Price reaction to earnings announcements a study of vietnam stock market

  DƯƠNG THÚY AN PRICE REACTIONS TO EARNINGS ANNOUNCEMENTS: A STUDY OF VIETNAM STOCK MARKET MASTER THESIS Ho Chi Minh City – 2011... The study investigates the price reactions to a

 

DƯƠNG THÚY AN

PRICE REACTIONS TO EARNINGS

ANNOUNCEMENTS: A STUDY OF VIETNAM

STOCK MARKET

MASTER THESIS

Ho Chi Minh City – 2011

 

DƯƠNG THÚY AN

PRICE REACTIONS TO EARNINGS

ANNOUNCEMENTS: A STUDY OF VIETNAM

Ho Chi Minh City – 2011

ACKNOWLEDGEMENT

Firstly, I would like to express my profound gratitude and deepest appreciation to my supervisor, Ph.D Trương Tấn Thành, for his intensive support, precious guidance, insightful comments as well as constructive direction throughout my thesis Due to his support, I can overcome many problems during doing this thesis

I’m very grateful to nvtrungkr (an admin of forum http://vikool.org) for his help in finding many valuable journal articles Special thanks to Vu Khanh Hoang for his assistance in collecting the data

I would like to express my appreciation to all instructors at Faculty of Banking and Postgraduate Faculty, University of Economics Ho Chi Minh City for their support and valuable knowledge during my master course

Finally, I wish to thank my beloved family for their boundless support, abundant love and encouragement in my life

ABSTRACT

A substantial number of empirical studies has investigated the information content of earnings announcements While most of studies focus on United of States data or data from other developed countries such as United of Kingdom, Europe, Australia etc, other emerging and developing markets are largely unexplored This motivates us to further explore this issue in Vietnam stock market The study investigates the price reactions to annual earnings announcements on the Vietnam stock market We conduct

an event study method on a sample of 109 listed firms of Ho Chi Minh Stock Exchange (HOSE) over the two fiscal years (2009 – 2010) to examine stock price reactions to annual earnings announcements We find that significant abnormal returns before earnings announcements date It provides evidence of informational value of earnings information release The results of this study also provide that stock prices respond more strongly to negative than positive earnings surprises Our findings are robust across several sensitivity analyses

Keywords: price reactions, earnings announcements, event study, Vietnam stock

market

Table of Contents

LIST OF ABBREVIATION vii

LIST OF FIGURES viii

LIST OF TABLES ix

CHAPTER 1: INTRODUCTION 1

1.1.Background 1

1.2.Research objectives and research questions 3

1.3.Research methodology and scope 3

1.4.Structure of the thesis 3

CHAPTER 2: LITERATURE REVIEW 5

2.1.Informational effects 5

2.2.Stock price reactions to earnings announcements 5

2.3.Post Earnings Announcements Drift 9

2.4.Other reactions to earnings announcements 10

2.4.1 Trading volume and earnings announcements 10 2.4.2 Liquidity, information asymmetry and earnings announcements 11 2.4.3 Volatility and earnings announcements 13 2.5.Information disclosure regulation on listed firm in Vietnam 13

2.6.Summary 15

CHAPTER 3: DATA AND METHODOLOGY 16

3.1.Data 16

3.1.3 Sample description 17

3.2.Methodology 18

3.2.1 Descriptive statistics 18 3.2.2 Event study 19 3.3.Hypotheses development 22

CHAPTER 4: DATA ANALYSIS AND FINDINGS 24

4.1.Descriptive statistics 24

4.2.Results of the price analysis 25

4.3.Results of price response to unexpected earnings 30

4.4.Sensitivity analysis 32

4.4.1 Concurrent disclosures 32 4.4.2 Rank test 34 CHAPTER 5: CONCLUSIONS AND REMARKS 37

5.1.Conclusions 37

5.2.Implications 37

5.3.Limitations 38

5.4.Recommendations 38

REFERENCES 39

APPENDIX 44

LIST OF ABBREVIATION

LIST OF FIGURES

Figure 3.1 Time line of event study 19Figure 4.1 Plot of cumulative abnormal return for earnings announcements from day -20 up to day 20 28Figure 4.2 Plot of cumulative abnormal return for earnings announcements from day -20 up to day 20 31

