HD2S .K414 ALFRED P. WORKING PAPER SLOAN SCHOOL OF MANAGEMENT EARNINGS INFORMATION CONVEYED BY DIVIDEND INITIATIONS AND OMISSIONS by Paul M. Healy and Krishna October, 1^87 G. Palepu WP #1943-87 MASSACHUSETTS INSTITUTE OF TECHNOLOGY 50 MEMORIAL DRIVE CAMBRIDGE, MASSACHUSETTS 02139 EARNINGS INFORMATION CONVEYED BY DIVIDEND INITIATIONS AND OMISSIONS by Paul M. Healy and Krishna October, 1987 G. Palepu WP #1943-87 EARNINGS INFORMATION CONVEYED BY DIVIDEND INITIATIONS AND OMISSIONS Paul M. Hcaly Sloan School of Management Massachusetts Institute of Technology Krishna (J. I'alcpu Graduate School of Business Administration Harvard University September 19R7 We wish to thank Paul Asquith and David Mullins for making their data available for The paper benefited from the comments of seminar participants at this University of California paper - Berkeley, Carnegie- Mellon University, University of Chicago, Harvard University, Massachusetts Institute of Technology, Northwestern University, University of Oregon, University of Pennsylvania, University of Rochester, and Stanford University. Paul Asquith, Bob Kaplan, Richard Ross Watts, and the We are particularly grateful to Lcftwich, John long. Stew Myers, Pat O'Brien, Rick Ruback, referee for their suggestions. M.I.T. MOV LIBRARIES 2 1987 RECEIVED 1 ABSTRACT This paper examines earnings changes surrounding firms' decisions to initiate or omit dividend payments. Firms changes both before and that initiate (omit) dividend after the dividend policy positively related to the dividend announcement announcement of subsequent earnings change. return. positive (negative) earnings The subsequent earnings changes arc Also, the stock price reactions arc smaller than usual. (1) dividend initiation/omission decisions reflect (2) the payments have These at the results suggest that: both past and future earnings performance; and market interprets the announcement of these decisions as managers' forecasts of future earnings changes. I. INTRODUCTION This paper examines whether dividend policy changes convey information about future earnings. The topic has been widely discussed in the finance literature. Modigliani and Miller (1958) demonstrate that, under conditions of perfect capital markets anil zero taxes, dividends do not affect the content if However, they contend value of the firm. managers have superior information to investors that information to set current dividends (sec Miller That is, a dividend that dividends change may indicate a change in on the may have information firm's future earnings and Modigliani (1961) and and use l.intncr (1956)).' management's expectations of future Dividend changes can thus be thought of as management forecasts of future earnings earnings changes substantiated by cash. The hypothesis on the information content of dividend changes has been formalized by Bhattacharya (1979, 1980), John and Williams (1985), Miller and Rock (1985) and Offer and Thakor (1987). In a Miller and Rock summarize this idea as follows: world of rational expectations, the firm's dividend (or financing) announcements provide just enough pieces of the firm's sources and uses statements for the market to deduce the unobserved The market's piece, to wit, the firm's current earnings. estimate of current earnings contributes in turn to the estimate of the expected future earnings Watts (1973), dividends, examines earnings; and (2) on which the issues: (1) the relation abnormal stock returns decreases in dividends. market value largely hinges (p. 1031). an one of the early empirical studies of the information content of in two firm's He concludes for firms that out, his study has stock price data, makes it two announce unexpected increases and that current dividends provide earnings and there are no abnormal returns As Watts points between unexpected current dividends and future in months surrounding limitations. difficult to distinguish First, the between the little information on future the dividend announcements. use of monthly, rather than daily effect of dividend and other Lintner provides empirical evidence that managers consider past as well as future earnings in setting current dividends. Gonedes (1978) also reports similar findings contemporaneous information model reduces the power of the A They use Second, the potential noise releases. in the dividend expectation tests. recent study by Asquith and Mullins (19R.1) attempts to mitigate the above daily stork price data to control for other announcements, such announcements as earnings lhat they believe are least likely to lie anticipated, contemporaneous information Also, they select a sample of dividend changes namely dividend Mullins find that there are significant positive abnormal returns announcements. Other studies also use daily data announcement of unanticipated dividend While recent studies document problems at Asquith and initiations the dividend initiation and document abnormal returns at the increases and decreases." a significant stock price reaction to dividend policy changes, they do not re-examine the relation between dividend policy changes and subsequent earnings, the second issue analyzed by Watts. Our this issue announcement inferred tests differ The purpose of from those of Watts in this paper two ways returns, rather than unexpected dividends, as a by the market from dividend announcements; and policy changes that have been announcement returns, documented in (2). is ( 1) to provide fresh evidence we on use dividend measure ol the we focus on information the two dividend the literature as having the largest average dividend initiations and omissions. 3 These include Aharoncy and Swary (1980), Bricklcy (1983), Kalay and and Dielman and Oppcnheimer ( 0R4). 1 owenstcin 1 and omissions have on average the largest announcement imply that they are the largest dollar or percentage dividend Restricting our analysis to initiations and omissions allows us to examine dividend While dividend initiations returns, this docs not necessarily changes. power of our tests beyond these extreme policy changes with the largest information content, thereby increasing the However, this research design restricts the gencralizabilitv, ol out results dividend changes relevant to all The reader should therefore be careful not to interpret our findings as dividend policy changes. ( 1985), Our sample comprises after a hiatus of at least ten 131 firms that years and 172 firms that omit dividend or after continuously paying for issues. First, pay dividends for the at least we document changes after a dividend initiation or in a firm's omission. The ten years. return Finally, payments for the first lime designed to examine three tests are earnings performance for fi\e years before and Next, we examine whether subsequent earnings changes arc related to the information released at the dividend announcement lime or resume payments first we analyze announcement, as measured by the the market reaction to earnings announcements subsequent to the dividend policy change to assess whether the market anticipates these earnings from the dividend announcement. The results of our increases in their annual earnings for as initiation. Firms that tests arc as follows. many initiate as five years before Firms that omit dividend payments have dividends have significant and the year of the dividend a significant decrease in their earnings two years before and in the year of the dividend omission consistent with those reported by I intner (1956), annual These findings arc Fama and Habiak (1%S) and Watts (1972), and suggest that dividend initiations and omissions can, in part, be predicted by changes in past and current earnings to the Similar to earlier studies, announcement of predicted and convey we find that there is a significant market reaction these dividend policy changes, indicating that they cannot be perfectly new information. Tests of the earnings performance subsequent to the dividend policy changes lead to three conclusions. for two years First, there arc significant increases in after this event. omission firms there is earnings for firms that initiate dividends These earnings increases appear to be a significant decrease in earnings for only permanent 1 or dividend one year following the event Further, this earnings decrease appears to be temporary since there are significant offsetting earnings increases in the subsequent two years Second, earnings changes in the year of and year following a dividend initiation or omission arc positively related to the information that is revealed by the dividend announcement as measured by reaction at the dividend initiation or omission available prior to the dividend relation lie two day abnormal stock announcement I his relation is price found to exist changes and information on future earnings performance after controlling for prior earnings is t For dividend omissions there announcements. is a negative between the announcement return and earnings changes three years subsequent event, consistent with the earnings recovery noted above. announcements following price reactions to earnings significantly less than at to the magnitude of the stock Finally, the the dividend initiation or omission are normal, indicating that these earnings changes anticipated by the market that the date of the dividend are, at least in part, announcement Together, the above three findings indicate that the information conveyed by dividend initiations and omissions announcement of to the is related to earnings in the year of ami one year subsequent these dividend policy changes. This evidence is The results arc also consistent with dividend information hypothesis. in changes making dividend policy decisions managers consider past, current I consistent with the miner's description that and future earnings. Investors therefore interpret dividend initiations and omissions as changes in managements forecasts of firms' future earnings.