Could Conservatism Reduces Earning Management Practice in IPO During the Current Covid-19 Pandemic?

The research examines the market reaction to the information prepared by two different accounting standards during the covid-19 pandemic, and the role of Conservatism in reducing the earning management practice in the initial public offerings (IPO). The response considers the rationality level, the avoidance risk, and accounting standards' role in preparing the information. An event study of IPO is conducted on the various primary market in the U.S., Europe, and Asia. Our conclusions are twofold. First, the market reacts quickly to the stated information, where the accounting standards influence the reaction level. Second, during the covid-19 pandemic, earnings management plays a role in increasing the market reaction intensity. At the same time, we have not found conservatism's role in reducing the earning management activity.


INTRODUCTION
Covid-19 cases were inevitably one of the most significant issues in 2020.The virus was firstly reported in Wuhan Province, China.Then, it spreads around the world.Based on the tally of worlddometers.info(2020), total cases per 12 th December 2020 reached 71.503.497total cases, 21.811.478active cases, and 1.602.509total death toll victims.Since the beginning of the pandemic, several studies have examined the impact of covid-19 in economic, social, and finance.Shehzad, Xiaoxing, and Kazouz (2020) observed whether the dangerous pandemic impact is a fact or a rumor, and they found that the covid-19 pandemic affected the monetary crisis.The U.S. and Europe suffered a worse effect.Based on S&P 500 data, there is a long time-range of the covid-19 pandemic, and it raises the financial volatility that challenges the management activity risk (Albulescu, 2020).On the other hand, China's evidence shows that the pandemic impact is different in each sector and finance market (He et al., 2020).Other studies show another variation of the behavior and the market responses to explain a pandemic's impact, e.g., communication media's role (Salisu dan Vo, 2020;Cepoi, 2020).
Furthermore, previous researches stated two concepts impacting the market reaction.First, the bounded rationality theory explains that the limitation of ability, time, and any other things make marketeers fails to take a rational decision as they want (Kahneman, 2003).Second, the risk-aversion theory determined that marketeers will avoid uncertainty risks (Scott, 2015) or act relatively on market volatility (Takkabutr, 2013).The interesting issue is how the measured market reaction can reveal the impact of the available events or information in the abnormal situation.
In terms of the managerial aspect, there are discussions on the market's pre-released information.Investors can gain a higher return by including new information in their portfolios when others did not get those messages (Easly and O'Hara, 2004).However, the data is related to organizational management (Bartov and Mohanram, 2004).Furthermore, the application of each accounting standard is considered to bring a distinct effect on the information.The conservatism principle is believed to benefit, though the accounting measurement is possibly irrelevant and requires rejuvenations (Wolk et al., 2017).Chen et al. (2006) stated that accounting standards' conservatism could improve the risk-sharing quality, resulting in an efficient contract.The economic earnings tend to be downward-bias as accounting earnings show in the absence of earning management.On the other hand, the unbiased accounting earnings resulted from the unbiased accounting standard will present unbiased economic gains.The fair value concept is admitted to bringing the unbiased accounting standard.Hence, the interesting issue is comparing market reaction on the information prepared by conservatism concept vis-a-vis fair value concept, particularly during the covid-19 pandemic.
An event study of initial public offering (IPO) during the pandemic in several countries is conducted to examine the market reaction on the covid-19 plague based on the different information given to the market.The selection of the method is an alternative way to test Chen et al. (2006) conclusion about a conservatism concept.The conclusion states that accounting value manipulation increases when a company sells stocks to the second or third parties.That problem can be eased by the conservatism in accounting standard.Lastly, there are two steps of research.First, we check the comparison level of market reaction on the IPO activities by companies implementing two different standards (GAAP vs. IFRS).Second, we analyze market reaction's comparison level to the companies allegedly practice the management earning to maximize the IPO's success.
The study proposes two hypotheses.First, there is a significant market reaction on the IPO activities during the covid-19 pandemic, where the reaction's level is different on each accounting standard applied.Second, the more significant response is found on the indicated companies that conduct earnings management.The finding is that earning management plays a role in the intensity of the market reaction during the covid-19 pandemic.However, the role of conservatism to reduce the earning management practices has not been detected yet.
The study is expected to fulfill the emptiness on the covid-19 impact to the market reaction, particularly on the events held by two different accounting standard users, when the abnormal situation exists.There are two intentions of the research.The first objective is to support existing theories, which are bounded rationality theory and risk-aversion theory.The second intention is to test the consistency of Chen et al. (2006) research conclusion related to the conservatism role in reducing the earning management practice in the abnormal situation.
Lastly, the study will be discussed in the following parts.Section 2 explains the theoretical background and hypotheses development.The third part contains the data and research methodology.Part 4 presents the result and analysis.The last section shows the conclusion of the research.

