Identification and Ranking of Financial, Non-Financial, and Behavioral Components Influencing Earnings Response Coefficient in the Iranian Capital Market (Data Mining Approach)
Abstract
|
Objective: The relationship between accounting profit and investor reactions, as well as stock price changes, is measured using the Earnings Response Coefficient (ERC). Previous research indicates a variation in ERC between companies and over time, which is attributed to specific company factors and characteristics. The aim of this study is to identify and rank the financial, non-financial, and behavioral components affecting companies' ERC. Methodology: Using data from 153 companies listed on the Tehran Stock Exchange over the period 2013 to 2022, and employing data mining techniques and two methods: stepwise forward regression and regression decision tree, the influential components on ERC were identified and ranked. Findings: The results of the stepwise forward regression method indicated that the components of auditor's opinion type, earnings per share (EPS), stock liquidity, growth opportunities, earnings stability, inflation rate, sales growth, operating profit ratio, relative strength index per share, and market index return respectively influence ERC. The regression decision tree method results also showed that the components of earnings stability, stock liquidity, EPS, stock price synchronicity, stock trading imbalance, adjusted trading volume, psychological line index, relative strength index per share, earnings conservatism, and financial statement restatements respectively influence ERC. The Wilcoxon test results also showed that the ranking of components influencing ERC is not the same in the two methods. Additionally, comparing the mean absolute error of the two methods indicated that the regression decision tree method identifies and ranks the influential components on ERC more accurately. Conclusion: The results of both methods confirm the high impact of EPS, stock liquidity, earnings stability, and relative strength index per share on the ERC of companies. |
Introduction
Investment decisions in financial markets are influenced by information sources. According to stock market theorists, one of the valuable sources of data is financial statements, whose main objective is to assist users and facilitate their decision-making process. One of the most critical factors in financial reporting is the announcement of profit-related information, which likely attracts the most investor attention. At the time of profit announcement, the market has certain expectations based on available information about the company's declared profit, and the difference between market expectations and the company's declared profit is called unexpected earnings (Widyaatmoko & Indarti, 2018).
Any event in the capital market elicits reactions from market participants, one of which is the profit announcement, to which the market reacts with stock price movements. The studies by Ball and Brown (1968) showed a significant relationship between a company's profit announcement and stock price changes. For instance, if the declared profit increases (positive unexpected earnings), there is a positive trend in stock price changes, and conversely, if the declared profit decreases (negative unexpected earnings), negative changes in stock prices occur. The Earnings Response Coefficient (ERC) is used to measure investors' reactions or the stock price response to accounting profit information (Sandi & Mulyana, 2024). The ERC, originating from Ball and Brown's seminal work in 1968, plays a crucial role in measuring the correlation between unexpected earnings and stock returns, providing valuable insights into the complex relationship between reported profits and stock price movements (Riyani et al., 2023).
The ERC related to accounting profit indicates the sensitivity of stock price fluctuations to accounting profit and shows how market analysis of reported profit information can be inferred from stock price volatility around the financial statement disclosure date. Each company has a unique ERC based on how reported information influences investor reactions and stock prices (Fatmawati et al., 2023). The ERC is defined as the impact of each unit of unexpected earnings on stock returns and is typically measured by the regression coefficient between abnormal stock returns and unexpected earnings. The ERC measures the stock price reaction or the company's market value to the information contained in the company's accounting profit. A low ERC indicates that the profit provides less information for investors to make economic decisions (Riyani et al., 2023). In other words, a high ERC suggests that the reported profit information is of good quality, causing a strong market reaction. Conversely, a low ERC indicates that the reported profit information is of low quality, causing a weak market reaction.
The importance of research on ERC primarily stems from the need to increase stakeholder trust in the disclosure of accounting information, allowing them to make informed stock decisions (Alang et al., 2023). Additionally, research on ERC is useful for investors in fundamental analysis to determine the market's reaction to a company's profit information. It is expected that if investors understand the factors affecting the ERC, they can better predict stock pricing based on profit information (Widyaatmoko & Indarti, 2018). Considering that, in addition to the quality of the declared profit, other financial factors as well as non-financial and behavioral factors can influence the ERC, this study attempts to identify and rank the financial, non-financial, and behavioral components affecting the ERC using a data mining approach.
