تاثیر اینترنت بر اعلام سود

This study examines the impact of the Internet on earnings announcements, in an Australian context, with relation to (1) information content (2) post-earnings announcement drift (PEAD) and (3) information asymmetry. Our results for the pre-Internet (1996) and the post-Internet (2000) periods provide mixed evidence on the changes in the information content of earnings announcements. However, we detect a reduction in both the magnitude and duration of the post-earnings announcement drift in the post-Internet period for the positive unexpected earnings decile. The analysis on abnormal volume, spread and market depth do not show a significant change in information asymmetry between the pre- and post-Internet periods for preliminary final announcements.

The body of empirical studies investigating the relationship between earnings announcements and share prices has been both voluminous and diverse. Previous literature, beginning with the seminar paper by Ball and Brown (1968), has placed much emphasis on the importance of earnings announcements due to its implications on the performance of a firm. The increased popularity of the Internet and online trading (introduced in 1997) has brought about a change in the investment environment. The influx of Internet websites providing financial content and investment-related services allows investors access to large volumes of timely, inexpensive information while simultaneously reducing trading costs. In view of the changes in the investment environment, share price reaction and trading around earnings announcements are likely to be different. Several papers have recently examined the share price reaction to and information content of earnings announcements over time. Johnson and Schwartz (2000) provided evidence of a decrease in the profitability of the post-earnings announcement drift (PEAD). Others such as Landsman and Maydew (1999) and Kim and Kross (1999) have found an increase in information content of earnings announcements over time. These papers, among others, have found that various aspects of earnings announcements have been altered with the change in the economic environment. Several explanations have been provided for these changes, but one that has been widely discussed is the change in information dissemination brought about by the improvements in technology. The decrease in the cost of computing and communications has provided market participants access to vast amounts of information, possibly diminishing the importance of accounting information (Landsman and Maydew, 1999). In addition to the possible improvements in information dissemination, which reduces information asymmetry, the Internet can affect investors' reaction to earnings announcements in several other ways. Recent papers have explored how the Internet impacts the behaviour of investors (Barber and Odean, 1999 and Choi et al., 2000). Others have examined the use of the Internet as an information dissemination tool (Dewally, 2000 and Wysocki, 1998). The empirical studies to date provide a consensus view that the Internet has had an impact on investors. This paper attempts to examine the impact that the introduction of the Internet has had on earnings announcements in Australia. Three related questions are asked: (1) Has the introduction of the Internet led to a decrease in the information content of earnings announcements? (2) Is there a reduction in the PEAD with the change in investment environment brought about by the Internet? (3) Is there an improvement in information asymmetry around an earnings announcement with the introduction of the Internet? A motivating factor for this study is the concern that regulatory bodies such as the Australian Securities and Investment Commission (ASIC) have regarding the growth in online trading and the corresponding increase in securities trading by retail investors (ASIC, 2000). Many claim that the Internet can have an adverse effect on these investors by increasing the amount of noisy information that they receive. Earnings announcements provide an ideal scenario to examine the possible effects generated by the Internet because of their regular and periodic nature. Furthermore, investors' reactions to earnings announcements have also been widely documented, and previous results provide a point of reference for comparisons. This paper builds on a recent paper by Ahmed et al. (2003) that has examined how online trading affects investors' reaction to earnings announcements. This study enhances their work by examining a wider aspect of earnings announcements, combining traditional capital market research methods with market microstructure research to investigate changes around the earnings announcements. The use of market microstructure measures, such as bid–ask spread and depth, will provide a greater insight by allowing us to observe changes within the trading day, which may have been obscured using daily closing price data. In contrast to Ahmed et al. (2003), our study takes place in an Australian context. Differences in investors' reactions may be observed as the Australian market operates in a different institutional setting. The Australian Stock Exchange (ASX) is an order-driven environment unlike the quote-driven environment in American markets. Australia was also slower in its introduction of online trading. The late introduction might lead to a faster rate of adoption of online trading as Australians attempt to catch up with their American counterparts, thus providing us with clearer results. The remainder of the paper is structured as follows. Section 2 reviews the relevant literature and examines possible ways in which it would change how investors react to earnings announcements. In Section 3, we develop the testable hypotheses for the study. Section 4 discusses the relevant research methods and Section 5 describes the data necessary for the study. Section 6 discusses the results of the data analysis and explores its implications. The last section provides a summary of our conclusions.