LITERATURE REVIEW 1. Theoretical Background a. Bounded-rationality Theory and Risk-aversion Theory
The bounded rationality and the risk-aversion theory are the branch of behavioral psychology on decision making.The approach has been applied to various sciences such as economic, accounting, and finance.Hence, those interesting theories are used to examine the panic market reaction on the covid-19.
Concerning economics, Adam Smith postulated the neo-classic economic paradigm (Fror, 2007).In this paradigm, a rationality instrument forms the self-interest, and the bounded rationality happens because of limitation on the 'attention-capacity'. Bounded rationality defines rationality as a part of appropriate behavior to reach a specific goal in the boundary of pressure and current condition (Simon, 1972).The limitation of rationality is possibly from the complexity of calculation, the obstacle to find the alternative design, or the uncertain situation.The uncertainty situation means that in the future, there are events that unpredicted and unable to be predicted (Dequech, 2001).An individual will adapt to the external transformation and will innovate to advance.On the psychologic approach, the unity of preference and confidence on the rationality is the fear to predict the harmful condition, the optimism to take a risk, and the likeor-dislike with the real facts (Kahneman, 2003).To decide on how to solve and to decline the problems' degree, Lin (2014) offered the adaptive rationality to the decision making of higherlevel issues.
Based on those explanations, the conclusion is that bounded rationality can explain the behavior or the market reaction to an event.The investors will select two options on the finite information, the unpredicted condition, and the difficulty to calculate the estimation.The choice is either to take risks optimistically or to take risks carefully.Lastly, investors will adapt to choose rational decisions.
In terms of risk aversion, it happens naturally when the risk is systematic or connected to all individuals (Zhang et al., 2014).The risk ordinary stochastic ratio determines the aversion, and the statistic discrepancy will affect the difference of its utility function.One of the earlier studies states that risk aversion can be measured by the component importance, the risk, and the behavior (Menezes dan Hanson, 1970).Those components, in the different comparative static context, determines the behavior of risk-premium.Furthermore, Kimball (1993), on the risk aversion standard study, conveys the Von Neumann-Morgenstern utility function.There is a negative relation between a risk with low-decreased asset and an unintended independent risk in that function.On the uncertainty condition, the utility function defines the risk standard to maximize the appropriateness.The utility function can also decrease the absolute risk (Nielsen, 2004).The interesting point is that the risk aversion depends on the gambling choice in deciding the marginal risk (Tust dan Lybbert, 2009).In conclusion, risk aversion occurs when the risk is systematic and in a general scope, and the maximization of risk aversion utility function possibly reduces the risk.However, there is a weakness that the individuals use the gambling reference to determine the risks.

b. The Accounting Information (Conservatism vs. Fair Value)
There are many efforts to define the basic accounting framework for its users (Wolk, 2017).The failure in accounting concept is to build a general theoretical framework in presenting and interpreting the accounting financial data.Hence, it generates crucial issues on equity meaning and any other problems.There are arguments on the studies of the conservatism and fair value principles, and it covers a large area of researches.Thus, to narrow the issue, the study's background focuses on the topic of accounting information.
There are three accounting information streams related to delivering financial information to the market (Easley dan O'Hara, 2004).The first stream is related to the role of private information on the rational expectation model.While conducting a transaction with the informed investors, the uninformed investors need risk-premium to compensate for the adverse selection.The second stream is related to the role of incomplete but still symmetric information.The last stream is associated with the company's disclosed information, where it changes the marketeer's risk.
Later, Barth et al. (2012) explain that the mandatory application of IFRS can be compared with GAAP in the context of differences in financial standards.The source of comparability is in terms of smoothing earnings, accrual quality, and timeliness.However, there are still significant differences between the two concepts.One of the problems between the two standards is related to the 'grey area' in measuring values (Badia et al., 2016).The concept of historical cost is considered to be biased in calculating value.Thus, an idea of unbiased fair value is needed.However, the difficulty in measuring fair value causes managers to exercise their discretion by applying conditional fair value.That is a form of earning management used by managers in providing financial information to the market.Furthermore, there is a relationship between earnings management's role and information that affects the market (Bartov dan Mohanram, 2004).The connection is that earning manipulation and the timing of providing private information to the market can affect the company's unfair return.This unnatural return is then used by managers who have stock-exercise rights to maximize their stock returns.
Lastly, Chen et al. ( 2006) revealed conservatism's role in reducing managers' incentives from practicing earning management.The role of conservatism increases in two conditions.First, parties with personal interests (for example, top managers or major shareholders) influence the accounting process's preparation.Second, the accounting value plays two roles: the investors in calculating their business performance and the stewardship in increasing their salaries.Additionally, conservatism improves the quality of risk-sharing, so it produces efficient contracts.However, the actual economic earnings will lead to a downward bias, as shown by the accounting earnings result in the absence of earning management.On the other hand, the unbiased economic gains will be delivered by the accounting earnings generated from the unbiased accounting standards.