- One we examine bias is effect limitation of our study firms' is that there is an ex post selection bias which have post-dividend earnings data. One I lie is is organized as follows In the next section, research does not address whether investors value dividends per full managers consciously use dividends we describe the sc. that is whether Also, our results cannot be interpreted higher payouts arc associated with higher stock prices as indicating that However, the unknown remainder of the paper Our our sample since possible manifestation of this the pattern of earnings recovery for the dividend omission firms of the bias on our resulls in as signals" of future earnings. data employed in our empirical analysis. their results a summary and tests and discussion of the results DATA 2. 2. The paper concludes with The third section describes the empirical 1 Dividend Initiation lest Sample Our dividend initiation sample comprises the firms used by Asquith and Mullins (1983) announcements on shareholders' wealth their study of the effect of initiating dividend and Mullins define an resumption of initiating dividend as the All first and the period studied extends to at least one year after the firm was listed Their dividend payments 198(1. on For all either the Asquith and Mullins' sample of 168 firms is initial in the New York ten year screen initial after dividend was paid or American Stock Exchange. Moody's from several sources: Dividend Record, Standard and Poor's Dividend Record, the Center for Research Prices, was sample therefore occur sample firms, the selected history or the in a firm's a dividend after a hiatus of at least ten years. January 1954 to December 1963. 1963, dividend first Asquith in Security and the Wall Street Journal. The dividend announcement date, defined as the date when news of the forthcoming dividend first appeared in the Wall Street Journal, the stock price two days before the dividend announcement and stock returns for the day before and the day of the dividend announcement arc collected by Asquith and Mullins for each of the sample firms For the above 168 firms we earnings announcement dates collect the following additional data: prior to the dividend initiation subsequent annual earnings announcement dates. collected from the Wall Street Journal Index. (2) dates. These data are collected I) The announcement and 'These earnings six fiscal year the five announcement dates arc- Annual earnings per share before extraordinary items and discontinued operations reported announcement ( at the above eleven earning from the 1984 Compustat Annual Industrial and in Finns arc included Research tapes. in the final announcement dates and earnings data Our and Mullin sample, 14 arc not listed if at lcasl eight ol ihc eleven earning 1 ; arc available. usable sample comprises 131 firms s sample Of on Oompustat the 37 firms that arc eliminated from Asquilh files; ten do not have the required number of Wall Street Journal earnings announcement dates; three have insufficient earnings data before the dividend announcement new as they were listings insufficient earnings data after the dividend initiation on the NYS1 announcement oi AS1 and ten have ; (eight of these firms were acquired, one was involved in an exchange transaction and one was delisted). We define the first fiscal year earnings announced prior to the dividend initiation announcement as earnings for year -1; the announcement is the year -1 first annual earnings announced defined as earnings for year 0. announcement The five dividend annual earnings announced prior to arc defined as earnings for years -6 to -2. earnings announced subsequent to the year after the announcement Similarly, the four annual are defined as earnings for years 1 to 4 Since dividend announcements occur throughout the above include quarterly earnings One of our announced that are fiscal year, year earnings defined before, as well as after the dividend compares markcl reactions to annual earnings announcements before he data and after the dividend announcements using time-series data for each sample firm for each firm, imposed restriction, that at least eight earnings observations arc available is primarily to perform this test. This restricted sample is used in all tests reported in the paper so that the results are consistent. However, we also perform all tests other than this market tests I reaction test using available earnings data for the 16R firms reported in the The results arc similar to those paper. As noted above, excluding bias for our study. Discussion of the these companies from the analysis effect of this bias on the results is creates an ex post survival deferred to later in the paper. B . . .... announcement announcements is assigned to ycar-1, thcicbv ensuring that all the years f) to 4 arc announced subsequent to the dividend announcement Six firms report annual earnings concurrently with the dividend initiation ,ach of these earnings ; I earnings assigned to announcement. This introduces a potential hias to our findings regarding earnings performance subsequent to the dividend To initiation. address this problem, wc perform additional tests using quarterly earnings data. We announcement date collect the first quarterly earnings date from the Wall Street Journal Index prior to the dividend initiation Quarterly earnings per share before extraordinary items and discontinued operations reported on this date, and earnings reported quarterly announcements summed and are defined as earnings for year summed earnings reported following the dividend date arc announcement, and earnings -1 are for year announced arc strictly announced and Poor's Compustat Quarterly Industrial tapes Quarterly data arc available for 129 of our 131 The four quarterly this procedure. Defined this before the dividend initiation after the dividend date. We use Standard as the source of quarlcrly earnings data. test firms from a search of these tapes 131 dividend initiation firms are in 38 different 2-digit evidence of industry clustering within the sample initiating dividends The to construct earnings for year 0. Farnings for year -2 arc also constructed from quarterly data using way, earnings for years -2 and -1. the three prior at Table 1 SIC presents the by year. The most frequent years of dividend There industries. number of sample initiation in the no is firms sample arc 1976 (32 firms) and 1977 (25 firms). 'The dividend initiating year for the remaining 74 firms in the sample ranges from 1970 to 1979. 2.2 Dividend Omission Sample Our initial Compustat Annual Wc 1 dividend omission sample Industrial is identified by searching the 1984 Standard and Poor's and Research tapes, and the CRSP tapes. Wc first list all the New are grateful to the referee for the suggestion leading to this analysis. Standard and Poors do not construct information for companies that were delisted, Quarterly Industrial tape. we a Research tape for quarterly data searched old copies of the Compustat To collect York and American Stock Exchange firms on From period 1969- 1980. history (if this list \vc select firms that they have been listed for paying dividends for less years at least ten these tapes which omitted dividends during the omit dividends for the first We (if they have been listed for ten years or more). Out The thirty firms date, the date when news of announcement date from the CRSP (3) stock returns for the from the CRSP data files; (4) list of a dividend sample thus comprises 210 companies. above firms: (1) the dividend the forthcoming dividend omission the Wall Street Journal Index; (2) the stock price Record; initial collect the following data for each of the announcement in their than ten years) and firms that omit after continuously the 240 firms identified from this search, the Wall Street Journal Index does not omission announcement date for time two days before omission first appeared in the dividend omission Daily Master Tape or Standard and Poor's Daily Stock Price day before and the day of the dividend omission announcement annual earnings announcement dates and reported earnings per share before extraordinary items and discontinued operations for years -6 to 4, using the procedures described above for the dividend initiation