The aim of this study is to assess the impact of the Internet on the various aspects of earnings announcements. Our first set of tests examined the changes in information content ofearningsannouncements using abnormal volume andabnormalvolatility measures. There was some evidence of a decrease in abnormal volume and an increase in abnormal volatility with the multiple factor regression tests. This finding is divergent from previous studies, including Landsman and Maydew (1999) , who showed an increase in both metrics over time, and Ahmed et al. (2003) , who showed the same after the introduction of online trading. Interestingly, several of the changes in the preliminary and interim earnings announce- ment subsamples were in opposing directions. Interim announcements were shown to have increased in abnormal volume and abnormal volatility reaction, while preliminary final announcements exhibited greater price reaction, but lesser volume reaction, with the introduction of the Internet. A plausible explanation for the divergent finding is that interim announcements have greater information content in recent periods. McNichols and Manegold (1983) and Shores (1990) suggest in their studies that market reactions to annual earnings announcements decline as interim information increases. Interim announcementsthus provide guidance for investors to form their judgment on the firms’ earnings for the year. This, together with the increase in availability of other forms of nonearnings information on the Internet, reduces the informativeness of the preliminary final earnings announcement. The results in our study corroborate the findings of Johnson and Schwartz (2000) , who showed adecreaseinthe PEAD over time. Both univariateandmultiple factorregression test results were strongly in support for our alternative hypothesis. In contrast to both our findings and that in Johnson and Schwartz (2000) , Ahmed et al. (2003) showed an increase in the PEAD in a period with online trading. Their test of the drift, however, was conducted for a [+2,+5] event window. This short 4-day event window does not fully encompass the full extent of the drift. As we showed in Section 6.2, the drift for the pre-Internet period lasted for the whole 180-day event window. Previous empirical literature on the PEAD, such as Bhushan (1994) , showed that the magnitude of the drift is positively correlated with the degree of trading frictions, which makes attemptstoexploit the drifteconomically less attractive. With onlinetrading,the costs of trading have substantially dropped. Our findings thus suggest to us that the Internet has indeed changed the dynamics of the investing environment by improving the dissemination of earnings announcements and lowering trading costs; all these have allowed investors to take advantage of the drift. This, in turn, has led to a reduction of both the magnitude and duration of the PEAD. Bartov et al. (2000) examined whether the magnitude of the drift is decreasing in investor sophistication, as proxied for by the extent of institutional ownership in a stock. They conclude that the market fails to recognize the autoregressive earnings property only for firms that have relatively less predisclosure information (i.e., small firms with relatively unsophisticated investors). The increase in investor sophistication, although questionable, would be possible if retail investors are able to utilise the increase in information availability brought about by the Internet adequately. This explanation, however, would require further work to be substantiated. Finally, our tests of information asymmetry around earnings announcements did not show evidence of a reduction in information asymmetry brought about by the Internet. The Internet has provided retail investors with tools such as real-time order book information and company announcements. Again, the ability of retail investors to process utilise the information is debatable. To further examine this issue, individual trades can be examined to establish the profile of the investors that trade after an earnings announcement with the changes in technology. In conclusion, our findings suggest that the Internet has undeniably led to changes in how investors react to earnings announcements. The improvement in the trading environment has brought about a more level playing field for retail investors. This finding lends support to a survey on the online trading industry in Australia by ASIC (2000) , which concluded that the online trading industry in Australia is effective and efficient. While we have attempted to control for changes in other factors that could affect our inferences and our event (introduction of Internet) occurs simultaneously for all firms, we can never completely rule out that our results are driven by some other contemporaneous economic event unrelated to the Internet. This problem is endemic to most empirical studies that evaluate such changes. Also, it is possible that the new equilibrium induced by this change is yet to stabilise. Onlyfuture research can assess whether a new equilibrium is induced and, if so, whether our study predates it. Further insights to the impact of the Internet could be obtained by examining the change in the profiles of traders after an announcement using the Clearing House Electronic Sub-register System (CHESS) database, which captures the individual shareholdings on a daily basis 37 . This would complement this current paper that only examined market depth and order flows on an aggregate level