LIST OF TABLES

Table 3.1 Sample selection 17Table 3.2 Sample distributions by timing of announcement 17Table 3.3 Firm sample distributions by sector 18Table 4.1 Descriptive statistics for stock price, stock return, and market value in event window 24Table 4.2 AARs with their t-test around event window [-20; 20] 25Table 4.3 Cumulative abnormal returns of four periods: pre-event period, announcement day, post-event period and whole period 27Table 4.4 AAR with their t-test for positive and negative portfolios 30Table 4.5 AARs with their t-test around event window for sample without major concurrent disclosures 33Table 4.6 Rank test around event window [-20; 20] 35

CHAPTER 1: INTRODUCTION

1.1 Background

Companies that trade stocks in the public financial markets are required to report their financial statement Earnings announcement1 is an official public statement of a company's profitability for a specific time period, typically a quarter or a year

Earnings announcements play an important role in the stock market Earnings statements provide much useful information to shareholder, investor, and analyst It reveals how healthy a company’s operation It is one of main factors influenced the value of its stock Investors are dependent on information disclosure of company to determine whether buying or selling stocks of that company

A large body of researches in finance and accounting studies the impact of earnings announcements on financial markets Specifically, many researchers have provided evidences on the market response to earnings announcements particularly in developed countries For example: Ball & Brown (1968), Beaver (1968), Morse (1981), May (1971), Kiger (1972), Beaver, Clarke & Wright (1979), Morse (1981), Patell & Wolfson (1984), Foster, Olsen & Shevlin (1984), Jennings & Starks (1985), Bamber & Cheon (1995) They find the effect of earnings announcements on stock price, liquidity, volatility, and information asymmetry

Vietnam stock market was established lately in 2000, but it has been a rapid growth It has two stock exchanges: Ho Chi Minh Stock Exchange (HOSE) and Ha Noi Stock Exchange (HNX) HOSE was firstly established in July 28th 2000 with 2 listed firms

At the end of 2010, there are 275 listed firms in HOSE The total capitalization value

on HOSE at the end of 2010 has reached 591.345 billion VND, 19.4% higher than of

2009 The world financial crisis in 2008 has been impacted strongly the performance

1

According to definition of website www.investopedia.com

of Vietnam stock market The trading volume of 20102 has reached more than 11 billion securities, equivalent to a value of VND 362,000 million, a 6% rise in volume but an 11% fall in value compared to those of 2009

Previous researches on the information content of earnings announcements mostly focus on developed countries such as United States, United Kingdom and other countries in Europe To date, more and more studies provide new evidences from developing and emerging markets For example: Fan-fah, Mohd and Nasir (2008) in Malaysia, Alzahrani (2010) and Alzahrani & Skerratt (2010) in Arab Saudi, Mlonzi, Kruger & Nthoesane (2011) in South African However, there is lack of empirical evidence about this issue in Vietnam stock market Trần (2010) has investigated the efficiency of stock market to dividend announcements of three listed firms: STB, VNM, and FPT in Vietnam This motivates us to analyze the stock price reactions to annual earnings announcements in Vietnam stock market

The thesis contributes to existing literature and reality in several aspects as follows: First, the study explores the stock return reactions to annual earnings announcements

on the Vietnam stock market By using daily data, the thesis provides new empirical evidences in the Vietnamese context that has some distinguished characteristics with other markets such as order – driven trading method, no derivative products, forbidden short-selling, low liquidity and lack of analysts’ forecast We employ standard event study method to investigate the effect of the earnings event

Secondly, our findings may be useful caution for listed firms in disclosing accounting information in practice Earnings information is an important issue that largely effect price behaviour of investors Thus, listed firms should put more attentions to information disclosures

Last but not least, the results of this study also provide evidence for policy makers to enforce strictly laws on earnings information disclosure and enhance corporate governance of listed firms

2

According to HOSE’s annual report in 2010

1.2 Research objectives and research questions

Research objectives

This study is conducted to investigate the association between stock prices and annual earnings announcements on the Vietnam stock market in short-term horizon On the other hand, we examine whether the earnings information releases possess informational value

Research questions

The research objectives defined above lead to the following research questions:

 Do stock prices respond to annual earnings announcements before and after

release?

