The Information Content in Trades of Inactive Nasdaq Stocks
|
|
- Brett West
- 6 years ago
- Views:
Transcription
1 The Journal of Entrepreneural Fnance Volume 8 Issue 2 Summer 2003 Artcle 4 December 2003 The Informaton Content n Trades of Inactve Nasdaq Stocks Peter Chen Youngstown State Unversty Kasng Man Syracuse Unversty Chunch Wu Syracuse Unversty Follow ths and addtonal works at: Recommended Ctaton Chen, Peter; Man, Kasng; and Wu, Chunch (2003) "The Informaton Content n Trades of Inactve Nasdaq Stocks," Journal of Entrepreneural Fnance and Busness Ventures: Vol. 8: Iss. 2, pp Avalable at: Ths Artcle s brought to you for free and open access by the Grazado School of Busness and Management at Pepperdne Dgtal Commons. It has been accepted for ncluson n The Journal of Entrepreneural Fnance by an authorzed admnstrator of Pepperdne Dgtal Commons. For more nformaton, please contact paul.stens@pepperdne.edu.
2 The Informaton Content n Trades of Inactve Nasdaq Stocks Peter Chen + Youngstown State Unversty, Kasng Man ++ Syracuse Unversty and Chunch Wu +++ Syracuse Unversty In ths paper we analyze the frequency and nformaton content of small Nasdaq stock trades and ther mpacts on return volatlty at the ntraday nterval. We employ an autoregressve condtonal duraton (ACD) model to estmate the ntensty of the arrval and nformaton content of trades by accountng for the determnstc nature of ntraday perodcty and rregular tradng ntervals n transacton data. We estmate and compare the prce duraton of thnly and heavly traded stocks to assess the dfferental nformaton content of stock trades. We fnd that the number of transactons s negatvely correlated wth prce duraton or postvely correlated wth return volatlty. The mpact of the number of transactons on prce duraton or volatlty s hgher for thnly traded stocks. On the other hand, the persstence of the mpact on prce duraton adjusted for ntradaly perodcty s about the same for thnly and heavly traded stocks on average. + Peter Huayu Chen s an Assstant Professor of Fnance at Youngstown State Unversty. Hs current research nterests nclude fxed ncome nvestments, market mcrostructure and credt rsk analyss. ++ Kasng Man s an Assstant Professor n Manageral Statstcs at the Martn J. Whtman School of Management, Syracuse Unversty. Hs research nterests nclude tme seres analyss and forecastng methods. +++ Chunch Wu s a Professor of Fnance at the Martn J. Whtman School of Management, Syracuse Unversty. Hs current research nterests nclude fxed ncome nvestments, market mcrostructure and entrepreneural fnance.
3 26 Introducton The subject of prce formaton has always been ntrgung to fnancal researchers. A vast fnancal lterature has been devoted to the study of the pattern of nformaton arrvals and how new nformaton s ncorporated nto prce. These studes range from smple event studes of the market response to news announcements, to more sophstcated nformaton flow studes analyzng how nformaton nnovatons are mpounded nto securty prces. Interest n ths ssue has been fueled by recent advances n market mcrostructure theory and the avalablty of ultra-hgh-frequency data, thanks to modern technology. The shft to hgh-frequency data analyss has posed sgnfcant challenges to emprcal studes. A major dffculty faced n hgh-frequency data studes s that transactons arrve n rregular tme ntervals. Most emprcal mcrostructure studes have employed data wth a fxed tme nterval (e.g., hourly or half-hourly) to test the mplcatons of market mcrostructure theory (see, for example, Foster and Vswanathan, 1995; Andersen and Bollerslev, 1997). Ths s because standard tme-seres econometrc technques buld on the premse of fxed tme ntervals. The selecton of the tme nterval s often arbtrary. Large heavly-traded stocks typcally have transactons every few seconds whereas small thnlytraded stocks may not have transactons every hour or day. If a short tme nterval s chosen, there wll be many ntervals wth no transactons for thnly traded stocks and heteroskedastcty of a partcular form wll be ntroduced. On the other hand, f a long nterval s chosen, we may lose most of the mcrostructure features of the data. In partcular, when transactons are averaged, the tmng relaton and characterstcs of trades wll be lost. Emprcal mcrostructure studes that examne transacton-by-transacton data (see, for example, Hasbrouck, 1991; Madhavan et al., 1997; Huang and Stoll, 1997) face a dfferent estmaton problem. Data ponts n these studes correspond to the transacton (event) tme and so they are rregularly spaced. However, these studes have typcally gnored the problem of rregular ntervals when applyng standard tme-seres econometrc technques. Assumng that data ponts are equally spaced, when n fact they are uneven, leaves out much of the mportant nformaton about trade clusterng, temporal order flow patterns and the nformaton assmlaton process. Fortunately, new econometrc methods have been developed recently to cope wth the estmaton problems of rregularly spaced data (see Engle, 2000; Dufour and Engle, 2000). Two tme-seres methods were developed to model rregularly spaced data: Tme Deformaton models (TD) and Autoregressve Condtonal Duraton (ACD) models. The TD approach uses auxlary transformatons to relate observatonal or economc tme to calendar tme. In contrast, the ACD approach drectly models the tme duraton between events (e.g., trades). The ACD model typcally adopts a dependent pont process sutable for modelng characterstcs of duraton seres such as clusterng and overdsperson. In ths paper, we employ the ACD model proposed by Engle and Russell (1998) to examne nformaton clusterng and tradng responses to nformaton at the ntraday level. There are advantages of usng ths model. Frst, ths model provdes a framework for measurng and estmatng the ntensty of transacton arrvals that s partcularly suted for the tradng process. The model accounts for the rregular tme nterval, typcally encountered n stock tradng, by treatng t as a random varable that follows a pont process. Ths treatment resolves nfrequent or nonsynchronous tradng problems n emprcal estmaton usng ntraday data of thnly traded
4 27 stocks. Second, the estmaton procedure s relatvely straghtforward and the model can be easly adapted to test varous mcrostructural hypotheses. The prmary objectve of ths paper s to examne the patterns of nformaton arrval of small thnly-traded versus large heavly-traded stocks, and ther mpact on prce movements. The tradng pattern and the ntensty of nformaton-based tradng of small thnly-traded stocks often devates sharply from that of large heavly-traded stocks. Asde from the sheer dfference n tradng frequency, large stocks enjoy much hgher lqudty than small stocks. Studes have shown that a good proporton of trades for large heavly-traded stocks s for lqudty purposes (see Easley et al., 1996). Thus, trades for large heavly-traded stocks may not always have hgh nformaton content. Conversely, small stocks are not as lqud and are not traded as heavly as large stocks. Fewer analysts are nterested n these stocks and so less nformaton s avalable for nvestors. Due to lack of nformaton and lqudty, there are typcally no transactons for a good porton of the open market tradng perod. However, tradng of these stocks often causes a sgnfcant prce movement. Thnly traded stocks also have hgher varatons n order flow. Trades are clustered n that the occurrence of a trade nduces another trade n a rather short tme nterval. Once these stocks are traded, there s a hgh probablty that nformed traders may trade to mnmze prce mpacts (see Admat and Pflederer, 1988). As nsders prvate nformaton s mpounded nto prce, return volatlty ncreases. Snce the number of trades s low for small nactve stocks, the nformaton content per trade may be hgher for these stocks. In ths paper we focus on the ntensty of transacton arrvals and ts effects on prce movements of thnly-traded stocks n the Nasdaq market. Prevous studes have shown that the prce dscovery process and bd-ask spread behavor of a dealer market such as Nasdaq dffer from those of a aucton market lke the NYSE (see Hasbrouck, 1995; Huang and Stoll, 1996). Dfferences n market structures and tradng mechansms cause varatons n tradng costs, order flows and the speed of nformaton transmsson. Therefore, emprcal fndngs of NYSE stocks do not necessarly characterze Nasdaq stocks. One dstnct feature of Nasdaq s that the depth of the market often vares wdely among stocks. Ths greater dsperson n tradng actvtes provdes an excellent opportunty for comparng the ntensty of trade arrvals and the extent to whch trades convey nformaton for heavly- and thnly-traded stocks. Most emprcal mcrostructure studes have not accounted for the uneven ntervals n stock trades n examnng the ssues of order flow and nformaton assmlaton. An excepton s Dufour and Engle (2000). However, ther study covers only the most actvely traded stocks at the NYSE. Unlke ther study, we examne both the actve and nactve stocks on Nasdaq. The remander of ths paper s organzed as follows. Secton I presents the emprcal model and methodology for estmatng the ntensty of trade arrvals and the effects of mcrostructure varables on the tme duraton of trades and prce changes. Secton II dscusses data and emprcal results. Fnally, Secton III summarzes the man fndngs of ths paper. I. The Model Informaton arrvals nduce trades and prce changes (see Admat and Pflederer, 1988; Easley and O Hara, 1992). To analyze nformaton flow at rregular arrval tmes, we employ the autoregressve condtonal duraton (ACD) model proposed by Engle and Russell (1998).