sample. The final sample consists of 172 reasons, 17 firms had missing earnings firms. Of the 38 firms excluded for data availability announcement dates in the Wall Street Journal Index; ten have insufficient data prior to the dividend omission as they were new ASF,; and eleven firms have insufficient data after the dividend listings announcement on the NYSE or (four were acquired, six were delisted and one firm's exchange listing was suspended). Fifteen firms report annual earnings concurrently with the dividend omission announcement. that all Each of these earnings announcements is assigned to year- 1, therein ensuring to 4 arc announced subsequent to the dividend the earnings assigned to years announcement Discussion of the excluding these firms is effect on our results of the ex post survival bias that arises deferred to later in the paper. from As noted for dividend initiations, our method of aligning annual earnings relative to the dividend announcment date leads to a potential bias in our tests of post-omission earnings correct this problem, variables for years -2, we adopt -1 and procedure described in Section 3.1 to Quarterly data arc available for 129 of our 172 172 dividend omission firms arc in 42 different 2-digit evidence of industry clustering within the sample. The most omitting dividends by year. and 1971 (33 construct earnings using quarterly earnings per share before extraordinary items and discontinued operations. The tlie Table 1 SIC test firms. industries presents the number There is no of sample firms frequent years of dividend omission are 1970 (50 firms) The dividend omission years for the remaining S9 firms firms). To in the sample range from 1969 to 1980 2.3 Comparison Samples A matched evaluating test firms' same industry at sample of comparison firms as Each comparison firm earnings performance. its lest Industry matches are based firm match. the date of the dividend initiation or omission sample is required to: (1) be have stock price data on the listed on CRSP same earnings data requirements cither the announcement New York Master Tape or Record two days prior to the dividend the collected to provide an earnings is initiation or in randomly on the selected test firm's Each firm in the for from the SIC code comparison or American Stock Exchange; (2) Standard and Poor's Daily Stock omission by as the initiation is benchmark its and omission match test test firm; and 1'ricc (3) meet firms Quarterly data are available for relatively fewer dividend omission firms than for the initiating sample. As can be seen in Table when quarterly data available on occur in the I, Compustat the omissions arc clustered in the early 1970s, arc less complete, whereas the initiations tend to mid-1970s. For test firms currently listed on the Compustat Research tape, we select the comparison firms from the Research tape This is done to avoid survivorship bias in the comparison samples. For the remaining test firms we select the comparison firms from the regular Compustat tape. We omission For DO for dividend initiation test tests that require quarterly earnings 101 per share wc are able to find data for omission comparison firms. TESTS AND RESULTS The results of four tests arc reported in this section, reaction to the announcement of dividend initiations first, wc describe the market Second, wc examine and omissions. earnings changes in the five years before and after these dividend events. between the market reaction analyzed. Finally, announcements 3.1 firms and 171 dividend firm each in the dividend initiation and omission samples could not comparison firms and 18 initiating 3. One test firms. he matched. 1 comparison firms arc able 1o find is we test to the dividend is whether the the market reaction to the subsequent earnings Market Reaction to Dividend We announcements and subsequent earnings changes than the normal reaction. less Third, the relation Initiations These and tests results arc described below. and Omissions estimate abnormal returns for dividend initiation and omission firms for the period 60 days before to 20 days after the announcement. market-adjusted returns, that is, Abnormal returns are defined as the difference between firms' returns and returns on the CRSP equal-weighted market portfolio. Mean abnormal are reported in Table returns for various holding periods surrounding the dividend The mean announcement 2. return (days -1 and announcements 0) for the initiation firms Where possible, SIC matches are based on 4-digit codes Wc arc able to find 122 matches for the initiation sample, and 159 such matches for the omission sample. The remaining matches are based on 2-digit industry codes. 4-digit ' The dividend services, except motion initiation firm that picture, industry. is unmatched is in the The dividend omission amusement and recreation is unmatched is in firm that the apparel and accessory stores industry. ' The Abnormal results reported in to the definition of returns were also estimated as risk-adjusted returns from a market Table 2 and the other abnormal tests returns. ic model reported later in the paper arc not sensitive is 3.9% and is 1% statistically significant at the level. There also evidence that initiating is Mean firms have significant positive returns in the prc-announccmcnt period -60 to -21, -20 to 1% level. -1 These 1 and -10 to -1 are results are similar to those reported sample of 168 firms. 1 significant -1 are our 1% respectively, all by Asquith and Mullins significant at the for their full 8 For the dividend omission significant at the 3.5%, 1.1% and 4.0% returns for days As level. mean two day announcement firms, the in the case of dividend initiations, Mean pre-announcement returns -7.0%, -2.7% and -7.0% respectively, wc return is -9 5%, find evidence of returns for days -60 to -21, -20 to -11 and -11 to all significant at the 1% percent level. Once again results are similar to those reported in earlier studies. The above findings indicate that investors partially anticipate dividend initiations and omissions from other information available prior to the announcement of the dividend policy change. However, these events arc not fully anticipated: the actual announcement of the policy changes conveys information to the market. 3.2 framings Changes Surrounding Dividend Initiations and Omissions Studies by Ball and Brown (1968), Ball and Watts (1972), and Watts and suggest that annual earnings follow a random sample of firms is random walk. initiations (omissions) arc preceded by a number of eftwich (1977) Thus, the average earnings changes However, expected to be zero. I I for a intner (19S6) implies that dividend years of earnings increases (decreases) Asquith and Mullins report a two day announcement return of 3.7 percent for the full Dielman and Oppcnhcimcr report a two day announcement return of 3.5 percent for a sample of 39 firms that resume cash dividends. sample. For example, Dielman and Oppcnhcimcr report -8.1 a percent for a sample of 53 firms that omitted cash dividends 1 I mean two day announcement return of Further, if dividend policy changes convey information on future earnings, dividend initiations (omissions) arc expected to he followed by earnings increases (decreases). To examine whether we or omit dividends, there arc systematic earnings patterns exhibited by firms that initiate calculate earnings changes for five years before (years -5 to -1), the year of (year 0) and the four years after (years results across firms, two days prior price firm year in j t, AF: ( AF where j in we is to 4) the dividend policy announcement, P:. To aggregate The standardized change in earnings for therefore defined as: = (Ejt jt -Ejt )/Pj t = -5 4 F:, arc earnings per share before extraordinary items year change express earnings changes in these years as a percentage of the stock to the dividend , 1 (1) and discontinued operations for firm t. Standardized earnings changes are computed for years -5 to 4 for the dividend initiation and omission firms and mean and median addition, we for the same fiscal years for the comparison mate lies Our tests standardized earnings changes for the intiation and omission firms. examine In analyze industry-adjusted standardized earnings changes for these firms, defined as the difference in standardized earnings changes for the initiation/omission and comparison firms. Dividend initiation results Mean and median earnings changes as a percentage of equity price are reported of Table 3 for the dividend initiating firms for years -5 to on 4. we only require that firms have earnings change data in seven 12 Panel Panel M reports mdustry- The number of firms with available earnings change data since in (if differs across years -5 to 4 the ten years. A adjusted numbers for these same firms are insignificant in years -5 to 1% initiating firms are -1. As the results in Panel matched by mean 4.3% and is annouccment, there B is These findings indicate significant at the is a further 5.5% increase in indicate, the earnings increases of the similar earnings increases for the industry However, the earnings