 How quick are stock price reactions to annual earnings announcements?

1.3 Research methodology and scope

The object of this research is all listed non-financial companies in Ho Chi Minh Stock Exchange (here after HOSE) of period from 2009 to 2010 Our sample is unbalanced panel data, including 211 annual earnings announcements

Standard event study is employed to test the relationship between stock prices and annual earnings announcements Market model is used to estimate expected return This model was widely used in previous studies (Beaver, 1968; Bhattacharya, Daouk, Jorgenson, & Kehr, 2000; Chen, Firth, & Gao, 2002; Cheung & Sami, 2000; Morse, 1981; Sponholtz, 2008)

We use Stata 11 as a data analysis tool to implement the study

1.4 Structure of the thesis

The layout of the thesis is organized as follows:

 Chapter I introduces background, motivation and objectives of this thesis

 Chapter II discusses a literature review of the market reaction to earnings announcements

 Chapter III performs a general task to describe the sample coverage, data

selection process, and methodology approach used in this thesis

 Chapter IV presents the empirical results and explanations

 In conclusion, Chapter V summarizes the findings, implications and put forwards recommendations for further research

CHAPTER 2: LITERATURE REVIEW

The purpose of this chapter is to summarise reviews and discuss the literature related

to the information contents of earnings announcements and the links with our study Specifically, we mainly focus on the relation between stock prices and earnings announcements Moreover, we also briefly review other reactions to earnings announcements As the theory and empirical work in this area is so vast, we only provide a review of a few most relevant ones for our limited scope

2.1 Informational effects

Accounting information plays an extremely important role for all market participants

in stock market Market – based accounting research has developed fast since 1960s It investigates the relation between published accounting information and individual investor’s behaviours as well as characteristics of stock such as share prices stock return, trading volume (Lev & Ohlson, 1982) Since then, numerous studies in literature have documented the relationship between market reactions and accounting reports Obviously, accounting data conveys information content to investors

According to Fama (1970), in an efficient market, prices fully reflect all available information He proposes three subsets of efficient market: weak-form with prices just reflect historical prices; semi-strong form with prices adjust to all publicly available information and strong-form with prices absorb all information relevant Among variety of accounting information, earnings are one of the most important source for investors and analysts

2.2 Stock price reactions to earnings announcements

The seminal paper by Ball & Brown (1968) is the first to investigate empirically the information content of earnings announcements Employing the monthly data, they examine the association between accounting income numbers and stock prices The sample consists of 261 firms on the New York Stock Exchange covering the period

1957 - 1965 They also divide earnings announcements into two types: bad news if

income forecast error is negative, good news if income forecast error is positive Applying Abnormal Performance Index (API), they show that there are the relationships between the sign of income forecast error and stock return residual Surprisingly, they also report that the drift of API persists for two months after the announcements Later on, this phenomenon is namely Post Earning Announcement Drift

In the same year, Beaver (1968) uses weekly dataset of New York Stock Exchange to study the information content of annual earnings announcements from 1961-1965 He employs Sharpe model to remove the effect of market-wide events upon the stock price change He finds that volume and price changes are higher than average in the weeks of announcements In addition, he realizes market responses very quickly such as finding

of Fama et al (1969)

In a theory article, Kim & Verrecchia (1991) show that the price reaction at the time of announcements is proportional to both the unexpected portion of the announcement and its relative importance across the posterior beliefs of traders They explain that the relative importance is increasing with the precision of the announcement and decreasing in the precision of the preannouncement information

Since 1960s, a large body of empirical researches on stock market reaction to earning disclosure has been conducted in the context of United States such as findings of May (1971), Kiger (1972), Beaver, Clarke & Wright (1979), Morse (1981), Patell & Wolfson (1984), Foster, Olsen & Shevlin (1984), Jennings & Starks (1985), Bamber & Cheon (1995)