Hypotheses Development
The previously described theories can explain some research related to the market reaction to the Covid-19 pandemic.There is an asymmetric dependence between the stock market rush and information associated with Covid-19, such as the outbreak, media coverage, and false information (Cepoi, 2020).Furthermore, the impact of Covid-19 is different in each sector, as shown in China's capital market (He et al., 2020).While some industries are affected positively, the others negatively suffered from the effect.The United States stock market also shows another evidence of each sector's different reactions (Mazur et al., 2020).The rationality theory explains the two evidence above, where uncertainties raise confusion, limited information, and adaptation.Lastly, the risk avoidance theory can also explain how the Covid-19 pandemic got a quick reaction from China's stock market with a different response variation at each stage of the outbreak (Ashraf, 2020).Shehzad et al. (2020) found similar cases in the stock markets of the United States (S&P 500, Nasdaq), Germany (DAX 30), Japan (Nikkei 225), and Italy (FTSE MIB).
The conclusion from the explained market reaction above is that the Covid-19 pandemic impacts markets, either applying IFRS or GAAP standards.However, no specific studies examine the market reactions' differences between conservative information and fair value information in an uncommon situation.There is also no prediction of the differences' level to the data.By using an IPO event that occurred during the Covid-19 pandemic, the hypotheses is proposed as follows: H1a: There is a significant market reaction to the IPO activities during the Covid-19 pandemic.H1b: The difference in applying the accounting standard causes the distinct market reaction's level.Furthermore, Chen et al. (2006) stated how the IPO becomes a medium for personal interests to manipulate the accounting values when a company offers shares to the second or third parties.In this part, the role of conservatism in reducing earnings management arises.Therefore, the differences in the market reaction can be used to investigate conservatism's role in abnormal conditions.The indication of earning management's existence or absence during an IPO measures the difference's level.Hence, the following hypothesis is proposed: H2: The more significant market reaction on the IPO activities is found in the companies indicated to practice earnings management.

RESEARCH METHOD 1. Methodology
The event study research has been conducted to test stock prices' reaction to several events (Binder, 1998).There are two primary purposes of the study.First, to test the null hypothesis that the information is efficient on the market.Second, to examine the event's impact on the shareholder asset values.Regarding the Covid-19 pandemic, the event study is widely carried out for the Covid-19 effects, such as the reaction in each market sector (He et al., 2020), the general market reaction (Cepoi, 2020;Mazur et al., 2020;Shehzad. et al., 2020;Ashraf, 2020), stock market volatility (Albulescu, 2020), and the stock return predictions (Salisu, and Vo, 2020).There are four models commonly used in the event studies (Basdass and Oran 2014).Those are zero adjustment returns, mean adjusted returns, market-adjusted returns (index model), and market and risk-adjusted models.This study will use a market and risk-adjusted model in the Covid-19 cases because it is reliable in predicting abnormal returns (He et al., 2020).
The normal return on the IPO share price is calculated as follows: Ri,t = α + βiRmt + εit (1) The average abnormal return (AR) is calculated as follows: (2) Meanwhile, the accumulated abnormal return (CAR) is calculated as follows: Where i is the IPO's company, t is the trading day, and t1&t2 is the event window period.
The statistical testing uses the method proposed by Patel (1976) and Campbell et al. (1997), referred as J1 where: In alternative, J2 is used by calculating the CAR based on its standard deviation as follows: n is the total of estimation days.