5 28 Denote the nterval between two arrval tmes, x = t - t -1, as duraton. The expectaton of the th duraton condtonal on past x s s gven by, where E x x, x,..., x ) ( x, x,..., x ; ) (1) ( where s the vector of the parameters of the duraton process. Assumng that the stochastc process of the duraton s x (2) where s an..d. error term wth a dstrbuton whch must be specfed. Followng Engle and Russell (1998), we specfy the condtonal duraton by a general model: m j 1 q x, (3) j j j1 j j whch follows an ACD (m, q) process wth m and q referrng to the orders of the lags, and, j, ), j = 1, 2,, m and k = 1, 2,, q, are parameters to be estmated. Ths model ( k has a close connecton wth GARCH models and shares many of ther propertes. The model s convenent because t can be easly estmated usng a standard GARCH program by employng the square root of x as the dependent varable and settng the mean to zero (see Engle and Russell, 1998). In general, f duratons are condtonally exponental, the condtonal ntensty s (t x (4) N (t),...,x1) 1 N(t) 1 It can be shown that the hgher the condtonal ntensty, the hgher the volatlty of returns. There are several ways to estmate the system of (2) and (3). The smplest way s to assume that the error term follows an exponental dstrbuton and the lagged orders equal to one. Ths model s called the EACD(1,1) where E stands for the exponental dstrbuton. Another way s to assume that the condtonal dstrbuton s Webull, whch s equvalent to assumng that x s exponental where s the Webull parameter. Smlarly, we can estmate the Webull model wth the lagged orders equal to one, that s, WACD(1,1). The Webull dstrbuton functon can be wrtten as F( x ) = ( / ) x -1 exp[-( x / ) ] for, > 0 (5) When = 1, x / follows an exponental dstrbuton. The Webull dstrbuton s preferred f the data show an overdsperson wth extreme values (very short or long duratons) more lkely than the exponental dstrbuton would predct (see Dufour and Engle, 2000). Gven the
6 29 condtonal densty functon, we can estmate the parameters of the ACD model by maxmzng the followng log lkelhood functon: T L() = ln( / x ) + ln[ ( 11/ ) x / ] [ ( 11/ ) x / ] (6) 1 where (.) s the gamma functon, and s a column vector contanng the parameters to be estmated. Engle and Russell (1998) commented that clever optmzaton can avod repeated evaluaton of the gamma functon. Ths tactcs s useful when the sample sze s very large. The ACD model s essentally a model for ntertemporally correlated transacton (event) arrval tmes. The arrval tmes are treated as random varables followng a pont process. In the context of securty tradng, assocated wth each arrval tme are random varables such as volume, prce or bd-ask spread. These varables are defned as marks. Fnance researchers are often nterested n modelng these marks assocated wth the arrval tmes. For example, not all transactons occur because of the arrval of new nformaton. Instead, some are trggered by pure lqudty or portfolo adjustment reasons, whch are not related to changes n the expected (fundamental) value of stock. On the other hand, there are tmes when transactons occur as a result of new nformaton arrval that s not publcly observable. Market mcrostructure theory suggests that traders possessng prvate nformaton wll trade as long as ther nformaton has value. Ths results n clusterng of transactons followng an nformaton event. To examne ths hypothess, we can defne the events as a subset of the transacton arrval tmes wth specfc marks. For example, to examne the effect of nformaton events, we can select data ponts for whch prce has moved beyond the bd-ask bound. Ths process s called dependent thnnng. To dstngush nformed from unnformed trades, we modfy transacton arrval tmes nto prce arrval tmes. The basc dea s to leave out those transactons that do not sgnfcantly alter prce. The prce movements can be classfed ether as transtory or permanent movements. Defne the mdpont of the bd-ask spread or mdprce to be the current prce. Followng Engle and Russell (1998), we defne a permanent prce movement as any movement n the mdprce (mdquote) greater than or equal to $0.25 or 2 tcks. 1 Once we defne the prce arrval tmes, we can apply the ACD model to these new event arrval tmes. In ths case, we are modelng how quckly the prce s changng rather than the arrval rate of transactons. The ntensty functon s now called prce ntensty, whch measures the nstantaneous probablty of a permanent prce change. The basc formulaton of the ACD model parameterzes the condtonal ntensty of event arrvals as a functon of the tme between past events. It can be easly extended to nclude other effects such as characterstcs assocated wth past transactons or other outsde nfluences. For example, prevous studes have shown that mportant nformaton s contaned n the number of trades, and the trade sze whch s the average volume per transacton. To examne ths hypothess, we can modfy the ACD model to nclude these two varables: 1 A tck s 1/8 dollar.
7 30 p j1 j q x + #Trans + Volume/Trans (7) j j1 j j where the duraton s now between two consecutve prces wth a movement greater than or equal to two tcks, and the number of transactons per duraton and trade sze per transacton are added as determnants of duraton. Market mcrostructure theory contends that trades contan nformaton that affects prce movements (or volatlty). Includng the number of transactons and trade sze allows us to test ths mportant hypothess. In addton, dvdng the accumulated volume by the number of transactons yelds the average volume per transacton (or trade sze) at the nterval x. Prevous studes have ndcated that trade sze may contan nformaton. The ACD model n (7) now descrbes how quckly the prce changes, by takng nto consderaton the effects of transacton rate and trade sze. The ntensty functon becomes a measurement of the nstantaneous probablty of a prce movement called prce ntensty. It can be shown that prce duraton s nversely related to the volatlty of prce changes. In addton to transacton frequency and trade sze, we also test the ACD model wth the bd-ask spread varable. Mcrostructure theory suggests that the specalst s (or dealer s) bd-ask spread reflects the ntensty of nformed tradng. It wll be nterestng to see whether ths varable wll ncrease the explanatory power of the model. Thus, we also estmate the followng extended model: p j1 j q x + #Trans + Volume/Trans + Spread (8) j j1 j j where Spread s the bd-ask spread dvded by md-quote. It s wdely known that ntraday return volatlty exhbts sgnfcant determnstc (perodc) patterns. Snce prce duraton s the nverse of volatlty, the duraton measure s expected to contan a determnstc component. Ths determnstc component needs to be separated from the stochastc component n emprcal estmaton. The strategy followed here to elmnate the ntraday pattern s a smple seasonal adjustment approach. The tme span wthn a tradng day s dvded nto non-overlappng tme ntervals of 15 mnutes each. The mean of prce duratons wthn each nterval s computed over the entre sample perod. The adjusted prce duraton s then computed as the prce duraton dvded by the average prce duraton wthn that nterval. The adjusted prce duraton seres now has a mean approxmately equal to one. If the adjusted duraton s greater (less) than one, the duraton s greater (less) than the average duraton n that tme nterval. We estmate the ACD model usng these adjusted prce duratons, as well as the raw (unadjusted) duratons. 2 II. Data and Emprcal Estmaton Data on prce, sze, and tradng tme for Nasdaq stocks are obtaned from the TAQ database over the perod of July 1 to September 30, Trades and quotes are selected 2 We have also tred the splne method to flter the determnstc ntraday components. The results usng ths method are qute smlar.
8 31 strctly for Nasdaq-lsted frms, thus excludng NYSE stocks traded on Nasdaq and stocks lsted on regonal exchanges. We also exclude all preferred stocks, stock funds, stock rghts, warrants and ADRs. 3 Prevous studes (see, for example, Easley et al., 1996; Wu and Xu, 2000; Wu, 2003) have used tradng volume as a measure for defnng the actveness of stocks. Followng the nfluental paper by Easley et al. (1996), we use tradng volume to classfy the actveness of stocks for the purpose of comparng wth ther results. Tradng volume s a preferred measure for ths classfcaton because t contans the nformaton of frequency and sze of trades, both of whch are mportant ndcators of the actveness or depth of stocks. We rank all Nasdaq common stocks by the average daly tradng volume over the sample perod, and then dvde the sample nto volume decles. The frst volume decle ncludes the hghest-volume stocks and the tenth decle contans the lowest-volume stocks. To nsure enough tradng actvtes for purposes of emprcal estmaton, we choose stocks from the frst, ffth and eghth volume decles. To control for the prce effect, we construct a matched sample of stocks havng transacton prces close to each other at the begnnng of the sample perod (July 1), but at dfferent levels of tradng volume. Stocks from the three selected decles are ranked n order of ntal prce and adjacent trplets of stocks are matched. We randomly choose fve matched stocks from each of the three volume decles to perform emprcal estmaton. We choose only fve stocks from each volume decle for emprcal estmaton to allevate the computaton burden. Transacton duraton can be easly computed as the tme dfference between consecutve trades. Consecutve trades wth same tme stamp and prce are aggregated and treated as one trade. We then thn the transacton data by constructng prce duraton wth prce changes greater than or equal to two tcks. Volume s expressed n terms of the number of shares traded at each nterval. Table 1A shows the summary statstcs after dependent thnnng where any mdquote movements less than two tcks are gnored. More heavly traded stocks have lower spreads, more transactons (or shorter trade duratons) and hgher volume. Note that the daly number of transactons (or tradng frequency) n the hgh-volume group s hgher than those n the medum- and low-volume groups for all stocks except DURA. Smlarly, the daly number of transactons (or tradng frequency) n the medum-volume group s hgher than that of lowvolume group for all stocks except PSUN. In the analyss to follow, we compute the parameter estmates wth and wthout these two stocks. The averages wthout these two stocks represent the average parameter estmates of hgh and medum tradng frequency groups. After the data are thnned by prce, the prce duraton stll tends to be lower for more actvely traded stocks. On the other hand, trade sze or the average volume per transacton s about the same for both actve and nactve stock groups. Table 1B lsts the names of sample stocks. Fgure 1 shows the average prce duraton throughout a typcal tradng day for three selected stocks. The vertcal axs ndcates the prce duraton n seconds, and the horzontal axs ndcates the ntraday ntervals. We dvde each tradng day nto 25 ntervals of 15 3 To avod the problem at the market open (e.g., stale quotes, and delay of the open), data for the frst ffteen mnutes are dropped as suggested by Mller et al. (1994). Ths avods serous stale quote problems, especally for thnly traded stocks.
9 32 mnutes each. The average prce duraton wthn each nterval s computed over the entre sample perod. As shown, prce duraton exhbts an nverted U-shape pattern. Ths s not a surprse snce prce duraton s an nverse of prce volatlty, and ntraday prce volatlty exhbts a pronounced U shape. Prce duraton s negatvely related to tradng frequency or number of transactons. As ndcated, the prce duraton (n seconds) of ASND s much shorter than that of WIND because the former has a much greater number of daly transactons (see Table 1A). 2.1 Model Estmaton Usng Unadjusted Data We frst estmate the baselne ACD models wth no mcrostructure varables. We use the Polak-Rbere Conjugate Gradent (PRCG) to obtan the MLE estmates of the ACD parameters. The model s frst estmated usng the unadjusted prce duraton and then the adjusted duraton. The adjusted duraton s the prce duraton adjusted for the ntraday determnstc pattern. Table 2 reports the emprcal estmates for the EACD(1,1) model usng unadjusted data. As shown, most parameter estmates are statstcally sgnfcant. The ARCH and GARCH parameters, and, are postve n most cases, consstent wth the predcton and ther values fall n the theoretcal range. The results ndcate that a short prce duraton s lkely to be followed by another short prce duraton. Or equvalently, hgh prce volatlty n the current tradng nterval s lkely to brng hgh prce volatlty at the next tradng nterval. The sum of and represents the persstence of prce duraton. The results do not show a materal dfference n persstence for hgh and low tradng volume groups. Table 3 reports the estmates of the WACD(1,1) model. Agan, the estmates of and are postve n most cases and most of them are sgnfcant. The Webull parameter s hghly sgnfcant. The values of the Webull parameter are all less than one and tend to be smaller for less heavly traded stocks. The results suggest that the EACD model s not sutable because the error term does not follow exactly the exponental dstrbuton. The persstence of prce duraton s measured by the sum of and. Ignorng CBSS, the result agan does not show a materal dfference n persstence for hgh and low tradng volume groups. 4 We next test the mplcatons of market mcrostructure theores. On theoretcal grounds, Easley and O Hara (1992) predct that the number of transactons would nfluence the prce process through the nformaton-based clusterng of transactons. Admat and Pflederer (1988, 1989) predct that the number of transactons wll have no mpact on prce ntensty. Glosten and Mlgrom (1985) and Kyle (1985) predct that volume tends to be hgher as the probablty of nformed tradng ncreases. Most emprcal studes have documented a postve relatonshp between volatlty and volume for both ndvdual securtes and portfolos. Schwert (1989) and Gallant, Ross, and Tauchen (1992) fnd a postve correlaton between volatlty and tradng volume. Jones, Kaul, and Lpson (1994) show that the postve volatlty-volume relatonshp actually reflects the postve relatonshp between volatlty and the number of transactons. Based on ths fndng, they conclude that trade sze carres no nformaton beyond that contaned n the frequency of transactons. None 4 Note that although the estmate of for CBSS s sgnfcantly negatve, ths estmate mproves when adjusted prce duraton s used as dependent varable as shown n Table 6 below.