May (1971) conducts empirical study to test the influence of the quarterly earnings announcements on stock price changes Using weekly dataset of American Stock Exchange from 1964-1968, he provides evidence that stock price responds significantly with quarterly earnings announcements within weeks of disclosure than non announcement weeks In addition, he also finds that stock price response to annual earnings disclosure is greater than response to quarterly earnings announcements Following May (1971), Kiger (1972) works on a sample of 87 firms in New York Stock Exchange during the period 1966 – 1969 Using correlation analysis, he also

documents the positive correlation between adjusted price changes and hypothesized price changes surrounding the announcement date of interim report

Applying Spearman rank correlation, Beaver, Clarke and Wright (1979) indicate a positive relationship between unsystematic security returns and the magnitude of earnings forecast errors Their results base on two earning forecast models: one based

on the process of a Martingale with drift and one assumed that expected earnings have

a linear relation with average earnings of market

The subsequent study by Morse (1981) on daily data reports large price changes on the day prior, the announcement day and two days following the quarterly earnings announcement in The Wall Street Journal from 1973 – 1976 His findings imply it seems to take several time to absorb and adjust price at the individual and portfolio level He also finds no considerable differences for both exchange securities portfolio and over the counter securities portfolio

Using intraday data, Patell & Wolfson (1984) test mean return, return variance and serial correlation of stock price to examine the speed of stock price adjustment to earnings and dividends announcements occurring from 1976 - 1977 Quarterly and annual earnings expectations are Value Line Investment Survey forecasts Their results show that there are quickly reactions to disclosure Stock returns are positively detected

in the announcement day, the overnight period and the opening of trading on the next day, particularly at 30 minutes before and after announcement For dividends announcements, the stock price response is similar but weaker than earnings announcements

Empirically evidences of price reactions to earnings announcements can be witnessed not only in United State data, but also in other markets

In the context of Australia, Brown (1970, according to Cheung & Sami (2000)) provides evidence that there is more price reaction to annual and semiannual earnings announcements in the announcement month This evidence can be explained by some reasons: Australian firms issue semiannual reports instead of quarterly reports and the size of the Australian firms are much smaller than NYSE firms

In a working paper, Louhichi (2004) uses intraday data to investigate market reaction around earnings announcements in the Paris Bourse during the period 2001 – 2003 Based on the difference between actual earning and financial analysts’ expectations, the sample is split into three categories: good news, bad news and no news Applying event study method, he finds that returns react positively to good news and negatively to bad news In addition, returns do not react significantly to no news He also offers evidence that Euronext Paris is efficient at semi-strong form level

By replicating Morse (1981), Cheung & Sami (2000) investigate the price changes and trading volume reactions to annual earnings releases of 191 firms listed on the Stock Exchange of Hong Kong (SEHK) over the years 1992 to 1995 Their results materially confirm the findings of Morse (1981) Price changes are observed significantly within four days starting the day of announcements Expanding Morse (1981), they also find that price reactions of non blue chip stocks prolong over time than reactions of blue chip stocks This implies that analyst and investors give less attention for non blue chip stocks

Su (2003) uses a sample of 183 earnings announcements during the period from 1997 –

1998 on the Shanghai and Shenzhen securities exchange, including firms issue both A and B shares: domestic A-shares and international B-shares Based on the outcome of the event, firms in the sample are assigned to two portfolios: group I if actual EPS exceeds last EPS or group II if actual EPS is equal or less than last EPS Using event study methodology, he finds that A-shares investors react positively and negatively for group I and group II, respectively, during two days before announcements The reaction trend for group I and group II keep persisting after announcements These imply that A-share investors fail to reflect earnings information rapidly and B-share investors quickly incorporate earnings release These findings can be explained by the following reasons: most A-share investors are individuals with short-term investment horizon while most B-share investors are large institutions that can access more detailed and accurate financial information