Data, Sample Selection, and Event Windows
The data is from companies that conducted IPOs in all capital markets during 2019-2020.We use the Osiris program to collect the list of companies.Next, we retrieve the companies' historical stock prices from yahoo finance.The data of companies that conducted IPOs in 2015-2020.5.419 2.
The sorted companies that contain the data of IPO dates, ticker identities (that is used for the data collection on yahoo finance), and listed stock markets. 5.149

3.
The sorted companies that conducted IPOs during the Covid-19 pandemic.The period is measured based on the estimated length of the testable event window (November-December 2019, and January-August 2020).The determination of the accounting standard type used by the company is obtained from Osiris data.We found three groups of companies based on the accounting standard used.The groups of companies are IFRS, GAAP, or unidentified accounting standards.These companies have implemented the same accounting standards in the last six years (2015)(2016)(2017)(2018)(2019)(2020).
Furthermore, we identify the companies indicated to have performed earning management shortly before the IPO.Chen et al. (2006) found to use the parameter of stewardship salary increases prior IPO activities.However, the study did not use this data due to data limitations.For this reason, we propose a ratio of book equity to net sales re/p.Then, we calculate the difference between the current year ratio t and the previous year ratio t-1.This proxy's choice is made because an increase in book equity can reflect the result of earning management.Meanwhile, net sales is a more stable parameter because it can be reflected in the year-to-year buyers and sellers' contracts.The calculation of earning management detection is as follows: The earning management is detected when Δre/p in the IPO year is greater than the previous year.Then it is followed by a decreasing Δre/p immediately in the following year after the IPO.Table 2 shows the results of sample selection and data processing.

Table 2. The summary of data selection
Furthermore, there are variations in selecting the event windows and the estimation windows of IPO activities.The event studies of IPO activities in Turkey show the variations (Basdas and Oran, 2014).The criteria of the estimation window are before or after the event window (Hartono, 2010).There is no benchmark for the estimation window.Short estimation windows may not be able to capture the full event.On the other hand, the long estimation window potentially raises confounding events.Jiang and Leger (2010) used some variations of the event window [0,20,40,60,90,120] and the estimation window [120,140,160,190].Hence, we determine the length of the event period [0.20] and the estimation period [20,80].This selection period is to minimize any missed events and potential confounding events.Figure 1 shows the model of the estimation period and event period.

RESULT AND DISCUSSION
Table 3 shows the descriptive statistics on the test results.Twenty-nine samples are eliminated during the test because it did not fulfill the range of event windows and estimation windows.Based on the company's differences in accounting standards, the average CAR for GAAP users is greater than IFRS users (0.0828 vs. 0.0298).The result also shows that the standard deviation of CAR on IFRS is greater than GAAP, which indicates that the information's volatility is more significant.Meanwhile, the average CAR for companies indicated to practice earning management is greater than those that did not (0.0078 vs. 0.0483), and it has a lower standard deviation (0.3713 vs. 0.4399).Finally, the average CAR in each country varies, where Taiwan has the highest average (0.3205), and Sweden has the lowest average (-0.1188).However, Taiwan records the highest standard deviation of variance (0.8935), while the Netherlands has the lowest standard deviation (0.0185).   4 presents the comparison of the CAR value and its significance based on the J1 and J2 tests.The test results show that during the Covid-19 pandemic, the IPO announcement by using either GAAP or IFRS standard receives a high market reaction.However, the GAAP standard has a bigger advantage on both the average CAR value (0.0828 vs. 0.0298) and the significance result (114.0378 vs. 20.7700).Campbell et al. (1997) stated that the J2 test has advantages over the J1 test because it increases the test's strength by giving a weight to the lower residual variance.The J2 test's strength is proven to bolster the IFRS standard results on the 4th day and afterward.The CAR results are positive on GAAP and IFRS, while CAR results are negative for companies whose accounting standards are not identified (mean = -0.1304;min = -0.2750;max = -0.0025).It shows that the market is rational towards the financial statement information, is risk-averse, and is cautious about weak information.The result is supported by the unidentified standard CAR's insignificant result in the first three to six days after the IPO announcement.It indicates that the initial market reaction is weak.The results in table 4 reflect that the first hypothesis can be accepted.