10 33 of these studes has addressed the ssue of uneven tradng ntervals or nfrequent tradng. In the followng, we re-examne ths ssue usng the ACD model at the ntraday level. We estmate the ACD model wth two addtonal explanatory varables: the number of transactons per duraton and average trade sze. Table 4 reports the results of estmaton. The coeffcents of the number of transactons are mostly negatve. The results suggest that the expected prce duraton tends to be shorter, or equvalently the volatlty s hgher, followng an nterval of hgh transacton rates. Ths relatonshp s much stronger for less-heavly traded stocks. Ths concluson holds regardless of whether DURA and PSUN are ncluded or not. On the other hand, the effect of trade sze s less conclusve. The sgn of the coeffcents of average volume per transacton (or trade sze) s negatve for more-heavly traded stocks but postve for less-heavly traded stocks. Table 5 reports the estmates of the ACD model when the bd-ask spread s added as an addtonal explanatory varable. The coeffcents of the number of transactons contnue to be qute sgnfcant wth a predcted negatve sgn. The coeffcents of trade sze agan have mxed sgns. The coeffcents of spreads are generally negatve, suggestng that hgher spreads generally lead to shorter prce duraton (or hgher volatlty). Excludng PSUN n the mddle group does not change the concluson Model Estmaton Usng Adjusted Data We next turn to the estmaton of the ACD model usng the adjusted data where prce duraton s adjusted for the ntradaly perodcty. Table 6 reports the estmates of the baselne WACD(1,1) model where no mcrostructure varables are ncluded. As shown, after removng the ntraday determnstc effect to retan the stochastc component of prce duraton, parameter estmates of the WACD(1,1) model become much more stable. The parameters and are now all wthn the theoretcal range wth a sum less than one. The results suggest that t s necessary to account for the ntraday perodc pattern n emprcal estmaton. Agan, the results show lttle dfference n the persstence of prce duraton between the hgh- and low-volume stocks. On average, the sum of and s qute close for the three groups. Table 7 reports the results of the WACD model wth mcrostructure varables. The coeffcents of the number of transactons are all negatve, ndcatng that the hgher the number of transactons, the shorter the prce duraton. The sze of the coeffcents (n absolute value) s much larger for less-heavly traded stocks. The average value of the coeffcent for the number of transactons s 0.51 for the lowest-volume group compared to for the hghest-volume group. Excludng DURA and PSUN does not affect the results materally ( for the hgh volume group). Thus, the mpact of the number of transactons (#Trans) s hgher not only for low volume group but also for low trade frequency group. Another nterestng fndng s that the coeffcents of trade sze have mxed sgns. The sgn tends to be negatve for most-heavly traded stocks. As the tradng volume decreases, the sgn becomes postve. Thus, trade sze does not necessarly decrease prce duraton (or ncrease prce volatlty). 5 Note that DURA n the frst group does not converge. Therefore, the results for the hgh-volume group also represent the results for stocks wth hgh-trade frequency.
11 34 Table 8 reports the results when bd-ask spread s added as an addtonal varable. Results show that the effect of spread s much hgher for less-heavly traded stocks. The coeffcents of spread are mostly negatve and sgnfcant for thnly traded stocks. On average, the absolute value of the spread coeffcent for the thnly traded group (-0.77) s much hgher than that of the heavly traded group, whch s close to zero. Furthermore, the absolute value of the coeffcent of the number of transactons s agan much hgher for the thnly traded group. The average value of ths coeffcent s for the thnly traded group compared to for the heavly traded group. Excludng DURA and PSUN does not change the results materally (-0.06 for the heavly traded group). Moreover, the coeffcent of trade sze (average volume per transacton) s postve, whch contrasts wth the negatve sgn of the trade sze coeffcent for the heavly traded group. Thus, the prce ntensty appears to be qute dfferent between actve and nactve stocks. The response of prce ntensty to the arrval of nformaton, as captured by the spread and transactons, s much stronger for the thnly traded stocks than for heavly traded stocks. The results support the contenton that trades of thnly traded stocks have a larger mpact on ther duraton (or volatlty); that s, a trade of thnly traded stock tends to trgger another trade much faster. Ths result also suggests a greater trade clusterng for thnly traded stocks. 2.3 One-Step-Ahead Forecast of Prce Duratons Fgure 2 shows the one-step-ahead forecast of prce duratons n a tradng day for three selected stocks ASND (July 2), SEBL (July 18) and WIND (July 14), respectvely. These dates are chosen such that they are among the days wth more transactons after the thnnng process. The one-step-ahead forecasts are obtaned as follows. Consder the ACD(1,1) model for duraton n (3). Defne x, whch s a martngale dfference sequence wth mean 0. Rewrte the ACD (1,1) equaton n (3) as x ( ) x 1 1 (3a) The prce duraton x therefore has an ARMA(1,1) representaton, and ts forecast can be obtaned usng the usual ARMA method. We employ the adjusted duraton and compute ts one-step-ahead forecast, usng the WACD(1,1) estmates n Table 6. These forecasts are shown n the left-hand panel of Fgure 2 for three selected stocks. They are then multpled back by the ntraday perodc pattern (.e., average duraton wth a typcal shape as shown n Fgure 1) accordng to the tme of day the transacton occurs. The resultng forecast of the unadjusted duraton s shown n the rghthand panel of Fgure 2. The horzontal axs ndcates the sequence of transactons for ths partcular tradng day and the vertcal axs ndcates the duraton. Fgure 2 shows that actual duratons are n general subject to hgher fluctuaton than the forecasted duratons. As shown, the one-step-ahead forecast obtaned by multplyng the adjusted duraton forecast and the ntraday perodc component performs reasonably well for the less-heavly traded stocks. The forecasts for stock ASND, whch s n the heavly traded group, s comparatvely more stable at the level of around 1000 (seconds). Incorporatng the
12 35 determnstc ntraday perodc component nto the forecastng enhances the forecastng precson. The results suggest that one needs to consder the ntraday determnstc components to provde a good forecast for prce duraton. Engle and Russell (1997) establsh a relatonshp between prce duratons and volatlty. In bref, they assume the underlyng prce process s a bnomal process wth ncrements of c whch takes expected tme. They show (see eq. (22) of ther paper) that the expected varance per unt of tme s nversely related to expected duraton; n partcular, ˆ 2 2 c /. Usng ths relatonshp, we can transform our estmates of prce duraton to volatlty. 2.4 Impulse Response Functon We next employ the concept of mpulse response functon to examne the mpact of nformaton shock on prce duraton. Consder the ACD(1,1) model wth a mcrostructure varable z, x z 1 1 (9) Let x, and rewrte the ACD equaton (9) as x ( ) x 1 1 z Denote /( 1 ) as the mean, and B the back-shft operator such that By y 1. It follows that x 1 B z 1 ( ) B 1 ( ) B Let be the sum of the ARCH and GARCH parameters. The expanson 2 B B 1B 2... k B k... s useful for studyng the lastng effect of the mcrostructure varable z on duraton. k Furthermore, the lagged k term s, whch measures the mpact of one unt ncrease n the mcrostructure varable z on prce duraton x (mean adjusted) k-lags (or k-transactons) later. We refer to ths term as the mpulse response functon at lag k. It s clear that for less than 1 n magntude, the mpact goes down to zero eventually. But the mpact decreases wth a slower rate or s more persstent, when s closer to 1. Also, we refer to
13 36 2 k... (1 k1 ) /(1 ) (10) as the k-cumulatve mpulse response functon of a unt ncrease n z on prce duraton x (mean adjusted) after k transactons. It goes to a lmtng value of /( 1 ). Note that the k (frst) dfference n the k-th and (k-1)th cumulatve mpulse response functon gves. Table 9 reports the k-cumulatve mpulse response functons of two mcrostructure varables, number of transactons and average volume per transacton, based on the parameter estmates for adjusted duraton n Table 7. 6 For each stock, the frst row reports the estmates of,,, +,, and. The second row reports the k-cumulatve mpulse response functon of one unt ncrease n #Trans (the number of transactons per duraton) for k = 0, 1,.., 9 lags later, and wth the lmtng value (when k s nfnte) gven as the last value. Smlarly, the thrd row s the cumulatve response functons that correspond to Volume/Trans (or trade sze). The results show that an ncrease of one unt n #Trans has a hgher (negatve) mpact on prce duraton for stocks that are less actve. On the other hand, the mpact of a unt ncrease n Volume/Trans on prce duraton s mxed and has no clear pattern across stock groups. The response functon for #Trans s most nterestng and s plotted n Fgure 3 for each stock group of hgh, medum and low tradng actvtes. The graph on the top left-hand corner shows the k-cumulatve mpulse response functon (k = 0, 1,, 99) for adjusted prce duraton after a unt ncrease n #Trans. These response functons are obtaned based on the average WACD(1,1) and estmates for each stock group as reported n Table 7. As shown, the effect of a unt ncrease n #Trans has a hgher (negatve) mpact on prce duraton for stocks that are less actve. k The two graphs on the bottom show the mpulse response functon. The graph on the lower rght-hand corner s smlar but wth the value set equal to 1 for all three groups. Ths settng ams at snglng out the effect caused by the sum of the ARCH and GARCH parameters by standardzng the mpact of #Trans. For llustraton, consder the thnly traded group n the lower left-hand corner of Fgure 3. If #Trans ncreases by one unt, after one transacton (k = 1), the change n adjusted duraton s -0.51*( )= (see Table 7), whch s the second pont on the mpulse response functon for ths group. We need to dvde ths number by 100 to obtan the actual amount of reducton because the unt of adjusted duraton was multpled by 100 n Table 7. In other words, the reducton n adjusted duraton due to a trade nnovaton s 0.44% of the average prce duraton (depends on the tme of day) for ths group after one transacton. From Table 1A, we can compute the average duraton for ths group, whch s 2,308 seconds. Therefore, the reducton amounts to about 10 seconds per #Trans. Snce the average #Trans s 19 (from Table 1A) for ths group, the overall reducton n prce duraton s about 190 seconds. Smlar calculaton shows the overall reducton s about 56 seconds after k = 10 transactons. From the lower left-hand corner of Fgure 3, we agan see that the effect of a unt ncrease n #Trans has a hgher (negatve) mpact for stocks that are less actve, whch eventually goes down to zero. After standardzng the mpact of number of trades by keepng 6 Results are qualtatvely the same f we add the spread varable.