increase industry factors. standardized earnings changes for the initiation In year -1 the -2. In year 0, the year of the dividend level. standardized earnings. year Mean firms. comparison firms in cannot be attributed to for initiating firms in year that dividend initiating firms arc in growth industries but have superior earnings performance in the year of the dividend initiation. One announcement we year subsequent to the dividend earnings change is 2.2% (p = 0.07); the following year the earnings changes are insignificant in years 3 and significantly larger than that for the industry' The 4. find thai the mean change is mean 3.5% (p= earnings increase for the comparison firms only in standardized Mean 0.01). firms test is year 2 Conclusions from median earnings changes are generally consistent with the above findings. For the test -5, -4, -2 periods before (years 2) the firms there are significant median earnings increases in four of the dividend announcement. significant in two of and -1), the year of (year 0) and two years after (years 1 and Median industry-adjusted standardized earnings changes the years before (years -4 and -3), the year of (year 0) and arc two years following the dividend event. Thus, firms that prior to the dividend initiate dividends experience earnings growth starling as announcement. announcement and two subsequent The years. many as five years earnings growth continues in the year of the dividend These findings arc consistent with the hypothesis thai The mean earnings change for year 1, reported in Table 3, includes an observation which has an earnings decline that is 86% of price. This company is Valmac Industries, which in this year acquired Rite Care Corporation and recorded a large loss on this new business. If this observation is excluded, the sample mean in year is significant at the 1% level in both Panels A and B. 1 1 5 managers consider past and current performance as well as expectations of future earnings in the dividend initiation decision. One limitatioti of the above results is that they arc based on earnings data that arc reported annually whereas changes in dividend policy arc reported throughout the year. median number of trading days between the announcement of the dividend first subsequent annual earnings announcement date is 171 trading days. initiation earnings in Table 3 announcements that precede the dividend To examine replicate the strictly above whether our tests may have been reported We using quarterly earnings. earnings are created from the four to the dividend change; year and year -1 for quarters -5 to -R. we redefine annual earnings so that they are -1 earnings are constructed from the four quarterly earnings announced prior to the event; and year year quarterly earnings results arc sensitive to the use of fiscal year earnings data, announcements subsequent from announcements in fiscal year. announcement aligned with the dividend announcements: year quarterly and the This indicates that dividend initiation announcements occur after approximately one quarter of the Therefore, part of the year The -2 earnings arc created Realigned earnings changes arc then calculated for and arc standardized by the stock price two days prior to the dividend announcement. Mean and median values of raw standardized earnings changes initiation firms are report in Panel Panel B A of Table In contrast to the year D earnings in announced strictly after the We adopt this in years -1 and Industry-adjusted figures are reported 4. Table 3, year (1 earnings for the in in this table are dividend announcement. approach since we do not have enough time estimate firm-specific quarterly earnings expectation models. unlike annual earnings, quarterly earnings do not follow a 14 scries observations to Previous results suggest that, random walk (Sec Foster (1977)) Mean year standardized earnings changes for the initiation firms arc These are both 0. significant at the 1% one year prior earnings, the earnings change in year -I and 4.5% In contrast to results for fiscal yeai level. announcement to the dividend earnings change in the event year, indicating that the 5.4% is larger than the year findings understate earnings fiscal changes prior to the dividend initiation and overstate earnings changes subsequent to the event. The mean at the 5% industry-adjusted standardized earnings change The mean level. for year is 1.23% and is attributable to a small number of extreme observations difference for year 1.72% and is is significant at the the significant annual earnings changes for year 0, is 2.85% in vent -I not significant in the 5% in is However, comparison sample: significant this finding the is median These findings indicate that level reported and Table 3, cannot be fully announced before the dividend announcement. attributed to quarterly earnings Dividend omission results Panel A of Table 5 presents summary standardi7.cd earnings changes in years -5 to statistics for the Panel 4. B dividend omission firms presents the industry-adjusted standardized earnings changes for the same years. The The means There is test firms' for years -2 mean and also a significant omission. The standardized earnings changes are insignificant -1 arc -1.2% and -7.7% respectively and -13.5% mean earnings change results in Panel B in year 0, years -5 to -3. arc statistically significant the year of the dividend indicate that the earnings decline in year -2 could be However, attributed to industry-related factors in years -1 and earnings declines even after adjusting for industry performance initiations, dividend in the omission firms have Thus, similar to dividend omissions follow significant earnings changes Subsequent to the dividend announcement, the omitting firms experience two years of significant positive earnings: 6.3%. in year I and 9.4% is in year 2 The earnings increases cannot in be attributed to industry factors as they earnings patterns following dividend initiations. initiation persist for several years omissions reverse These persist in Panel B. The results differ from the earnings increases prior to a dividend subsequent, whereas the declines in earnings prior to dividend subsequent years. in The above findings indicate that dividend omissions are preceded by declines However, these declines do not persist beyond year 0. Year between the announcement of the dividend omission and the cannot 136 days, about two is (The median number of trading first subsequent annual earnings Based on the above fiscal quarters.) infer that there arc earnings declines following the dividend omission. issue further, Once we announcements. we To Year (-1) earnings two days prior are reported in Panel to the dividend A of Table 6. These values are after the in year -1 and -6.6% announcement in raw earnings in Industry-adjusted numbers are reported in Panel significant at the in year 0. Since 1% all level. year I he mean in years P> years up anil The -1 and earnings changes here arc announced to four quarters subsequent to the dividend omission. 16 -1 industry-adjusted changes dividend omission date, these results indicate that there are significant earnings declines for in the These numbers are standardized by omission firms have mean standardized earnings changes of -10.3% and -10.0% -9.0% explore this changes are Ihc sum of earnings changes announced Dividend omission mean and median standardized changes are we redefine annual earnings so that they are strictly aligned with the dividend four quarters subsequent (prior) to the dividend omission respectively. results, analyze quarterly earnings data around the dividend omission date again, the stock price earnings earnings include quarterly earnings announced both before and after the dividend omission. announcement date in In summary, the above changes for up to results indicate that dividend initiating firms five years prior to and in the have positive earnings year of the dividend announcement; the dividend omitting firms exhibit negative earnings changes for up to two years before and the year of the dividend event. these firms' industries. They These patterns even after controlling for the patterns. Evidence on the post-announcement earnings patterns is earnings for these firms appears to be permanent. The the 3.2 wc find In contrast, the dividend omission firms have by these firms before and shortly after the post-announcement performance of the omission firms is Further, the earnings omissions announcement appears One temporary as indicated by the subsequent earnings recovery. This issue initiations, increased level of earnings declines only for one year after the dividend announcement. to be For mixed. announcement earnings growth for two years after the dividend decline experienced performance of are consistent with the proposition that these dividend decisions by systematic earnings are preceded persist possible explanation for the survival bias in our sample. discussed later in the paper. is Relation Between Dividend Information and Earnings Changes We next test whether the post-announcement earnings changes documented section are related to the market reaction to the omission. If announcement of the dividend changes A previous initiation or dividend policy changes are based on managers' expectations of future earnings, there will be a positive relation between dividend changes. in the simple regression framework in years to 4 is announcement returns ami subsequent earnings used to test this prediction for fiscal year earnings and annualized quarterly earnings changes year earnings are described below. in year Discussion of modifications to these quarterly earnings arc deferred to the results section. 