Sponholtz (2008) in a study in Denmark finds significant positive abnormal returns accompanying the earnings announcements In addition, abnormal returns persist

several days after earnings announcements; show a signal of inefficiency market The study also shows that a positive correlation between the information content and predisclosure information, contrary to previous evidence of a low level of pre-announcement information The results confirm that analyst forecast models are the best proxy for unexpected earnings In this study, she uses lumped procedure, uniform ones, and trade-to-trade ones for stock return due to thin trading of small stock market Denmark

Recently, Mlonzi, Kruger & Nthoesane (2011) in a study in South African find empirical evidence that there is large negative share price reaction to earnings announcements on the ALtX stock market These results come from earnings announcements during a recessionary period, leading to share market erosion of the selected sample The study also shows the weak form of market efficiency of ALtX stock exchange

Evidences of price reactions to earnings announcements keep growing in other markets

in the world See more: Ariff, Loh & Chew (1997) in Singapore, Chen, Cheng & Gao (2005) in China, Abad, Sanabria, & Yagüe (2005) in Spanish, Chan, Watson and Wee (2005) in Australia, Louhichi (2008) in France, Fan-fah, Mohd and Nasir (2008) in Malaysia, Alzahrani (2010) and Alzahrani & Skerratt (2010) in Arab Saudi

2.3 Post Earnings Announcements Drift

Post Earnings Announcements Drift (here after PEAD) was first reported in a seminal study of Ball & Brown (1968) This phenomenon is described as abnormal return of good news (bad news) continue to drift up (down) in positive (negative) direction after earnings announcements (Nguyen, 2009)

PEAD has been confirmed for the last four decades and continued to a challenge to the efficient market hypothesis For example: Foster, Olsen & Shevlin (1984), Bernard & Thomas (1989, 1990), Abarbanell & Bernard (1992), Ray & Ball (1992), Bhushan (1994), Rangan & Sloan (1998), Liu, Strong & Xu (2003), Booth, Kallunki & Martikainen (1996), Forner, Sanabria & Marhuenda (2009) Some explanations for the existence of PEAD are under-reaction to earning surprise, information biased

processing, deficient asset pricing model, misspecification of risk, and stock liquidity (Nguyen, 2009)

2.4 Other reactions to earnings announcements

Previous studies also provide evidence of other reactions to earnings announcement, as measured by volatility, trading volume, liquidity, and information asymmetry As mentioned at the introduction of this chapter, we just brief some key studies concerned with these reactions

2.4.1 Trading volume and earnings announcements

The important difference between price and volume reaction is that the price reaction reflects changes in the expectation of the aggregate market while the volume reaction reflects changes in the expectation of individual investors (Beaver, 1968)

Kim & Verrecchia (1991) in their theory study also show that the volume reaction at the time of the public announcement is proportional to both the absolute price change and a measure of differential precision across traders The newly public announcement just is important to investors with less precise private predisclosure information It means that the differential precision of private information before announcements cause differential reaction of investors in trading volume In their following theory study, Kim & Verrecchia (1994) also present that trading volume respond to earnings announcements due to informed opinions resulting from disclosure

Various empirical studies have confirmed the link between trading volume and earnings announcements3, for example: Beaver (1968), Kiger (1972), Morse (1981), Bamber (1983, 1987), Gajewski (1999), Cheung & Sami (2000), Landsman & Maydew (2002), Otogawa (2003), Abad, Sanabria, & Yagüe (2005), Alzahrani (2010)

Beaver (1968) is the first to examine volume reactions to earnings announcements He states that there is a large mean volume residual in the announcement weeks than the mean volume residual during the non-report period

3

See Bamber, Barron & Stevens (2011) for a more detailed literature review

Using two different tests (a parametric chi square test and a non parametric binomial test), Morse (1981) also investigates trading volume reaction to interim and annual earnings announcements Similar to price reaction, he observes significant abnormal trading volume during four days surrounding quarterly earnings release

As discussed above, Cheung & Sami (2000) extend Morse’s study They also find substantial volume reactions around earnings announcements date for blue chip stocks portfolio as well as non blue chip stocks one in the context of Hong Kong