14 37 the same (= -1), the graph on the lower rght-hand corner shows more clearly that the persstence of mpact on the adjusted prce duraton for the heavly and thnly traded stock groups s about the same. Thus, whle trades of thnly traded stocks have a larger mpact (or a greater nformaton effect), the duraton of the mpact s close to that of heavly traded stocks on average. III. Summary In ths paper, we examne the frequency of nformaton arrvals of small thnly-traded stocks and ts mpacts on prce duraton or return volatlty at the ntraday level. We employ the autoregressve condtonal duraton (ACD) model to estmate the ntensty of nformaton arrvals and nformaton content of trades. The unque feature of ths model s ts ablty to handle hgh-frequency transacton data recorded at rregular tme ntervals. We fnd that ntraday perodcty must be consdered n the transacton data analyss. Our results show that the data adjusted for the ntraday determnstc pattern produce much more stable parameter estmates. In addton, the accuracy of forecasts s enhanced when the ntraday pattern s accounted for n the one-step-ahead forecastng. Our results show that there are dfferences n transacton and prce duratons between heavly and thnly traded stocks. The mpact of the number of transactons on adjusted prce duraton s much larger for thnly traded stocks than for heavly traded stocks. On the other hand, the persstence of the mpact on adjusted prce duraton s about the same between heavly and thnly traded stocks on average. The results show that the number of transactons has hgher explanatory power than average trade sze. We also examne the mpact of spread on prce duraton. The results show a consstent sgnfcantly negatve (postve) relatonshp between spreads and prce duraton (volatlty) only for thnly traded stocks. In addton, the effect of spreads s much stronger for thnly traded stocks, suggestng a larger mpact of asymmetrc nformaton for these stocks. Overall, we fnd that the number of trades contans most of the relevant nformaton affectng prce duraton or volatlty, and the trades of thnly traded stocks have greater mpact on prce duraton. The results suggest that the trades of thnly traded stocks contan more prvate nformaton than the trades of heavly traded stocks.
15 38 REFERENCES Admat, Anat R. and Paul Pflederer, 1988, A theory of ntraday patterns: Volume and prce varablty, Revew of Fnancal Studes 1, Admat, Anat R. and Paul Pflederer, 1989, Dvde and conquer: A theory of ntraday and day-of-week mean effects, Revew of Fnancal Studes 2, Andersen, Torben G., and Tmothy Bollerslev, 1997, Intraday perodcty and volatlty persstence n fnancal markets, Journal of Emprcal Fnance 4, Dufour, Alfonso, and Robert F. Engle, 2000, Tme and the prce mpact of a trade, Journal of Fnance 55, Easley, Davd and Maureen O Hara, 1992, Tme and the process of securty prce adjustment, Journal of Fnance 19, Easley, Davd, Ncholas M. Kefer, Maureen O Hara and Joseph B. Paperman, 1996, Lqudty, nformaton, and nfrequently traded stocks, Journal of Fnance 51, Engle, Robert F., 2000, The economcs of ultra-hgh-frequency data, Econometrca 68, Engle, Robert F. and Jeffrey R. Russell, 1997, Forecastng the frequency of changes n quoted foregn exchange prces wth the autoregressve condtonal duraton model, Journal of Emprcal Fnance 4, Engle, Robert F. and Jeffrey R. Russell, 1998, Autoregressve condtonal duraton: A new model for rregularly spaced transacton data, Econometrca 66, Foster, F. Douglas, and S. Vswanathan, 1995, Can speculatve tradng explan the volumevolatlty relaton? Journal of Busness and Economc Statstcs 13, Gallant, A. Ronald, Peter E. Ross, and George E. Tauchen, 1992, Stock prces and volume, Revew of Fnancal Studes 5, Glosten Lawrence R., and P. Mlgrom, 1985, Bd ask and transacton prces n a specalst market wth heterogeneously nformed agents, Journal of Fnancal Economcs 14, Hasbrouck, Joseph, 1991, Measurng the nformaton content of stock trades, Journal of Fnance 66,
16 39 Hasbrouck, Joseph, 1995, One securty, many markets: Determnng the contrbutons to prce dscovery, Journal of Fnance 50, Huang, Roger D. and Hans R. Stoll, 1997, The components of the bd-ask spread: A general approach, Revew of Fnancal Studes 10, Jones, M. Charles, Gautam Kaul, and Marc Lpson, 1994, Transactons, volume and volatlty, Revew of Fnancal Studes 7, Kyle, Albert 1985, Contnuous tme auctons and nsder tradng, Econometrca 53, Madhavan, Ananth, Mathew Rchardson and Mark Roomans, 1997, Why do securty prces change? A transacton-level analyss of NYSE stocks, Revew of Fnancal Studes 10, Mller, Merton H., Jayaram Muthuswamy and Robert Whaley, 1994, Mean reverson of Standard & Poor s 500 ndex bass changes: Arbtrage-nduced or statstcal lluson?, Journal of Fnance 49, Schwert, G. Wllam, 1989, Why does stock market volatlty change over tme? Journal of Fnance 44, Wu, Chunch, 2003, Informaton flow, volatlty and spreads of nfrequently traded Nasdaq stocks, Quarterly Revew of Economcs and Fnance, forthcomng. Wu, Chunch and X. Eleanor Xu, 2000, Return volatlty, tradng mbalance and the nformaton content of volume, Revew of Quanttatve Fnance and Accountng 14,
17 40 Table 1A Summary Statstcs Stock Symbol No. of Duratons Ave. Prce Ave. Spread /Duraton Ave. # Tran /Duraton Ave. Daly # Trans Ave. Vol/Trans Ave. Prce Duraton Ave. Daly Volume (shares) ASND 1, , , , ,431,497 ORCL , , , ,131,497 NSCP , , , ,551,877 SBUX , , ,025 DURA , , ,185 CLST , , ,837 ADTN , , ,343 IRIDF , ,698 PSUN , , ,440 SEBL , , ,112 SDTI , , ,817 APOL , , ,208 LHSPF , , ,928 WIND , , ,510 CBSS , , ,073 Ths table provdes summary statstcs for stocks n three groups classfed based on tradng volume. The frst group s the hgh-volume or heavly traded group and the thrd group s the low-volume or thnly traded group. The medum volume group s n between these two groups. The frst group can be classfed as the frequently traded group, f DURA s excluded. The second group can be classfed as the medum-frequency group f PSUN s excluded whle the thrd group can be classfed as the nfrequently traded group. The data are thnned by gnorng prce movements less than two tcks ($0.25). Duraton s the tme nterval between two trades. The duraton calculated after thnnng s called prce duraton. Prce duraton s measured n seconds. Volume s measured n number of shares. The number of duratons s the number of observatons for the duraton varable; average prce and spread are expressed n dollars; average #Trans/Duraton s the number of transacton per duraton; average daly #Trans s the mean transacton number per day; and average Vol./Trans s the average volume (n shares), or trade sze per transacton.
18 41 Table 1B Company Names Hgh- Volume Group Medum- Volume Group Low-Volume Group ASND ORCL NSCP SBUX DURA CLST ADTN IRIDF PSUN SEBL SDTI APOL LHSPF WIND CBSS ASCEND COMMUNICATIONS ORACLE CORP NETSCAPE COMMUNICATIONS CORP STARBUCKS CORP DURA PHARMACEUTICALS, INC. CELLSTAR CORP ADTRAN INC IRIDIUM LLC PAC SUNWEAR CA SIEBLE SYSTEMS SECURITY DYNAMICS APOLLO GROUP LERNOUT & HAUSPIE SPEECH PRODUCTS WIND RIVER SYSTEMS COMPASS BNCSHRS
19 42 Table 2 Estmates of the EACD(1,1) model for unadjusted prce duraton Stock ASND 8.34(4.01) 0.26(5.13) 0.41(4.08) ORCL 7.63(3.54) 0.31(4.81) 0.46(4.64) NSCP 6.66(3.09) 0.20(4.29) 0.54(4.95) SBUX 7.45(1.79) -0.03(-0.93) 0.83(8.72) DURA 0.70(1.09) 0.01(0.84) 0.97(37.07) Average Average (wthout DURA) CLST 1.97(3.39) 0.37(5.64) 0.63(12.68) ADTN 8.34(4.01) 0.26(5.13) 0.41(4.08) IRIDF 3.10(4.44) 0.25(5.26) 0.67(15.65) PSUN 6.27(3.48) 0.16(3.89) 0.64(8.16) SEBL 3.80(2.31) 0.06(2.82) 0.79(11.11) Average Average (wthout PSUN) SDTI 14.78(1.38) 0.02(0.84) 0.51(1.52) APOL 5.56(2.30) 0.07(2.28) 0.76(8.32) LHSPF 2.63(2.95) 0.59(5.41) 0.46(7.66) WIND 7.81(5.82) 0.45(6.69) 0.35(5.52) CBSS Average Ths table reports the parameter estmates of the EACD(1,1) model and the t-values (n parentheses) for three stock groups descrbed n Table 1. The EACD(1,1) model for duraton s where x x 1 1 x s the duraton and s the condtonal mean of the duraton between two arrval tmes t and t -1. The prce duraton was dvded by 60 n estmaton. The parameter estmates for CBSS dd not converge. Average parameter estmates are mean estmates for each group. Averages wthout DURA or PSUN are mean parameter estmates excludng each of these two stocks whch are removed because ther tradng frequences are too low to be qualfed n the hgh and medum frequency groups. The mean estmates excludng these stocks represent the group average usng the measure of trade frequency to defne the actveness of stocks.