17 0. Tests using tests for fiscal annualized examining the relation between the earnings changes and the market reaction In to the dividend announcement, our tests control for information on future earnings from sources other than the dividend announcement. First, Tables 3 and dividend initiation and omission firms deviates from a changes may 5 indicate that the random walk be used to forecast subsequent earnings changes earnings in year t-1 is included as an independent variable in earnings time scries of Therefore, prior earnings The standardized change the regression model in for year to t control for this earnings pattern. The second to the market we source of information on future earnings between the earnings announcement omission announcement. for year - 1 is and the dividend show Results reported in section 3.1 control for information released initiation or that there arc significant abnormal returns prior to the dividend announcements, indicating that information regarding the sample companies' future performance is released during this period. The market-adjusted return cumulated from one day subsequent to the earnings announcement to the dividend date We AF,: is is = B + B,DRETj + B 2 AI ; j ., t + B PRBTj + 3 the standardized earnings change for firm PRUT: is two days prior j in year Uj, j = I t, t = to 4: N (2) as defined in equation (1); t the cumulative market-adjusted return from one day subsequent to the earnings announcement for year initiation or \ to the market-adjusted return for one day before and the day of the dividend initiation or omission announcement; and i 1 used to proxy for this information. ( DRET: - estimate the following cross-sectional regression separately for each year AFj, where is for year -1 to two days prior to the dividend omission announcement. . . ... This information could be from prior quarterly earnings announcements or anticipation of the dividend news. IK The primary interest coefficient of equation (2) in omissions convey information about future earnings, In addition, if earnings changes in year Dt earnings, the coefficient Bi. If this coefficient dividend initiations and is positive can he used to forecast the change t-1 be non-zero. will is Finally, Dividend D-^ will in significant year information released prior to if dividend announcement but after the previous earnings announcement earnings performance, the coefficient and is t tin- related to subsequent be positive. initiation results Equation (2) estimated cross-sectionally, using the 131 dividend initiating firms, for each is of the five years subsequent to the dividend initiating announcement. reported in Panel A of Table and coefficients (Bi) in years The 7. 1 estimates for the dividend announcement return are 0.197 and 0.356 and arc 26 levels respectively using a two-tailed test. This evidence is significant at the year of and the year following the dividend initiation: dividend initiation implies a 0.2% and 0.36% a 1% 10% and 5% consistent with the hypothesis announcements convey information about that dividend initiation These estimates arc firms' earnings prospects in the abnormal price reaction to a increase in standardized earnings in these years number of observations used to estimate the regression in each year varies The number for each year is different from the number of observations reported by year in Table 3 for two reasons. First, we use earnings changes in two successive years in each regression. Second, wc require the annual earnings announcement The and is actual reported in Table 7. date prior to the dividend reduce the number of announcement to calculate PRET. These additional data requirements usable observations. z- White tests for hctcrosccdasticity arc not significant for any of the five regressions White (19R0) for a description of this test). Belslcy, Kuh and Wclsch diagnostics, which assess the effect of extreme observations on the regression coefficients, are also examined (see Rclsley, Kuh and Welsch (19R0) for a description of these diagnostics). The reported coefficient estimates do not appear to be influenced by extreme observations. (see Since the information hypothesis predicts the sign of the dividend return coefficient to be positive, a one-tailed coefficients for years in the tables are for a test is and 1 probably more appropriate. are significant at the more conservative 5% two-tailed 19 level. test. Under a one-tailed test the The significance levels reported . The coefficients for year 2 to 4 arc insignificant, indicating that dividend no information on earnings changes beyond year The coefficient estimate for the 10% significant at the level in the year negative and significant at the 1% change announcements convey 1 in earnings in year The estimated regression. fW t-1, is positive arid coefficient for year 3 Fstimatcs for other years are not significant. level. is Finally, the coefficient estimates for the market adjusted stock return, cumulated from the earnings announcement at the 1% in year -1 to the level in the year dividend initiation announcement (B^) regression. The estimates of R^ arc not is 0.12 and is significant significant for years 1 to 4. We between the earnings change also replicate the test of the relation in year and the market reaction to the announcement of the dividend policy change using annualized quarterly earnings. for Fquation (2) is estimated cross-scctionally for year which quarterly earnings data are return for firm is redefined as the market-adjusted cumulated from one day following the previous quarterly earnings j announcement PRET: available. for the 129 initiating test firms to two days prior announcement. to the dividend Tabic 8 reports the estimated regression coefficients. The estimate of the dividend significant at the 5% level using a two-tailed price increase at a dividend initiation standardized earnings year information do not return coefficient The test. announcement is 0.31 and estimate implies that a associated with a The earnings change measure for year the dividend is announcement. is statistically 1% abnormal 0.3% change in corresponds to Therefore, this finding provides Quarterly earnings announcements for 16 initiating firms arc reported on the same date as the dividend results 0. strictly released after 27 announcement announcement. Fquation from those reported. (3) is estimated after excluding these firms and the differ 20 further support for the hypothesis that dividend initiations provide information on subsequent earnings. Dividend omission results The parameters of regression equation (2) are estimated cross-scctionally using the 172 dividend omitting firms in the sample. years These estimates to 4. The and 1 Separate equations arc estimated for each of the are reported in Panel B of Table 7. estimates for Bi, the dividend omission return coefficient, arc 0.42 and 0.39 in years and are 5% significant at the 10% and levels respectively in a two-tailed test. findings are consistent with the information hypothesis: the estimates indicate that a unexpected decline the omission in price at announcement The estimate for year 3 is The in years 0, 3 the 1% and 4. This , The estimated For years is 1 and 2, the estimated coefficient is The is is not significant negative anil significant at consistent with the turnaround in the earnings performance of the The coefficient estimates for the cumulative market-adjusted return between the earnings announcement and the following coefficients for years 2 the coefficient of the earnings change in year t-1, dividend omission sample in these years. 2R announcement and significant. estimate for Bt level. 1% negative and significant, reflecting the earnings recovery subsequent to the dividend omission documented above. and 4 are not These accompanied by about a 0.4% is decrease in standardized earnings in the year of the dividend year. fiscal actual reported in Table number of observations used The number for each year 7. observations reported by year in Table 4. This is in year -1 and the dividend omission to estimate the regression in each year vanes is different from the number of due to additional dala requirements discussed earlier. Once again, White hctcrosccdasticity tests and diagnostics indicate that the regressions are well-specified. 