2.4.2 Liquidity, information asymmetry and earnings announcements

Both theoretical and empirical studies have examined liquidity and information asymmetry surrounding earnings announcements Kim & Verrecchia (1994) suggest that there may be more information asymmetry at the time of an announcement than in non announcement periods due to the different ability of processing public information

of traders As a result, it implies that bid-ask spreads increase, suggesting that market liquidity decreases at the time of an earnings announcement

There are a lot of empirical papers studying the impact of earnings announcements on liquidity and information asymmetry Nevertheless, findings from these studies are mixed See Morse & Ushman (1983), Venkatesh & Chiang (1986), Lee, Mucklow & Ready (1993), Krinsky & Lee (1996), Yohn (1998), Gajewski (1999), Acker, Stalker and Tonks (2002), Affleck-Graves, Callahan & Chipalkatti (2002), Otogawa (2003), Ranaldo (2003), Hegde & McDermott (2003), Wael (2004), Abad, Sanabria, & Yagüe (2005), Alzahrani (2010)

Morse & Ushman (1983) find that no significant changes in bid-ask spreads around earnings announcements

Venkatesh & Chiang (1986) use bid-ask spread as a proxy of information asymmetry in earnings and dividend announcement They classify three group of announcement: joint announcement, fist announcement – announcement is announced prior to another announcements in thirty days, second announcement – announcement followed the first announcement by at least ten days and no more than thirty days The results find an

increase in information asymmetry only before the second announcement and almost

no increase before the joint and first announcement

Lee, Mucklow & Ready (1993) study a sample of 230 NYSE firms, covering period January 4 to December 30 in 1988, and using intraday data They offer that spreads widen and depths fall before quarterly earnings announcements In addition, increased spreads are observed during and after the announcement

Similarly to Lee, Mucklow & Ready (1993), Krinsky & Lee (1996) show the impact of earnings announcements on three components of bid-ask spread: (1) adverse selection costs increase substantially before and after the announcements, (2) inventory holding component is lower during event and predisclosure period, (3) order processing cost are decreased significantly surrounding earnings release

Affleck-Graves et al (2002) find that an increase in adverse selection costs on the day prior to earnings announcements affect only NASDAQ firms with less predictable earnings

Acker, Stalker and Tonks (2002) investigate the determinants of inside spreads and their behaviour around corporate earnings announcement dates, using a sample of 195 less-liquid stocks on the London Stock Exchange over the period 1986 - 1994 They observe spread start to narrow 15 days before an earnings announcement, and narrow further by the end of the announcement day Consistent with the findings of Acker, Stalker and Tonks (2002), Otagawa (2003) also finds evidence that there are large decreases in daily bid-ask spreads and slight increases in daily depth during the announcement period and the period just after releasing quarterly earnings in Japan

In contrast, Ranaldo (2003) conducts a detailed analysis of market behaviour around real-time news release arrival for a sample of 30 liquid stocks on Paris Bourse The result indicates that the spread is tighter and the market depth is thick, especially in few minutes around news release arrival It also documents a decrease in adverse selection cost around the public information disclosure

2.4.3 Volatility and earnings announcements

Employing intraday data, Krinsky & Lee (1996) find that volatility increase before and following quarterly earnings announcements in the context of United States

Acker (2002) examines volatility increases following annual earnings announcements

by a sample of 379 announcements in UK’s FTSE 100 over period November 1989 and October 1996 Sixty firms in the sample have option quoted on their shares Based on the sign of earning surprise and the standard deviation of firm’s reported EPS changes, the earnings announcements are split into portfolios: good news and bad news, easy-to-interpret and difficult-to-interpret He applies implied standard deviations by options prices to measure volatility of stock price The results report that the volatility reactions

to easy-to-interpret and good news announcements are highest on the day of announcements For difficult-to-interpret and bad news portfolios, the volatility reactions are lower than good news announcements on the day of announcements Moreover, the reactions to bad news also are delayed until the following day He argued that the delays of bad news may reflect market uncertainty about the implications of the news

In despite of raised concern about the decrease of informative of earnings information, Landsman & Maydew (2002) report evidence of an increase overtime in the informativeness of quarterly earnings announcements in the past three decades since the study of Beaver (1968), as measured by abnormal trading volume and abnormal volatility