20 43 Table 3 Estmates of the WACD(1,1) model for unadjusted prce duraton Stock ASND 7.07(5.19) 0.40(7.58) 0.23(2.34) 0.91(47.99) ORCL 7.71(3.36) 0.30(3.88) 0.45(4.29) 0.94(33.37) NSCP 0.67(1.66) 0.06(3.28) 0.91(30.81) 0.81(36.10) SBUX 7.84(1.57) -0.03(-0.89) 0.83(7.49) 0.78(25.51) DURA 0.53(0.64) 0.01(0.47) 0.98(26.01) 0.71(23.58) Average Average (wthout DURA) CLST 1.79(2.51) 0.40(4.33) 0.61(9.13) 0.68(35.59) ADTN 8.62(2.72) 0.22(3.33) 0.42(2.81) 0.72(33.64) IRIDF 2.96(3.05) 0.29(4.02) 0.64(10.65) 0.68(31.04) PSUN 5.91(2.36) 0.21(2.71) 0.61(5.16) 0.57(29.19) SEBL 4.82(1.93) 0.12(2.38) 0.70(5.70) 0.61(37.17) Average Average (wthout PSUN) SDTI 16.46(1.09) 0.04(0.75) 0.45(0.96) 0.68(29.54) APOL 5.11(1.60) 0.07(1.58) 0.78(6.66) 0.65(29.48) LHSPF 2.53(2.30) 0.61(4.49) 0.44(5.82) 0.77(25.05) WIND 7.80(4.19) 0.46(4.86) 0.33(3.83) 0.67(34.95) CBSS 41.10(1.85) -0.24(-4.09) 0.60(2.13) 0.75(11.52) Average Ths table reports the parameter estmates of the WACD(1,1) model and the t-values (n parentheses) for three stock groups descrbed n Table 1. s the parameter of the Webull dstrbuton. Estmaton s based on the lkelhood functon n equaton (6). The WACD(1,1) model for duraton s where x x 1 1 x s the duraton and s the condtonal mean of the duraton between two arrval tmes t and t -1. The prce duraton was dvded by 60 n estmaton. Average parameter estmates are mean estmates for each group. Averages wthout DURA or PSUN are mean parameter estmates excludng each of these two stocks whch are removed because ther tradng frequences are too low to be qualfed n the hgh and medum frequency groups. The mean estmates excludng these stocks represent the group average usng the measure of trade frequency to defne the actveness of stocks.
21 44 Table 4 Estmates of the WACD(1,1) model for unadjusted prce duraton wth number of transactons and trade sze Stock #Trans Ave Vol/ #Trans ASND 9.18 (5.21) 0.39 (6.67) 0.21 (2.43) ) 0.02 (0.13) -1.11(-1.81) ORCL 5.74 (8.08) 0.32 (2.51) 0.44 (4.31) 0.94 (33.97) (-0.01) 1.55 (1.51) NSCP 15.07(4.45) 0.39 (3.94) 0.35 (2.91) 0.81 (36.97) (-2.91) (-1.58) SUBX 15.46(2.73) -0.07(-1.17) 0.77 (7.87) 0.78 (28.33) 1.91 (0.75) (-2.26) DURA 11.44(1.37) 0.05 (0.96) 0.74 (3.91) 0.71 (28.00) (-1.08) (-0.31) Average Average (wthout DURA) CLST 4.90 (3.33) 0.51 (4.60) 0.57 (8.43) 0.68 (31.45) (-1.87) (-1.74) ADTN 4.40 (2.38) 0.03 (0.89) 0.89 (15.93) 0.73 (35.15) (-3.55) (-0.95) IRIDF 2.16 (1.19) 0.35 (3.62) 0.65 (9.94) 0.68 (28.09) (-1.72) 2.00 (0.79) PSUN 5.11 (1.69) 0.24 (2.61) 0.56 (4.50) 0.57 (28.79) (-0.64) 1.48 (1.10) SEBL 3.73 (1.24) 0.16 (2.57) 0.64 (4.27) 0.62 (36.75) (-1.47) 2.54 (1.95) Average Average (wthout PSUN) SDTI 6.72(9.89) 0.05 (0.90) 0.77 (7.97) 0.68 (30.02) (-1.50) 2.25 (1.24) APOL 2.47 (0.99) 0.08 (1.83) 0.80 (10.19) 0.65 (29.00) (-1.30) 3.35 (1.86) LHSPF -2.07(-2.28) 0.72 (5.11) 0.44 (6.26) 0.79 (25.52) (-2.22) 4.72 (3.48) WIND 6.76 (2.79) 0.55 (4.32) 0.33 (3.88) 0.67 (33.46) (-1.30) 1.38 (1.28) CBSS 41.40(5.23) -0.27(-4.15) 0.61 (6.19) 0.78 (11.39) (-0.76) 4.82 (1.17) Average Ths table reports the parameter estmates of the WACD(1,1) model wth the number of transacton and trade sze, and the t-values (n parentheses) for three stock groups descrbed n Table 1. s the parameter of the Webull dstrbuton. Estmaton s based on the model x 1 1 # Trans Volume / Trans where x s the duraton and s the condtonal mean of the duraton between two arrval tmes t and t -1 ; #Trans s the number of transactons per duraton and Volume/Trans s trade sze or average volume per transacton. The unadjusted duraton was dvded by 60, the number of transactons was dvded by 100, and the average volume over number of transactons was dvded by Average parameter estmates are mean estmates for each group. Averages wthout DURA or PSUN are mean parameter estmates excludng each of these two stocks whch are removed because ther tradng frequences are too low to be qualfed n the hgh and medum frequency groups. The mean estmates excludng these stocks represent the group average usng the measure of trade frequency to defne the actveness of stocks.
22 45 Table 5 Estmates of the WACD(1,1) model for unadjusted prce duraton wth spread, number of transactons and average trade sze Stock Spread #Trans Ave Vol/ #Trans ASND 3.80 (1.72) 0.02 (0.86) 0.91 (17.28) 0.73 (34.35) 0.01 (0.40) (-3.03) (-0.93) ORCL 9.52 (2.81) 0.31 (2.96) 0.48 (5.12) 0.94 (34.42) (-3.28) (-0.17) 0.98 (0.52) NSCP (6.67) 0.34 (4.22) 0.44 (4.96) 0.81 (37.93) (-6.85) (-4.38) (-0.96) SUBX (4.43) 0.33 (2.98) (-2.67) 0.79 (26.65) 0.69 (1.43) (-2.09) (-1.52) DURA Average CLST 7.91 (3.27) 0.54 (4.88) 0.55 (8.23) 0.68 (31.91) (-2.23) (-2.16) (-0.76) ADTN 3.80 (1.72) 0.02 (0.86) 0.91 (17.28) 0.73 (34.35) 0.01 (0.40) (-3.03) (-0.93) IRIDF 4.13 (1.40) 0.34 (3.83) 0.66 (10.58) 0.68 (33.69) (-0.92) (-1.87) 1.26 (0.52) PSUN 3.73 (1.06) 0.24 (2.52) 0.55 (4.77) 0.57 (28.01) 0.04 (0.56) (-0.67) 1.51 (1.30) SEBL 6.01 (1.61) 0.17 (2.77) 0.62 (4.72) 0.62 (36.87) (-0.80) (-2.01) 2.64 (2.16) Average Average (wthout PSUN) SDTI (1.80) 0.10 (1.28) 0.29 (0.66) 0.68 (30.91) (-1.31) (-1.49) 0.06 (0.05) APOL (1.10) 0.10 (1.68) 0.72 (3.73) 0.65 (29.09) (-1.49) (-1.14) 3.49 (1.53) LHSPF 2.62 (0.62) 0.53 (1.84) 0.52 (2.91) 0.80 (25.34) (-1.56) (-4.12) 4.21 (1.84) WIND (3.82) 0.54 (4.49) 0.34 (4.23) 0.68 (32.45) (-2.50) (-1.66) 1.15 (1.17) CBSS (1.56) (-5.18) 0.21 (1.86) 0.81 (11.98) (-2.24) (0.41) 4.59 (0.56) Average Ths table reports the parameter estmates of the WACD(1,1) model wth spread, the number of transactons, and trade sze, and the t-values (n parentheses) for three stock groups descrbed n Table 1. s the parameter of the Webull dstrbuton. Estmaton s based on the model x 1 1 Spread # Trans Volume / Trans Spread s the percentage bd-ask spread and the remanng varables are as defned n Table 4. The unadjusted duraton was dvded by 60, the number of transactons was dvded by 100, the average volume over number of transactons was dvded by 1000 and the spread was multpled by 100. The estmates for DURA dd not converge. Average parameter estmates are mean estmates for each group. Averages wthout DURA or PSUN are mean parameter estmates excludng each of these two stocks whch are removed because ther tradng frequences are too low to be qualfed n the hgh and medum frequency groups. The mean estmates excludng these stocks represent the group average usng the measure of trade frequency to defne the actveness of stocks.