21 Bclsley, Kuh and Welsch influence announcement, years 1 B-^, is and 0.13 in year is significant at the 5% The estimates of B3 for level. to 4 arc insignificant. Once again, we replicate the test of the relation between the earnings change in year and the market reaction to the announcement of the dividend policy change using annualized Equation 2 quarterly earnings. test rc-estimatcd cross-scctionally for year in firms for which quarterly earnings data are available." regression coefficients. and is statistically significant. is Table 8 reports the estimated estimate of the dividend announcement return coefficient The Since year omission for the 129 earnings arc announced strictly after is 0.39 the dividend date, this finding provides further support for the dividend information hypothesis. In summary, the indicate that there is results for both the dividend a positive relation announcements and earnings changes These year following. years, results are initiation and the dividend omission sample between the market reaction to the dividend for the year of the dividend policy change and for one obtained after controlling for the earnings changes in prior and information on future earnings available before the dividend announcement. consistent with the hypothesis that dividend initiations or omissions convey information They arc on future earnings performance. 3.4 Market Reaction to Earnings Announcements After Dividend Initiations/Omissions A number of accounting studies have documented that there announcement of unexpected reaction to the earnings.' We is a significant stock price use the following model to Quarterly earnings announcements for 30 omission firms are reported on the same date as the dividend results do not 31 Morse announcement. Equation from those reported (2) is estimated after excluding these firms and the differ See Ball and Brown (1968), Beaver, Clarke and Wright (1979) and Beaver, (1980). 22 lambcrt and represent the usual relation between earnings announcement returns and the si/e of unexpected earnings: ARETj where ARFTjj for firm j j + RijAi + ; ej t = -5 to 4. j = 1 to t j, N the market-adjusted return at the lime of the annual earnings is year in = D t and t AEu is (4) announcement unexpected earnings based on a random-walk earnings expectation model deflated by the firm's equity price two days before the earnings announcement. The parameters B n and Bj are the market reaction to unexpected earnings and assumed is to be firm specific. Bi is the elasticity of expected to be positive, consistent with the findings of the earlier studies. The results presented in the previous section announcements convey information on subsequent market to revise its show that dividend initiation or omission These announcements enable the earnings. expectation of earnings and thus reduce its measures of unexpected earnings based on a random walk model do not information earnings forecasts. in therefore, the elasticity of the model will be To ARET jt less B reflect this additional In years subsequent to the dividend initiation or omission, market reaction to unexpected earnings based on a random walk than "normal". test the = However, forecast errors. oj above prediction, we use the following modified form of equation + B,jAF jt +^^D i ti AEjt = 23 + e t jt = -5 to 4, j = 1 to N (4): (5) The parameters Uq to U4 arc cross-sectional average adjustments to Dj: five the elasticity of , the market reaction to unexpected earnings, in each of the five years following the dividend The initiation or omission. multiplicative dummy variable D.: takes the value following the dividend initiation or omission, and zero in other years. announcement leads the market to revise its Uq to The sample 114 will in years in year i the dividend policy forecast of subsequent earnings, earnings will be noisier estimates of unexpected earnings the parameters If one changes in to 4 than in -5 to -I, and be negative. distribution of estimated company-specific coefficients Bq: and Bj:. t statistics is used to test the significance For each parameter the following /. of the statistic is computed. N z = i/v^y^ i= where is t: is the t statistic for firm j /^kj/(kj-2) tj 1 associated with the estimate of the parameters B or Di; k: () the degrees of freedom in the regression for firm The sample. t statistic for firm Central Limit Theorem, the variance of N. The Z j is sum statistic for t test is used to test and distributed Student of the standardized each parameter the null hypothesis that the parameter (Bq or Bi) Student j; is is t t N is the number of firms in the with variance k:/(k:-2). statistics is Under the normally distributed with a therefore a standard normal variate under not significantly different from zero.' the significance of the parameters that arc A assumed constant across firms (uq to U4). For a detailed discussion of this test sec Christie (19S6). The test is based on the sample distribution of the parameter estimates. It is assumed that the parameters arc independent across firms in the sample. 24 Dividend initiation results The parameters of regression equation (5) are estimated jointly using ten years for the 131 dividend initiation coefficients Dq: and Bj:, estimate of U] The is 0.0039, and The sample mean of test. is -0.1280 and estimates of pj, of the regression is [ij, distribution of the estimated regression and the estimated values of the u coefficients arc shown The sample mean value of Bq two-tailed The test firms. observations over tlic is which 0.2894 and is U4 is the 5% are not significantly different These level Tabic level 1% also significant at the statistically significant. is 1% statistically significant at the statistically significant at M3. and 0.272, Dj is in using a zero. The level. using a two-tailed from 9. test. The adjusted R results arc consistent with the hypothesis that the magnitude of the market reaction to the earnings change during the one year following the dividend initiation is than the "normal' market reaction for a given level less J of earnings change.- Dividend omissions results The coefficients of regression equation (5) are estimated for the 172 dividend omission test firms using ten years' observations. These results are reported in values of Bq and Bi are 0.0038 and 0.2714 respectively. Table Both these The sample mean 10. values, which arc very similar to those obtained for the dividend initiation sample, are statistically significant at the level. The coefficient estimates for ug to and -0.1048. The estimates of Ug, Ui, estimate of \ia is significant at the u.4 arc, respectively, -0.1 148, -0.1 182, \ij, 5% and level u-^ are significant at the using a two tailed test. 1% These 1% -0.1323, -0.1 194, level, and the results are consistent with the hypothesis that the magnitude of the market reaction to the earnings change during the five years following the dividend omission dividend omission for a given level of earnings. which is statistically •" An significant at the 1% alternate explanation is The is less than the market reaction before the adjusted R of the regression is 0.296, level. that there is abnormally high noise to the dividend initiaton, reducing the earnings coefficient in these years. 25 in earnings subsequent The initiation year regression results for the dividend omission sample differ from those for the dividend sample in significantly negative for the dividend This is is |t significantly negative only in whereas the estimates of the for the dividend initiation sample, omission. The parameter one important respect omission sample in all five years |i parameters arc following the dividend noteworthy given that the earnings changes particularly omission sample change from negative to positive after year (I. The for the dividend negative value of u in later years for the dividend omission sample cannot be attributed to information revealed by the dividend omission announcement. the market receives possible explanation for the negative coefficients more non-accounting information than showing a dividend omitting firm starts earnings announcements 3.5 One a turnaround in its is that usual on the firm's performance once performance. If this is the case, years arc likely to convey less information than usual. in these Sample Selection Bias Our tests use earnings data for a subsequent to the date of the dividend requirement is initiation firms violated for one of and ten omission minimum initiation or five reasons: (1) of two years and a maximum omission announcement. new listing prior to the firms); (2) corporate control - of five years This data dividend date (three change subsequent to the dividend date (nine initiation firms and four omission firms); (3) dclistmcnl or suspension from the exchange subsequent to the dividend date (one Compustat coverage (14 Wall Street Journal. sample and 18% initiation firms- •); The of the initiation firm and (5) earnings and seven omission firms); announcement dates missing firms that arc excluded account for 22 initial (4) n n no in the of the original initiating omission sample. This requirement is imposed to ensure that we have sufficient observations to estimate firm-specific coefficients in the regression tests reported in Section 3.4. No omission firms are excluded for lack of coverage on Compustat since