The influence of earning management practice indication
Table 5 shows the market reaction test results on the IPO activities where the earnings management exists or is absent.The test results show that the market reacts strongly and positively at 99% significance almost on each day for both groups of companies.The positive CAR results in the first 20 days indicate that Covid-19 does not affect IPO events.The interesting point is that the average CAR on the companies group indicated to practice earning management is higher than those that did not (0.0708 vs. 0.0483).That is consistent with one of the risk avoidance characteristics where individuals rely on reference gambling in determining their marginal risk (Tust and Lybert, 2009).It supports Chen et al. (2006) research results related to earning management's role in manipulating the information to be more attractive and get a more significant market reaction.That also holds the proposed second hypothesis.7 on the first 20 days after the IPO activities.Table 6 shows 99% significant results on the two testing tools for companies applying GAAP standards.It is regardless of the existence or the absence of earning management practice.On the other hand, for companies that implement the IFRS standard, the J1 test result shows the significance only on the first day of the announcement.However, the J2 test presents a significant result on all event window days.Lastly, for companies with unidentified accounting standards, the result is similar to the previous explanation where the beginning event window days yield insignificance.
On average (Table 7), the companies that apply the GAAP standard have a higher CAR than those that implement the IFRS standard, whether the earning management indication exists (0.1074 vs. 0.0208) or not (0.0643 vs. 0.0360).The interesting point is, The CAR of IFRS companies that are not indicated to practice the earning management is higher than the opposite group (0.0360 vs. 0.0208).However, the J1 test resulted in insignificance at those values.Hence, the validity of that results is still weak, and there is still proof that the earning management practice plays a role in increasing the market reaction.The role of conservatism in GAAP in pushing down the earning management practice has not been found yet.Lastly, the negative results for firms whose financial standards are unidentified prove that the risk avoidance theory is valid.Table 6.The CAR significance on the cross-section between the accounting standard and the earnings management indication Table 7.

Mean
The study aims to examine the market reaction to the information conveyed by two different accounting standards.Using the IPO events during the Covid-19 pandemic, we want to investigate the rationality and the risk aversion of the presented information.The previous studies' results show that the market reacts quickly to all available information in the emerging uncertainties (Cepoi, 2020;Mazur et al., 2020;Shehzad et al., 2020;Ashraf, 2020).Meanwhile, the results of previous studies also argued on how differences in information based on the accounting standards applied can affect the market (Bartov and Mohanram, 2004;Chen et al., 2006).
The study found that the market reacts quickly to the information presented, where the accounting standard affects the reaction level.The highest reaction is found in companies that apply the GAAP standard.In contrast, there is no significant reaction at the beginning of the event windows for companies whose financial standards are unidentified.That proves rationality and risk avoidance behavior on the information whose validity is still unknown.
The research also found that earning management practice plays a role in increasing market reaction intensity during the Covid-19 pandemic.The surprise is that the increases are seen in GAAP companies indicated to practice the earnings management.The result is contrasted with the previous studies' results related to conservatism's role in reducing the earnings management practice (Bartov and Mohanram, 2004;Chen et al., 2006).

Contribution and Limitation
This research contributes twofold.First, the study proves that during the Covid-19 pandemic, the market reacted significantly to the company's information.The reactions also consider the accounting standards applied.Second, the research provides a new view of conservatism's role in reducing earnings management, particularly during the Covid-19 pandemic, where there are high uncertainties.On the other hand, several limitations are substantial for further research development.First, we only used a limited data sample and a finite period.Future research can expand the database used for the examination.Second, the research was conducted in a short period.Further studies can be developed by extending the research horizon.
that have daily transaction data on yahoo finance.The sorting for adequacy data testing includes the companies' daily stock prices, the stock market daily index, and the last five years' financial reports' incomplete data.The elimination criteria are the nonsynchronous data and the incomplete stock market index data.
The comparison of CAR values summary on the cross-section between the accounting standard and the earnings management 99% signifinance; ** 95% signifinance; * 90% signifinance.

Table 1 . The processes of sample selection
Table 1 explains of sample selection processes.

Table 5 . The comparison of CAR value based on the indication of earning management practice 3. The cross-section between the accounting standard and the earnings management indication.
The test analysis is based on Table6 and Table