MgtOp 215 Chapter 13 Dr. Ahn
MgtOp 5 Chapter 3 Dr Ahn Consder two random varables X and Y wth,,, In order to study the relatonshp between the two random varables, we need a numercal measure that descrbes the relatonshp The covarance
More informationEvaluating Performance
5 Chapter Evaluatng Performance In Ths Chapter Dollar-Weghted Rate of Return Tme-Weghted Rate of Return Income Rate of Return Prncpal Rate of Return Daly Returns MPT Statstcs 5- Measurng Rates of Return
More informationTHE VOLATILITY OF EQUITY MUTUAL FUND RETURNS
North Amercan Journal of Fnance and Bankng Research Vol. 4. No. 4. 010. THE VOLATILITY OF EQUITY MUTUAL FUND RETURNS Central Connectcut State Unversty, USA. E-mal: BelloZ@mal.ccsu.edu ABSTRACT I nvestgated
More informationNYSE Specialists Participation in the Posted Quotes
European Journal of Economc and Poltcal Studes NYSE Specalsts Partcpaton n the Posted Quotes Bülent Köksal 1 Abstract: Usng 2001 NYSE system order data n the decmal prcng envronment, we analyze how the
More informationNonlinear ACD Model and Informed Trading: Evidence from Shanghai Stock Exchange
Nonlnear ACD Model and Informed Tradng: Evdence from Shangha Stock Exchange Woon K Wong, Dun Tan and Yxang Tan Paper IMRU 080402 Nonlnear ACD Model and Informed Tradng: Evdence from Shangha Stock Exchange
More informationCHAPTER 9 FUNCTIONAL FORMS OF REGRESSION MODELS
CHAPTER 9 FUNCTIONAL FORMS OF REGRESSION MODELS QUESTIONS 9.1. (a) In a log-log model the dependent and all explanatory varables are n the logarthmc form. (b) In the log-ln model the dependent varable
More informationMonetary Tightening Cycles and the Predictability of Economic Activity. by Tobias Adrian and Arturo Estrella * October 2006.
Monetary Tghtenng Cycles and the Predctablty of Economc Actvty by Tobas Adran and Arturo Estrella * October 2006 Abstract Ten out of thrteen monetary tghtenng cycles snce 1955 were followed by ncreases
More informationMultifactor Term Structure Models
1 Multfactor Term Structure Models A. Lmtatons of One-Factor Models 1. Returns on bonds of all maturtes are perfectly correlated. 2. Term structure (and prces of every other dervatves) are unquely determned
More informationUnderstanding price volatility in electricity markets
Proceedngs of the 33rd Hawa Internatonal Conference on System Scences - 2 Understandng prce volatlty n electrcty markets Fernando L. Alvarado, The Unversty of Wsconsn Rajesh Rajaraman, Chrstensen Assocates
More information3/3/2014. CDS M Phil Econometrics. Vijayamohanan Pillai N. Truncated standard normal distribution for a = 0.5, 0, and 0.5. CDS Mphil Econometrics
Lmted Dependent Varable Models: Tobt an Plla N 1 CDS Mphl Econometrcs Introducton Lmted Dependent Varable Models: Truncaton and Censorng Maddala, G. 1983. Lmted Dependent and Qualtatve Varables n Econometrcs.
More informationThe Effects of Industrial Structure Change on Economic Growth in China Based on LMDI Decomposition Approach
216 Internatonal Conference on Mathematcal, Computatonal and Statstcal Scences and Engneerng (MCSSE 216) ISBN: 978-1-6595-96- he Effects of Industral Structure Change on Economc Growth n Chna Based on
More informationTests for Two Correlations
PASS Sample Sze Software Chapter 805 Tests for Two Correlatons Introducton The correlaton coeffcent (or correlaton), ρ, s a popular parameter for descrbng the strength of the assocaton between two varables.
More informationCapability Analysis. Chapter 255. Introduction. Capability Analysis
Chapter 55 Introducton Ths procedure summarzes the performance of a process based on user-specfed specfcaton lmts. The observed performance as well as the performance relatve to the Normal dstrbuton are
More informationUNIVERSITY OF VICTORIA Midterm June 6, 2018 Solutions
UIVERSITY OF VICTORIA Mdterm June 6, 08 Solutons Econ 45 Summer A0 08 age AME: STUDET UMBER: V00 Course ame & o. Descrptve Statstcs and robablty Economcs 45 Secton(s) A0 CR: 3067 Instructor: Betty Johnson
More informationoccurrence of a larger storm than our culvert or bridge is barely capable of handling? (what is The main question is: What is the possibility of
Module 8: Probablty and Statstcal Methods n Water Resources Engneerng Bob Ptt Unversty of Alabama Tuscaloosa, AL Flow data are avalable from numerous USGS operated flow recordng statons. Data s usually
More informationElements of Economic Analysis II Lecture VI: Industry Supply
Elements of Economc Analyss II Lecture VI: Industry Supply Ka Hao Yang 10/12/2017 In the prevous lecture, we analyzed the frm s supply decson usng a set of smple graphcal analyses. In fact, the dscusson
More informationII. Random Variables. Variable Types. Variables Map Outcomes to Numbers
II. Random Varables Random varables operate n much the same way as the outcomes or events n some arbtrary sample space the dstncton s that random varables are smply outcomes that are represented numercally.
More informationREFINITIV INDICES PRIVATE EQUITY BUYOUT INDEX METHODOLOGY
REFINITIV INDICES PRIVATE EQUITY BUYOUT INDEX METHODOLOGY 1 Table of Contents INTRODUCTION 3 TR Prvate Equty Buyout Index 3 INDEX COMPOSITION 3 Sector Portfolos 4 Sector Weghtng 5 Index Rebalance 5 Index
More information/ Computational Genomics. Normalization
0-80 /02-70 Computatonal Genomcs Normalzaton Gene Expresson Analyss Model Computatonal nformaton fuson Bologcal regulatory networks Pattern Recognton Data Analyss clusterng, classfcaton normalzaton, mss.
More informationChapter 3 Student Lecture Notes 3-1
Chapter 3 Student Lecture otes 3-1 Busness Statstcs: A Decson-Makng Approach 6 th Edton Chapter 3 Descrbng Data Usng umercal Measures 005 Prentce-Hall, Inc. Chap 3-1 Chapter Goals After completng ths chapter,
More informationChapter 3 Descriptive Statistics: Numerical Measures Part B
Sldes Prepared by JOHN S. LOUCKS St. Edward s Unversty Slde 1 Chapter 3 Descrptve Statstcs: Numercal Measures Part B Measures of Dstrbuton Shape, Relatve Locaton, and Detectng Outlers Eploratory Data Analyss
More informationRandom Variables. b 2.
Random Varables Generally the object of an nvestgators nterest s not necessarly the acton n the sample space but rather some functon of t. Techncally a real valued functon or mappng whose doman s the sample
More informationLecture Note 2 Time Value of Money
Seg250 Management Prncples for Engneerng Managers Lecture ote 2 Tme Value of Money Department of Systems Engneerng and Engneerng Management The Chnese Unversty of Hong Kong Interest: The Cost of Money
More informationClearing Notice SIX x-clear Ltd
Clearng Notce SIX x-clear Ltd 1.0 Overvew Changes to margn and default fund model arrangements SIX x-clear ( x-clear ) s closely montorng the CCP envronment n Europe as well as the needs of ts Members.
More information3: Central Limit Theorem, Systematic Errors
3: Central Lmt Theorem, Systematc Errors 1 Errors 1.1 Central Lmt Theorem Ths theorem s of prme mportance when measurng physcal quanttes because usually the mperfectons n the measurements are due to several
More informationAppendix - Normally Distributed Admissible Choices are Optimal
Appendx - Normally Dstrbuted Admssble Choces are Optmal James N. Bodurtha, Jr. McDonough School of Busness Georgetown Unversty and Q Shen Stafford Partners Aprl 994 latest revson September 00 Abstract
More informationDomestic Savings and International Capital Flows
Domestc Savngs and Internatonal Captal Flows Martn Feldsten and Charles Horoka The Economc Journal, June 1980 Presented by Mchael Mbate and Chrstoph Schnke Introducton The 2 Vews of Internatonal Captal
More information2) In the medium-run/long-run, a decrease in the budget deficit will produce:
4.02 Quz 2 Solutons Fall 2004 Multple-Choce Questons ) Consder the wage-settng and prce-settng equatons we studed n class. Suppose the markup, µ, equals 0.25, and F(u,z) = -u. What s the natural rate of
More informationA MODEL OF COMPETITION AMONG TELECOMMUNICATION SERVICE PROVIDERS BASED ON REPEATED GAME
A MODEL OF COMPETITION AMONG TELECOMMUNICATION SERVICE PROVIDERS BASED ON REPEATED GAME Vesna Radonć Đogatovć, Valentna Radočć Unversty of Belgrade Faculty of Transport and Traffc Engneerng Belgrade, Serba
More informationPivot Points for CQG - Overview
Pvot Ponts for CQG - Overvew By Bran Bell Introducton Pvot ponts are a well-known technque used by floor traders to calculate ntraday support and resstance levels. Ths technque has been around for decades,
More informationSpatial Variations in Covariates on Marriage and Marital Fertility: Geographically Weighted Regression Analyses in Japan
Spatal Varatons n Covarates on Marrage and Martal Fertlty: Geographcally Weghted Regresson Analyses n Japan Kenj Kamata (Natonal Insttute of Populaton and Socal Securty Research) Abstract (134) To understand
More informationA Comparison of Statistical Methods in Interrupted Time Series Analysis to Estimate an Intervention Effect
Transport and Road Safety (TARS) Research Joanna Wang A Comparson of Statstcal Methods n Interrupted Tme Seres Analyss to Estmate an Interventon Effect Research Fellow at Transport & Road Safety (TARS)
More informationECONOMETRICS - FINAL EXAM, 3rd YEAR (GECO & GADE)
ECONOMETRICS - FINAL EXAM, 3rd YEAR (GECO & GADE) May 17, 2016 15:30 Frst famly name: Name: DNI/ID: Moble: Second famly Name: GECO/GADE: Instructor: E-mal: Queston 1 A B C Blank Queston 2 A B C Blank Queston
More informationISyE 512 Chapter 9. CUSUM and EWMA Control Charts. Instructor: Prof. Kaibo Liu. Department of Industrial and Systems Engineering UW-Madison
ISyE 512 hapter 9 USUM and EWMA ontrol harts Instructor: Prof. Kabo Lu Department of Industral and Systems Engneerng UW-Madson Emal: klu8@wsc.edu Offce: Room 317 (Mechancal Engneerng Buldng) ISyE 512 Instructor:
More informationRisk and Return: The Security Markets Line
FIN 614 Rsk and Return 3: Markets Professor Robert B.H. Hauswald Kogod School of Busness, AU 1/25/2011 Rsk and Return: Markets Robert B.H. Hauswald 1 Rsk and Return: The Securty Markets Lne From securtes
More informationConsumption Based Asset Pricing
Consumpton Based Asset Prcng Mchael Bar Aprl 25, 208 Contents Introducton 2 Model 2. Prcng rsk-free asset............................... 3 2.2 Prcng rsky assets................................ 4 2.3 Bubbles......................................