Compustat was initial sample used to generate the 26 The exclusion of the above companies for data availability reasons leads to a potential selection bias. If the reason for exclusion is related to the dividend initiation or decision, the bias systematically influences the reported results. initiation firms if The most firms and 17 omitting firms There is little Compustat or the Wall is mean earnings changes for the population. uncertain. mean change mean in earnings. frequent reasons for exclusion are that firms arc not covered on Compustat or have missing earnings annoucement dates firms the dividend the best performing firms are excluded from the omission sample, the sample earnings decline overstates the population reasons. il which arc excluded are the worst performers, the sample mean earnings increases surrounding the dividend announcement overstate Similarly, For example, omission It in the (65% and 45% Wall Street Journal Twenty-four initiating of those excluded respectively) are excluded for these evidence on the characteristics of firms that arc not covered by Street Journal. Therefore, the effect of the bias from excluding these seems plausible that Compustat and the Wall that they consider to be of interest to the investment community. Street Journal cover firms If, for some reason, the worst performing initiating firms and the best performing omitting firms arc considered uninteresting, then the bias increases the probability of finding results similar to those reported in the paper. New 37% listings prior to, and acquisitions subsequent to the dividend date account for .12% of the firms excluded from the initiation and omission samples respectively. excluding these firms is again uncertain: newly listed in the omission sample, our and acquired firms The one arc The effect they could cither be performing well or poorly. poor performers in the initiation If of the sample and good performers results are biased. systematic bias that can be identified of dclistments subsequent to the dividend date. It is 27 is for firms that have been excluded because likely that these firms are poor and Hence, the seven omitting firms (19% of the excluded performers. excluded for this reason lead to an understatement of the The omissions. mean initialing firms) which arc earnings decline surrounding earnings recovery documented for dividend omission firms may, in part, he Since only one initiating firm attributed to this bias. excluded for this reason, the (3% of the excluded omitting firms) of the bias on the initiating sample results effect likely to is is be small In summary, the unavoidable in studies effect which of the selection bias on our results rely on ex post time-series data, is uncertain. and suggests This problem is our results be that interpreted with caution. 4 SUMMARY AND The tests DISCUSSION reported in this paper examine whether there arc significant changes in firms' earnings performance surrounding cither a dividend initiation or omission and, if so, these changes are consistent with the market reaction to the dividend policy changes examine a sample of years, 131 firms that and a sample of 172 firms paying dividends for pay dividends that at least ten years for the first omit dividends for the Wc time or after a hiatus of ten first All the dividend policy whether time or after continuously changes examined occur between 1969 and 1980. The statistical tests and results presented in the there arc significant earnings increases for as announcements and announcements. many paper lead to four conclusions. as five years prior to dividend initiation significant earnings decreases for two years prior Second, firms have earnings increases for the year a dividend initiation; these increases First, appear to be permanent to dividend omission of, and two years following Firms that omit dividends have earnings declines for only one year prior to the dividend date; subsequently, the omission firms experience a recovery in earnings Third, the abnormal stock price reactions to the dividend 28 initiations or omissions arc correlated with the firms' earnings changes year subsequent to the dividend announcements. for prior earnings This relation found to changes and information available to the slock market dividend announcement is less for five years following at year of and one exist after controlling the time of the Therefore, dividend initiations and omissions seem to provide incremental information firms' future earnings performance. earnings changes is in the than usual in the Finally, the market reaction to year following dividend initiation announcements, and announcements of dividend omissions. Once again, this is consistent with the hypothesis that the dividend initiation or omission announcements anticipate subsequent earnings changes. The dividend initiation findings provide strong support for lintner's (1956) description of managers' dividend decision-making process, and the dividend information hypothesis proposed by Modigliani and Miller (1961). Managers appear performance when they decide to initiate to consider both past and future earnings Dividend cash dividends. initiation decisions arc therefore interpreted by the market as managers' forecasts of future earnings increases. dividend omission results are The While firms appear to have earnings declines less conclusive. surrounding the dividend omission, most of the declines occur before the dividend announcement. The subsequent earnings declines arc short-lived and arc quickly reversed. There are two caveats to the interpretation of our post earnings time-series data, the results Second, while there is about future earnings, to communicate One may have been results. First, since our tests use ex influenced by a sample selection bias evidence that dividend initiation and omission decisions arc informative this docs not necessarily imply that managers make these decisions solely their earnings forecasts. possible extension of this paper is to examine earnings changes surrounding unexpected 29 dividend increases and decreases.initiations large and omissions. announcement As noted earlier, our sample comprises only dividend These arc relatively rare changes returns, thereby increasing the a dividend policy. power of our representative of the population of dividend policy changes. whether our findings can be generalized to in It tests, they would be may While they have not be interesting to examine wider class of dividend policy changes. Ofcr and Siegel (I9R7) examine a sample of unexpected changes in dividends. report evidence of revisions in analysts' forecasts of earnings subsequent to these dividend changes which arc consistend with our findings. JO They REFERENCES Aharoncy, and J. An returns: Swary, 19R0, Quarterly dividend and earnings announcements and stockholders' I. empirical analysis, Journal of Finance 31, 1-12 D.W. Mullins, 1983, The impact of initiating dividend payments on shareholders' Journal of Business 56, 77-96. Asquith, P. and wealth, and P. Brown, 1968, An empirical analysis of accounting income numbers, Journal of Accounting Research 6, 159-178. Ball, R.J. Some and R.F. Watts, 1972, Finance 27, 663-682. Ball, R.J. Beaver, W.H., R. Clarke, and W. time series properties of accounting income, Journal of Wright, 1979, The association between unsystematic security returns Journal of Accounting Research 17, 316-340. and the magnitude of earnings forecast errors, Beaver, W.H., R. Lambert, and D. Morse, 1980, of Accounting and Fconomics Kuh and Belsley, D.A., E. The information content of security prices, Journal 2, 3-28. Welscli, 1980, Regression diagnostics, (John Wiley and Sons, R.F. New York). Bhattacharya, S., 1979, Bell Journal of Bhattacharya, S., Fconomics Bricklcy, J., Imperfect information, dividend policy, and 'the bird in the hand' fallacy, Fconomics 259-270. 10, 1980, Nondissipative signalling structure 1983, Shareholder wealth, information signaling and the specially designated dividend, Journal of Financial Fconomics Christie, A., 1986, Aggregation of Diclman, T.E. and II. 187-209. An evaluation of the consistency of evidence on Unpublished paper, University of Southern California. An examination of investor behavior during periods of Journal of Financial and Quantitative Analysis, June, 197-216. R. Oppcnhcimcr, 1984, large dividend changes, II. 12, test statistics: contracting and size hypotheses, Fama, F.F. and and dividend policy, Quarterly Journal of 95, 1-24 Babiak, 1968, Dividend policy: Statistical Association, 63, An empirical analysis, Journal of the American 1132-1161. Foster, George, 1977, Quarterly accounting data: Time-series properties and predictive-ability results, Accounting Review 52,1-21. Goncdes, N.J., 1978, Corporate signaling, external accounting, and capital market equilibrium: Evidence on dividends, income and extraordinary items, Journal of Accounting Research 16, 26-79. John, K. and J. Williams, 1985, Dividends, dilution and taxes: Finance, September, 1053-1070. Kalay, A. and U. Lowenstein, 1985, A signalling equilibrium, Predictable events and excess returns: announcements, Journal of Financial Fconomics 31 14, 423-450. Journal of The case of dividend I intner, J., taxes. 