2.5 Information disclosure regulation on listed firm in Vietnam

Circular No 09/2010/TT-BTC of The Ministry of Finance dated 15 January 2010 on Disclosure of Information on the securities market regulate a standard on disclosure reporting for listed firms In particular, listed firms in Vietnam have obligations to disclose regulated information that involve periodical disclosure of information, extraordinary one, and disclosure of information on request from the SSC or Stock Exchange SSC also encourage public company and listed firms to announce voluntary disclosure related to firm’s activities and performances

Listed firms have to disclose three types of financial statements:

1 Annual financial statements

Annual financial statements must be audited by approved auditing firms which satisfies practicing conditions in accordance with regulations of the Ministry of Finance The date of completion of annual financial statements shall be no later than 90 days from the end of fiscal year Within 10 days from expiry of the date for completion of annual financial statements, listed firms must present their annual audited financial statements

to SSC, Stock Exchange In addition, listed firm must publish the full text of the annual audited report in 1 edition of a newspaper published nationwide accompanied by the website address where the financial statements are provided, in order to enable investors to refer to such statements

Disclosed annual financial statements comprise balance sheet, income statement, cash flow statement, and notes to the financial statements In the case where a listed firm is the parent company of another institution, disclosure of annual financial statements shall include the financial statements of the parent company as well as consolidated financial statements

2 Quarterly financial statements

Listed firms must provide their quarterly financial statements to SSC and Stock Exchange within 25 days from the end of a quarter If the listed firm is the parent company which is required to prepare consolidated financial statements, the period for disclosure of information shall be 50 days after the end of the quarter In addition, listed firms have to publish the full text of quarterly financial statements on their websites

Quarterly financial statements include balance sheet, income statement, cash flow statement, and notes to the financial statements If a listing firm has subsidiary companies, it must submit both quarterly financial statements of the parent company and consolidated financial statements to the SSC and Stock Exchange

If the profit after corporate income tax stated in the report on results of business operation as between the reporting period and the same reporting period of the previous year fluctuates 10% or more, then the listed firm must explain the reasons leading to such unusual fluctuation in its quarterly financial statements

3 Semi-annual financial statements

In addition to announce quarterly and annual financial statements above, listed firms must also prepare and disclose semi-annual financial statements which have been examined by an approved auditing firm within a period of 45 days following the end of the second quarter of each year If the listed firm is the parent company which is required to prepare consolidated financial statements, the period for disclosure of information shall be 60 days from the end of the second quarter of each year

Semi-annual financial statements must be disclosed on the information disclosure media of the SSC and SE and on the website of listed firms

In order to enhance more transparency and enforceable stock market, State Securities Commission intends to issue a new circular replaced Circular No 09/2010/TT-BTC in next year

2.6 Summary

In summary, the issue market reactions to earnings announcements have been discovered in many aspects of market since 1960s It can be seen that voluminous studies focus on developed market such as US, UK and other developed countries There are very few studies examined the impact of earnings announcements in emerging market, especially in Vietnam Trần (2010) investigates the efficiency of stock market to information in Vietnam She examines price behaviour of three listed firms: STB, VNM, and FPT to dividend announcements Obviously, there is lack of empirical evidence on informational effect in the context of Vietnam This motivates us

to conduct this study to investigate the association between stock price and annual earnings announcements in Vietnam stock market

CHAPTER 3: DATA AND METHODOLOGY

Based on the research objectives, research questions and scope in the first chapter, and literature review in chapter 2, chapter 3 mentions detailed data in the sample, research methodology applied in the thesis as well as development of hypotheses

3.1 Data

3.1.1 Sample selection

Vietnam has two stock exchanges: Ho Chi Minh Stock Exchange and Ha Noi Stock Exchange For this study, we choose firms listed on HOSE because HOSE was set up first and has been a four – year longer history before Ha Noi Stock Exchange

The study focuses on annual earnings announcements of listed firms for two fiscal years 2009 – 2010 The sample selection includes firms that have been listed on the HOSE through 2008 to 2011 inclusive We exclude financial companies, such as banks, insurance firms, funds and financial firms in the thesis because their characteristics are different We also require firms that have no missing data and a listing history longer than 270 trading days before earnings announcements Thin trading problem may cause misspecification of statistical tests (Cowan & Sergeant, 1996) Thus, we remove from the sample stocks that do not trade for at least 80% of the days in either the event or estimation period