More informationEDC Introduction
.0 Introducton EDC3 In the last set of notes (EDC), we saw how to use penalty factors n solvng the EDC problem wth losses. In ths set of notes, we want to address two closely related ssues. What are, exactly,
More informationASSET LIQUIDITY, STOCK LIQUIDITY, AND OWNERSHIP CONCENTRATION: EVIDENCE FROM THE ASE
ASSET LIQUIDITY, STOCK LIQUIDITY, AND OWNERSHIP CONCENTRATION: EVIDENCE FROM THE ASE Ghada Tayem*, Mohammad Tayeh**, Adel Bno** * Correspondng author: Department of Fnance, School of Busness, The Unversty
More informationNotes on experimental uncertainties and their propagation
Ed Eyler 003 otes on epermental uncertantes and ther propagaton These notes are not ntended as a complete set of lecture notes, but nstead as an enumeraton of some of the key statstcal deas needed to obtan
More informationPrice and Quantity Competition Revisited. Abstract
rce and uantty Competton Revsted X. Henry Wang Unversty of Mssour - Columba Abstract By enlargng the parameter space orgnally consdered by Sngh and Vves (984 to allow for a wder range of cost asymmetry,
More informationR Square Measure of Stock Synchronicity
Internatonal Revew of Busness Research Papers Vol. 7. No. 1. January 2011. Pp. 165 175 R Square Measure of Stock Synchroncty Sarod Khandaker* Stock market synchroncty s a new area of research for fnance
More informationSurvey of Math: Chapter 22: Consumer Finance Borrowing Page 1
Survey of Math: Chapter 22: Consumer Fnance Borrowng Page 1 APR and EAR Borrowng s savng looked at from a dfferent perspectve. The dea of smple nterest and compound nterest stll apply. A new term s the
More informationIntroduction. Chapter 7 - An Introduction to Portfolio Management
Introducton In the next three chapters, we wll examne dfferent aspects of captal market theory, ncludng: Brngng rsk and return nto the pcture of nvestment management Markowtz optmzaton Modelng rsk and
More informationHighlights of the Macroprudential Report for June 2018
Hghlghts of the Macroprudental Report for June 2018 October 2018 FINANCIAL STABILITY DEPARTMENT Preface Bank of Jamaca frequently conducts assessments of the reslence and strength of the fnancal system.
More informationFinance 402: Problem Set 1 Solutions
Fnance 402: Problem Set 1 Solutons Note: Where approprate, the fnal answer for each problem s gven n bold talcs for those not nterested n the dscusson of the soluton. 1. The annual coupon rate s 6%. A
More informationCreating a zero coupon curve by bootstrapping with cubic splines.
MMA 708 Analytcal Fnance II Creatng a zero coupon curve by bootstrappng wth cubc splnes. erg Gryshkevych Professor: Jan R. M. Röman 0.2.200 Dvson of Appled Mathematcs chool of Educaton, Culture and Communcaton
More informationLabor Market Transitions in Peru
Labor Market Transtons n Peru Javer Herrera* Davd Rosas Shady** *IRD and INEI, E-mal: jherrera@ne.gob.pe ** IADB, E-mal: davdro@adb.org The Issue U s one of the major ssues n Peru However: - The U rate
More informationChapter 5 Student Lecture Notes 5-1
Chapter 5 Student Lecture Notes 5-1 Basc Busness Statstcs (9 th Edton) Chapter 5 Some Important Dscrete Probablty Dstrbutons 004 Prentce-Hall, Inc. Chap 5-1 Chapter Topcs The Probablty Dstrbuton of a Dscrete
More informationWhich of the following provides the most reasonable approximation to the least squares regression line? (a) y=50+10x (b) Y=50+x (d) Y=1+50x
Whch of the followng provdes the most reasonable approxmaton to the least squares regresson lne? (a) y=50+10x (b) Y=50+x (c) Y=10+50x (d) Y=1+50x (e) Y=10+x In smple lnear regresson the model that s begn
More informationWenjin Kang and Wee Yong Yeo. Department of Finance and Accounting National University of Singapore. This version: June 2007.
LIQUIDITY BEYOND THE BEST QUOTE: A STUDY OF THE NYSE LIMIT ORDER BOOK Wenjn Kang and Wee Yong Yeo Department of Fnance and Accountng Natonal Unversty of Sngapore Ths verson: June 2007 Abstract We conduct
More informationMode is the value which occurs most frequency. The mode may not exist, and even if it does, it may not be unique.
1.7.4 Mode Mode s the value whch occurs most frequency. The mode may not exst, and even f t does, t may not be unque. For ungrouped data, we smply count the largest frequency of the gven value. If all
More informationReal Exchange Rate Fluctuations, Wage Stickiness and Markup Adjustments
Real Exchange Rate Fluctuatons, Wage Stckness and Markup Adjustments Yothn Jnjarak and Kanda Nakno Nanyang Technologcal Unversty and Purdue Unversty January 2009 Abstract Motvated by emprcal evdence on
More informationModule Contact: Dr P Moffatt, ECO Copyright of the University of East Anglia Version 2
UNIVERSITY OF EAST ANGLIA School of Economcs Man Seres PG Examnaton 2012-13 FINANCIAL ECONOMETRICS ECO-M017 Tme allowed: 2 hours Answer ALL FOUR questons. Queston 1 carres a weght of 25%; Queston 2 carres
More informationOn the Style Switching Behavior of Mutual Fund Managers
On the Style Swtchng Behavor of Mutual Fund Managers Bart Frjns Auckland Unversty of Technology, Auckland, New Zealand Auckland Centre for Fnancal Research Aaron Glbert Auckland Unversty of Technology,
More informationTRADING RULES IN HOUSING MARKETS WHAT CAN WE LEARN? GREG COSTELLO Curtin University of Technology
ABSTRACT TRADING RULES IN HOUSING MARKETS WHAT CAN WE LEARN? GREG COSTELLO Curtn Unversty of Technology Ths paper examnes the applcaton of tradng rules n testng nformatonal effcency n housng markets. The
More informationInformation Flow and Recovering the. Estimating the Moments of. Normality of Asset Returns
Estmatng the Moments of Informaton Flow and Recoverng the Normalty of Asset Returns Ané and Geman (Journal of Fnance, 2000) Revsted Anthony Murphy, Nuffeld College, Oxford Marwan Izzeldn, Unversty of Lecester
More informationMoney, Banking, and Financial Markets (Econ 353) Midterm Examination I June 27, Name Univ. Id #
Money, Bankng, and Fnancal Markets (Econ 353) Mdterm Examnaton I June 27, 2005 Name Unv. Id # Note: Each multple-choce queston s worth 4 ponts. Problems 20, 21, and 22 carry 10, 8, and 10 ponts, respectvely.
More informationQuiz on Deterministic part of course October 22, 2002
Engneerng ystems Analyss for Desgn Quz on Determnstc part of course October 22, 2002 Ths s a closed book exercse. You may use calculators Grade Tables There are 90 ponts possble for the regular test, or
More informationChapter 10 Making Choices: The Method, MARR, and Multiple Attributes
Chapter 0 Makng Choces: The Method, MARR, and Multple Attrbutes INEN 303 Sergy Butenko Industral & Systems Engneerng Texas A&M Unversty Comparng Mutually Exclusve Alternatves by Dfferent Evaluaton Methods
More informationRisk Reduction and Real Estate Portfolio Size
Rsk Reducton and Real Estate Portfolo Sze Stephen L. Lee and Peter J. Byrne Department of Land Management and Development, The Unversty of Readng, Whteknghts, Readng, RG6 6AW, UK. A Paper Presented at
More informationOCR Statistics 1 Working with data. Section 2: Measures of location
OCR Statstcs 1 Workng wth data Secton 2: Measures of locaton Notes and Examples These notes have sub-sectons on: The medan Estmatng the medan from grouped data The mean Estmatng the mean from grouped data
More informationISE High Income Index Methodology
ISE Hgh Income Index Methodology Index Descrpton The ISE Hgh Income Index s desgned to track the returns and ncome of the top 30 U.S lsted Closed-End Funds. Index Calculaton The ISE Hgh Income Index s
More informationDynamic Analysis of Knowledge Sharing of Agents with. Heterogeneous Knowledge
Dynamc Analyss of Sharng of Agents wth Heterogeneous Kazuyo Sato Akra Namatame Dept. of Computer Scence Natonal Defense Academy Yokosuka 39-8686 JAPAN E-mal {g40045 nama} @nda.ac.jp Abstract In ths paper
More informationUsing Conditional Heteroskedastic
ITRON S FORECASTING BROWN BAG SEMINAR Usng Condtonal Heteroskedastc Varance Models n Load Research Sample Desgn Dr. J. Stuart McMenamn March 6, 2012 Please Remember» Phones are Muted: In order to help
More informationTests for Two Ordered Categorical Variables
Chapter 253 Tests for Two Ordered Categorcal Varables Introducton Ths module computes power and sample sze for tests of ordered categorcal data such as Lkert scale data. Assumng proportonal odds, such
More informationCorporate Governance and Equity Liquidity: An Analysis of S&P Transparency and Disclosure Ranking
Corporate Governance and Equty Lqudty: An Analyss of S&P Transparency and Dsclosure Rankng We-Peng Chen Humn Chung Cheng-few Lee We-L Lao ABSTRACT Ths paper nvestgates the effects of dsclosure and other
More informationarxiv: v1 [q-fin.pm] 13 Feb 2018
WHAT IS THE SHARPE RATIO, AND HOW CAN EVERYONE GET IT WRONG? arxv:1802.04413v1 [q-fn.pm] 13 Feb 2018 IGOR RIVIN Abstract. The Sharpe rato s the most wdely used rsk metrc n the quanttatve fnance communty
More informationSYSTEMATIC LIQUIDITY, CHARACTERISTIC LIQUIDITY AND ASSET PRICING. Duong Nguyen* Tribhuvan N. Puri*
SYSTEMATIC LIQUIDITY, CHARACTERISTIC LIQUIDITY AND ASSET PRICING Duong Nguyen* Trbhuvan N. Pur* Address for correspondence: Trbhuvan N. Pur, Professor of Fnance Char, Department of Accountng and Fnance
More information4. Greek Letters, Value-at-Risk
4 Greek Letters, Value-at-Rsk 4 Value-at-Rsk (Hull s, Chapter 8) Math443 W08, HM Zhu Outlne (Hull, Chap 8) What s Value at Rsk (VaR)? Hstorcal smulatons Monte Carlo smulatons Model based approach Varance-covarance
More informationiii) pay F P 0,T = S 0 e δt when stock has dividend yield δ.