1956, Distribution of incomes of corporations among dividends, retained earnings, and American Economic Review, 46, 97-1 13. Mil, and K. Rock, 1985, Finance 40, 1031-1051. Miller, Ofer, A and D. Sicgel, 1987, Dividend policy under asymmetric information, Tests of rational signalling application to dividend signalling, models using expectations The Journal of data: An Journal of Finance, forthcoming. and W. I hakor, 1985, A theory of price responses to alternative cash disbursement methods: Stock repurchases and dividends, Unpublished paper, Northwestern University. Ofer, A. Watts, R., 1973, The information content of dividends, Journal of Business 46, 191-21 and R. Leftwich, 1977, The time Accounting Research 15, 253-271. Watts, R. White, llalbcrt, 1980, A scries 1. of annual accounting earnings. Journal of hctcroscedasticity-consistcnt covariance matrix estimator and a direct for hetcroscedasticitv, Fconomctrica 48, 817-838. 52 test Table Number of sample firms dividends by year in 1 and omitting initiating the period 1 970-1 979 a b . Companies initiating Number Year dividends Percent Companies omitting dividends Number Percent 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 131 Total a 100.0 172 100.0 dividend payments in a firm's history or dividend resumptions after a Dividend omissions are first time eliminations by firms that paid dividends continuously throughout their history or for at least ten years. Dividend initiations are first hiatus of at least ten years. D To be included in the sample, firms are required to meet the following requirements: (1) be listed NYSE or ASE; (2) have dividend initiation/omission announcement dates available in the Wall on the have stock price data available for two days before the dividend the CRSP daily Master Tape or Standard and Poor's Daily Stock Price Record; (4) have stock returns available on the CRSP data files for the day before and the day the dividend initiation/omission announcement; (5) have annual earnings per share before extraordinary items and discontinued operations available from the 1984 Compustat for 8 of the 11 years surrounding the dividend initiation/omission announcement date; and (6) have earnings Street Journal index; initiation/omission (3) announcement on announcement dates available in the Wall Street Journal Index for these same years. of Table 2 Abnormal returns dividend initiating and 172 dividend omitting firms for selected holding periods surrounding the public for 131 announcement date (t statistics in Dividend period b Holding PD-60 to PD-20 PD-10 to to PD-21 PD-11 PD-1 Dividend firms initiating omitting PD to (4.8) (-4.0) 1.1 -2.7 (2.7) (-3.1) 4.0 -7.0 (-8.0) -9.5 3.9 (-24.8) (15.4) PD+1 PD+11 to PD +10 to PD+20 firms -7.0% 3.5% (10.0) PD-1 parentheses) 3 1.4 -1 .2 (3.6) (-1.4) 0.6 -0.5 (1.4) (-0.5) a Abnormal returns prior surrounding the dividend announcement are market-adjusted returns using CRSP equal-weighted market returns. The sample firms initiate/omit dividends in the period 1969 to 1980. D PD is the date the dividend initiation or omission is announced in the Wall Street Journal. Table 3 Summary for statistics dividend Period relative to dividend initiation Panel A: Year Raw on changes initiating firms earnings per share as a percent of equity price years surrounding the dividend announcement 3 ^ in Number of firms earnings changes -5 in Student Mean Wilcoxon t probability Median probability Table 4 Summary of statistics equity price on annualized quarterly earnings changes as a percent dividend initiaition firms in years surrounding the for dividend announcements 3 Period relative to dividend initiation Number of firms Student Mean ' 13 Wilcoxon t probability Median probability Panel A: raw earnings changes Year -1 129 129 5.44% 0.01 3.73% 0.01 4.51 0.01 3.36 0.01 0.05 1.34% 0.01 1.72 0.03 Panel B: industry-adjusted earnings changes Year -11 118 2.85% 3 Earnings changes are estimated using quarterly earnings that are announced in the eight quarters prior to the dividend announcement and the four quarters subsequent. Changes in earnings per share before extraordinary items and discontinued operations for each firm are standardized by its stock price two days prior to the dividend announcement. b The dividend initiation sample comprises 129 firms that announce dividend initiations in the period 1970 to 1979. Industry-adjusted earnings changes are available for 118 of these firms. They are defined as the difference in standardized earnings changes for the initiation firms and for matched comparison firms that are randomly selected from the c same Student zero. The t industry. and Wilcoxon statistics test the hypotheses that the mean and median probability levels reported are for two-tailed tests of significance. earnings changes are different from Table 5 Summary for statistics on changes dividend omitting firms Period relative to dividend omission -5 ' Number of firms Panel A: raw earnings changes Year in earnings per share as a percent of equity price years surrounding the dividend announcement 3 6 in Student Mean Wilcoxon t probability Median probability Table 6 Summary statistics of equity price on annualized quarterly earnings changes as a percent dividend omission firms in years surrounding the for dividend announcements 3 Period relative to dividend omission Number of firms Panel A: raw earnings changes Year 1 Student Mean ' 6 Wilcoxon t probability Median probability Table 7 relation between standardized changes in earnings announcements of dividend initiations/omissions, and the dividend announcement return (t statistics in parentheses) 3 Tests of the following AEj, = Period relative to dividend announcement Panel A: dividend firms initiation 1 123 2 123 3 121 4 120 p,DRET j+ 161 1 159 2 158 3 152 4 141 [5u sample Panel B: dividend omission sample Year + P 2AE j|t .i + p 3 PRET, + ej,° Number of 124 Year (5 0.029 Pi Pi P3 R2 Table 8 Tests of the relation between standardized changes in annualized quarterly earnings one year following announcements of dividend initiations/omissions, and the dividend announcement return (t statistics in parentheses) 3 AE j0 = p + P,DRET j + p 2 AEj,.-, + p 3 PRETj + eJt ° Number firms of Dividend p R2 P3 P2 Pi sample initiation 129 0.029 0.315 -0.178 0.123 (2.85)C (2.49)d (-2.11)d (1.98)6 0.086 Dividend omission sample 140 3 The sample comprises 129 announced dividend omissions -0.079 0.394 (-3.42)C (2.26)0 firms thai -0.008 -0.062 (-0.40) announced dividend 0.032 (-0.51) initiations in the period 1970 to 1979, and 140 firms that the period 1969 to 1980. Annualized earnings are estimated using earnings for the eight quarters prior to the dividend announcement and the four subsequent earnings. D AEjt is firm j's announcement; change DRET; in is (omission) announcement; in earnings standardized by its stock price two days prior to the dividend the market-adjusted return for firm and PRETi is j for one day to the market-adjusted return for firm and the day j initiation from one day following the quarterly earnings announcement immediately prior to the dividend announcement to two days prior to the dividend date. c Significant at the 1% level using a two-tailed test. d Significant at the 5% level using a two-tailed test. e Significant at the 6% level using a two-tailed test. (omission) of the dividend initiation Table 9 Tests of the relation between unexpected stock returns earnings announcements and standardized changes in earnings for years surrounding initiation of at dividends 3 4 ARET jt = p oj +p AE + I M^D T tAEj t +e b jt jt Ij T=0 Po Mean 0.0039 Z statistic t statistic 0.0003 0.0013 0.1418 0.0194 0.3738 Adj. a for quartile R2 The M2 Mi m M3 -0.1280 -0.0422 -0.0957 -0.0223 -0.0134 -2.04 s -0.64 -1.49 -0.45 -0.27 7.26 d -0.0092 Median Third 0.2894 10.79 d quartile First Mo Pi 0.272 d results for coefficients (3q 131 firms that initiate and dividends in (3-| are for the cross-sectional distribution of time-series regression coefficients The the period 1970 to 1979. coefficients u. t (t=0 4) are assumed to be constant across firms. D ARETij is the market-adjusted return for announcement; aEjj is the change earnings announcement; and D^ in is a one day prior to earnings per share dummy in and the day year t variable that takes the value announcement and zero otherwise. c Under the d Significant at the 1% level using a two-tailed test. 9 Significant at the 5% level using a two-tailed test. null hypothesis, each Z statistic of is the Wall Street Journal annual earnings standardized by the stock price two days prior to the distributed unit normal. one t years following the dividend initiation Table 10 Tests of the relation between unexpected stock returns earnings announcements and standardized changes in earnings for years surrounding omission of at dividends 3 4 ARET jt = p oj +p AE lj jt + I ^D Tl AE jt + Ejl b T=0 Pfj Mean Pi Ho Hi H2 H3 H4 Date D I MRn' Lib-26-67 MIT LIBRARIES illinium 3 inn 111 111 1 ii iii inn =1060 DOM 134 243