According to the selection criteria, we remove all unsatisfied earnings announcements We have a final sample of 211 annual earnings announcements involving 109 listed firms The details of firms are given in the appendix

3.1.2 Data collection

We define annual earnings announcements date (event date) is the earliest date on which fourth quarter financial statement is publicised to the HOSE Firstly, these financial statements are published on the website of HOSE (www.hsx.vn) Then,

financial statements are publicised to Daily market review bulletin of HOSE If the

announcements are disclosed on Saturday and Sunday, we’ll employ the next

trading day as the announcement date

Annual earnings announcements date and financial statement covering two fiscal

years between 2009 and 2010 are collected from the HOSE announcements The

website of HOSE also provides us daily closing price of stocks and Vietnam Index

We adjust closing prices for any dividend and split distribution

Table 3.1 Sample selection

Exclusion for:

Not satisfied for trading less than 80% of the

3.1.3 Sample description

Table 3.2 Sample distributions by timing of announcement

Table 3.3 Firm sample distributions by sector

Electricity generation and

25 days following the end of the quarter This regulation creates clustering of event dates

In addition, we also assign sample into different sectors Due to sector criteria of HOSE, we decide to group firms into 7 sectors including manufacturing; mining; construction and real estate; whole trade and retail trade; agriculture, forestry and fisheries; transportation and warehousing; electricity generation and distribution The highest percentage belongs to manufacturing with 48.62% Other sectors take a small percentage in our sample

3.2 Methodology

3.2.1 Descriptive statistics

We provide a descriptive statistics of some variables before conducting event study method The descriptive statistics is to describe different fluctuations of stock price, stock return, and market value by year and whole sample According to the

results, we also give some explanations for normality of data

3.2.2 Event study

We apply standard event study methodology4 to examine the price reaction of earnings announcements This methodology was first introduced by Ball and Brown (1968) and Fama et al (1969)5 The event of interest is annual earnings announcements of listed firms on the HOSE

Definition of event window and estimation window

An event window is defined to capture the effect of annual earnings announcements on stock price An estimation window is used to estimate parameters of ordinary least-squares market model regression

We set the event window at 20 trading days before and 20 trading days after the earnings announcement date The announcement date is set day 0 Estimation period is defined 250 trading days preceding the event window, from day -270 to day -21 relative to the announcement day (Brown & Warner 1985; Corrado 2011; MacKinlay 1997)

Figure 3.1 Time line of event study

Measuring abnormal return

The abnormal return (AR) is the actual stock return minus the normal return over the event window

 ARit denotes the abnormal return on stock i at day t

 Rit denotes the return on stock i at day t

 E(Rit) denotes the normal return or expected return on stock i at day t in the absence of the event

In literature, there are many methods to estimate the expected return such as constant mean return model, market model, market-adjusted model, the capital assets pricing model (CAPM), Fama-French three factor model In this study, we use ordinary least-squares market model for estimation of normal return because it

is widely used in empirical research6

In market model, returns are estimated by using the following equation:

= + x + ℮ it (2)

Where:

 Rit denotes the return on security i on day t We calculate continuously compounded return, i.e Rit = ln(Pit / Pi,t-1), where Pit is the closing price at day t

 Rmt denotes the market return on day t Rmt is calculated with formula Rmt = ln(It / It-1) where I is the value weighted index of the Ho Chi Minh Stock Exchange

 and are constant intercept and regression coefficient Estimates of are taken from the OLS market model regression from estimation window [-270, -21]

 ℮ it is return residual for security i at day t with E(℮ it ) = 0

Testing average abnormal return (AAR)

In order to test price reaction to the annual earnings release, we average the AR across events (cross-section) for each trading day within the event window (-20, 20) The average abnormal return (AAR) for any day is computed by:

6

According to Brown & Warner (1985), OLS market model has similar power as other sophisticated models