Fnal s Wed May 7, 12:50-2:50 You are allowed 15 sheets of notes and a calculator The fnal s cumulatve, so you should know everythng on the frst 4 revews Ths materal not on those revews 184) Suppose S t
More informationInformational Content of Option Trading on Acquirer Announcement Return * National Chengchi University. The University of Hong Kong.
Informatonal Content of Opton Tradng on Acqurer Announcement Return * Konan Chan a, b,, L Ge b,, and Tse-Chun Ln b, a Natonal Chengch Unversty b The Unversty of Hong Kong May, 2012 Abstract Ths paper examnes
More informationHybrid Tail Risk and Expected Stock Returns: When Does the Tail Wag the Dog?
Hybrd Tal Rsk and Expected Stock Returns: When Does the Tal Wag the Dog? Turan G. Bal, a Nusret Cakc, b and Robert F. Whtelaw c* ABSTRACT Ths paper ntroduces a new, hybrd measure of covarance rsk n the
More informationTHE MARKET PORTFOLIO MAY BE MEAN-VARIANCE EFFICIENT AFTER ALL
THE ARKET PORTFOIO AY BE EA-VARIACE EFFICIET AFTER A OSHE EVY and RICHARD RO ABSTRACT Testng the CAP bols down to testng the mean-varance effcency of the market portfolo. any studes have examned the meanvarance
More informationMethod of Payment and Target Status: Announcement Returns to Acquiring Firms in the Malaysian Market
Method of Payment and Target Status: Announcement Returns to Acqurng Frms n the Malaysan Market Mansor Isa Faculty of Busness and Accountancy, Unversty of Malaya Lembah Panta, 50603 Kuala Lumpur, Malaysa
More informationMorningstar After-Tax Return Methodology
Mornngstar After-Tax Return Methodology Mornngstar Research Report 24 October 2003 2003 Mornngstar, Inc. All rghts reserved. The nformaton n ths document s the property of Mornngstar, Inc. Reproducton
More informationOPERATIONS RESEARCH. Game Theory
OPERATIONS RESEARCH Chapter 2 Game Theory Prof. Bbhas C. Gr Department of Mathematcs Jadavpur Unversty Kolkata, Inda Emal: bcgr.umath@gmal.com 1.0 Introducton Game theory was developed for decson makng
More informationSpurious Seasonal Patterns and Excess Smoothness in the BLS Local Area Unemployment Statistics
Spurous Seasonal Patterns and Excess Smoothness n the BLS Local Area Unemployment Statstcs Keth R. Phllps and Janguo Wang Federal Reserve Bank of Dallas Research Department Workng Paper 1305 September
More informationAn Empirical Study on Stock Price Responses to the Release of the Environmental Management Ranking in Japan. Abstract
An Emprcal Study on Stock Prce esponses to the elease of the Envronmental Management ankng n Japan Fumko Takeda Unversy of Tokyo Takanor Tomozawa Unversy of Tokyo Abstract Ths paper nvestgates how stock
More informationA copy can be downloaded for personal non-commercial research or study, without prior permission or charge
Sganos, A. (2013) Google attenton and target prce run ups. Internatonal Revew of Fnancal Analyss. ISSN 1057-5219 Copyrght 2012 Elsever A copy can be downloaded for personal non-commercal research or study,
More informationMeasures of Spread IQR and Deviation. For exam X, calculate the mean, median and mode. For exam Y, calculate the mean, median and mode.
Part 4 Measures of Spread IQR and Devaton In Part we learned how the three measures of center offer dfferent ways of provdng us wth a sngle representatve value for a data set. However, consder the followng
More informationAsset Management. Country Allocation and Mutual Fund Returns
Country Allocaton and Mutual Fund Returns By Dr. Lela Heckman, Senor Managng Drector and Dr. John Mulln, Managng Drector Bear Stearns Asset Management Bear Stearns Actve Country Equty Executve Summary
More information15-451/651: Design & Analysis of Algorithms January 22, 2019 Lecture #3: Amortized Analysis last changed: January 18, 2019
5-45/65: Desgn & Analyss of Algorthms January, 09 Lecture #3: Amortzed Analyss last changed: January 8, 09 Introducton In ths lecture we dscuss a useful form of analyss, called amortzed analyss, for problems
More informationCall & Put Butterfly Spreads Test of SET50 Index Options Market Efficiency and SET50 Index Options Contract Adjustment
Call & Put Butterfly preads est of E50 Index Optons Market Effcency and E50 Index Optons Contract Adjustment Woradee Jongadsayakul Abstract hs paper tests the effcency of E50 Index Optons market and nvestgates
More informationFORD MOTOR CREDIT COMPANY SUGGESTED ANSWERS. Richard M. Levich. New York University Stern School of Business. Revised, February 1999
FORD MOTOR CREDIT COMPANY SUGGESTED ANSWERS by Rchard M. Levch New York Unversty Stern School of Busness Revsed, February 1999 1 SETTING UP THE PROBLEM The bond s beng sold to Swss nvestors for a prce
More informationCongrès de l Association canadienne d économique Canadian Economic Association Meeting
Congrès de l Assocaton canadenne d économque Canadan Economc Assocaton Meetng May 2013, HEC Montréal Georges Donne, HEC Montréal Research n collaboraton wth Olfa Maalaou Chun, Kast Graduate School of Fnance,
More informationFM303. CHAPTERS COVERED : CHAPTERS 5, 8 and 9. LEARNER GUIDE : UNITS 1, 2 and 3.1 to 3.3. DUE DATE : 3:00 p.m. 19 MARCH 2013
Page 1 of 11 ASSIGNMENT 1 ST SEMESTER : FINANCIAL MANAGEMENT 3 () CHAPTERS COVERED : CHAPTERS 5, 8 and 9 LEARNER GUIDE : UNITS 1, 2 and 3.1 to 3.3 DUE DATE : 3:00 p.m. 19 MARCH 2013 TOTAL MARKS : 100 INSTRUCTIONS
More informationProblem Set 6 Finance 1,
Carnege Mellon Unversty Graduate School of Industral Admnstraton Chrs Telmer Wnter 2006 Problem Set 6 Fnance, 47-720. (representatve agent constructon) Consder the followng two-perod, two-agent economy.
More information3 Portfolio Management
Mathematcal Modelng Technques 69 3 ortfolo Management If all stock predctons were perfect, portfolo management would amount to the transfer of funds to the commodty that promses the hghest return n the
More informationChapter 5 Bonds, Bond Prices and the Determination of Interest Rates
Chapter 5 Bonds, Bond Prces and the Determnaton of Interest Rates Problems and Solutons 1. Consder a U.S. Treasury Bll wth 270 days to maturty. If the annual yeld s 3.8 percent, what s the prce? $100 P
More informationMaturity Effect on Risk Measure in a Ratings-Based Default-Mode Model
TU Braunschweg - Insttut für Wrtschaftswssenschaften Lehrstuhl Fnanzwrtschaft Maturty Effect on Rsk Measure n a Ratngs-Based Default-Mode Model Marc Gürtler and Drk Hethecker Fnancal Modellng Workshop
More informationHewlett Packard 10BII Calculator
Hewlett Packard 0BII Calculator Keystrokes for the HP 0BII are shown n the tet. However, takng a mnute to revew the Quk Start secton, below, wll be very helpful n gettng started wth your calculator. Note:
More informationConditional Beta Capital Asset Pricing Model (CAPM) and Duration Dependence Tests
Condtonal Beta Captal Asset Prcng Model (CAPM) and Duraton Dependence Tests By Davd E. Allen 1 and Imbarne Bujang 1 1 School of Accountng, Fnance and Economcs, Edth Cowan Unversty School of Accountng,
More informationFinancial mathematics
Fnancal mathematcs Jean-Luc Bouchot jean-luc.bouchot@drexel.edu February 19, 2013 Warnng Ths s a work n progress. I can not ensure t to be mstake free at the moment. It s also lackng some nformaton. But
More informationA Meta Analysis of Real Estate Fund Performance
A Meta Analyss of Real Estate Fund Performance A Paper Presented at the ARES Annual Meetng Aprl 00 Naples, Florda Abstract Stephen Lee, Unversty of Readng * and Smon Stevenson, Unversty College Dubln Ths
More informationAn Application of Alternative Weighting Matrix Collapsing Approaches for Improving Sample Estimates
Secton on Survey Research Methods An Applcaton of Alternatve Weghtng Matrx Collapsng Approaches for Improvng Sample Estmates Lnda Tompkns 1, Jay J. Km 2 1 Centers for Dsease Control and Preventon, atonal
More informationData Mining Linear and Logistic Regression
07/02/207 Data Mnng Lnear and Logstc Regresson Mchael L of 26 Regresson In statstcal modellng, regresson analyss s a statstcal process for estmatng the relatonshps among varables. Regresson models are
More informationBid-auction framework for microsimulation of location choice with endogenous real estate prices
Bd-aucton framework for mcrosmulaton of locaton choce wth endogenous real estate prces Rcardo Hurtuba Mchel Berlare Francsco Martínez Urbancs Termas de Chllán, Chle March 28 th 2012 Outlne 1) Motvaton
More informationHow diversifiable is firm-specific risk? James Bennett. and. Richard W. Sias * October 20, 2006
How dversfable s frm-specfc rsk? James Bennett and Rchard W. Sas * October 0, 006 JEL: G0, G, G, G4 Keywords: dversfcaton, dosyncratc rsk * Bennett s from the Department of Accountng and Fnance, Unversty
More informationDigital assets are investments with
SANJIV R. DAS s a professor at Santa Clara Unversty n Santa Clara, CA. srdas@scu.edu Dgtal Portfolos SANJIV R. DAS Dgtal assets are nvestments wth bnary returns: the payoff s ether very large or very small.
More informationEconomic Design of Short-Run CSP-1 Plan Under Linear Inspection Cost
Tamkang Journal of Scence and Engneerng, Vol. 9, No 1, pp. 19 23 (2006) 19 Economc Desgn of Short-Run CSP-1 Plan Under Lnear Inspecton Cost Chung-Ho Chen 1 * and Chao-Yu Chou 2 1 Department of Industral
More information