix Table 1: Equiy Marke Capiilizaion Ousanding (Millions of dollars a year end) End of Year # of REITs Equiy Morgage Hybrid Marke Capializaion # of REITs Marke Capializaion # of REITs Marke Capializaion , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , ,

22 , , , , , , , , , , , , Source: NAREIT websie. Available a hp:// (Marke capializaion equals price of shares muliplied by shares ousanding.) 2. Hisorical Securiies Issuance by US REITs (Millions of dollars) Toal Financing IPO Secondary Equiy Deb Offering Period Number Capial Raised Number Capial Raised Number Capial Raised Number Capial Raised Source: NAREIT websie. Available a hp:// 21

23 3. Figure 1: NAREIT equiy REITs price index NAREIT equiy REITs price index Year 4. Figure 2: S & P 500 price index S & P 500 Price index Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q Year 22 Q Q Q Q Q Q Q Q Q Q Q Q Q Q Q4 2004