Methods and Materials
This study is "applied" in terms of its objective and "descriptive-correlational" in terms of data collection and execution method. It is a "retrospective" study in terms of its temporal nature and a "quantitative" study due to the use of quantitative data. Furthermore, because the study reaches a general conclusion through analysis and inference from observations, it is "inductive." The theoretical foundations and research background were compiled using the "library" method, and data collection for the research variables was done using the "documentary" method. The theoretical foundations and the financial and non-financial data required for the statistical tests in this study were recorded and collected through the "card" tool. By studying the theoretical foundations and research background, the potential financial, non-financial, and behavioral components affecting the ERC (independent variables) were collected and classified into six groups as shown in Table 1:
Group of Factors | Variable Name | Measurement Method | Abbreviation |
|---|---|---|---|
Company Financial Status and Performance Characteristics | Earnings per Share (EPS) | Net profit after tax divided by the number of shares adjusted for end-of-period stock price | EPS |
Growth Opportunities | Market value of assets divided by book value of assets | GO | |
Return on Assets | Net profit after tax divided by book value of assets | ROA | |
Operating Cash Flows | Operating cash flow divided by book value of assets | CFO | |
Financial Leverage | Book value of liabilities divided by book value of assets | LEV | |
Sales Growth | Logarithm of current year sales relative to the previous year | SG | |
Operating Profit Ratio | Operating profit divided by sales | OP | |
Investment Environment Factors | Information Asymmetry | Venkatesh and Chiang model (1986); average daily bid-ask spread during the year | ASY |
Systematic Risk | Covariance of company stock returns and market index returns divided by the variance of market index returns | BETA | |
Stock Price Synchronicity | Piotroski and Roulstone market model (2004); R-squared from the extended market model regression (regression of stock returns and market and industry returns) | SYNCH | |
Delayed Stock Price Response | Hou and Moskowitz model (2005); ratio of R-squared from the restricted to the unrestricted market regression model | DELAY | |
Market Index Return | Geometric return of daily market index changes | MR | |
Inflation Rate | Annual inflation index announced by the central bank | IF | |
Capital Market Status and Performance | Free Float | The portion of a company's shares that are readily available for sale and calculated by the stock exchange organization | FF |
Stock Turnover Rate | Number of shares traded during the fiscal year relative to total shares | STV | |
Stock Liquidity | Inverse of Amihud's (2002) illiquidity measure; sum of absolute daily stock returns divided by the monetary volume of stock trades | LIQ | |
Number of Stock Trades | Logarithm of the number of stock trades during the year | STT | |
Corporate Governance Characteristics | Government Ownership | Percentage of shares held by the government or quasi-governmental institutions | GOW |
Institutional Ownership | Number of shares held by institutional shareholders relative to total common shares | IS | |
Financial Expertise of Audit Committee | Ratio of audit committee members with financial expertise (accounting or finance) to the total number of audit committee members | ACFEX | |
Audit Quality | If the company's auditor is the Audit Organization or Moafidrahbar Auditing Firm, assigned a value of 1, otherwise 0 | AQ | |
Board Independence | Ratio of non-executive (independent) board members to total board members | BI | |
CEO Duality | If the CEO is also the chairman or vice-chairman of the board, duality exists | DUAL | |
Financial Reporting Characteristics | Earnings Quality | Abnormal accruals based on the modified Jones model (1995) | PQ |
Timeliness of Reporting | Inverse of the time difference between the fiscal year-end and the date of the audited financial statement release | FRT | |
Auditor's Opinion Type | Unqualified opinion assigned a value of 0, and qualified, adverse, or disclaimer of opinion assigned a value of 1 | AO | |
Earnings Stability | Dechow and Dichev model (2002); regression coefficient of prior period profit with current period profit | ES | |
Financial Statement Restatements | Companies that have restated their income statement or balance sheet are assigned a value of 1, otherwise 0 | RS | |
Earnings Conservatism | Conditional conservatism (post-event or asymmetric earnings timeliness) based on the Khan and Watts model (2009) | CC | |
Financial and Behavioral Factors of Investors | Relative Strength Index per Share | Ratio of positive to negative monthly stock return changes for each company | RSI |
Psychological Line Index | Ratio of the number of days the stock closing price is higher than the previous day's to the total number of trading days | PSY | |
Adjusted Trading Volume | Stock trading volume to total issued shares adjusted by daily return | ATR | |
Trading Imbalance | Ratio of the difference between buy and sell trade volumes to the total buy and sell trade volumes | BSI |
The statistical population of this study comprises companies listed on the Tehran Stock Exchange from 2013 to 2022. Due to the heterogeneity of some members of the statistical population, sampling was performed using a screening or systematic elimination method. The following restrictions were considered for selecting the sample companies: 1) Listed on the Tehran Stock Exchange from the beginning of 2008 and continuously present until the end of 2022; 2) Fiscal year ending in March of each year; 3) No change in activity or fiscal year during the research period; 4) Not part of investment, holding, financial intermediary, banking, or insurance companies; 5) No trading halt longer than three months in any of the research years. After applying these restrictions, 153 companies were selected as the sample. Since the data is collected over 10 years for 153 companies, the sample size consists of 1,530 company-years. To prepare the data and perform calculations related to research variables, Excel 2016 was used. For statistical analysis of the impact of financial, non-financial, and behavioral variables on the ERC, data mining approaches using two methods—stepwise forward regression and regression decision tree—along with Clementine 12 software were employed.
Findings and Results
Table 2 shows the descriptive statistics of the research variables. For example, the mean of the Earnings per Share (EPS) variable in the research sample is 0.051, which ranges from a maximum of 1.568 to a minimum of -1.171. The median value is also 0.05, indicating that the distribution of this variable is close to a normal distribution given the small difference between the median and the mean. Furthermore, dispersion parameters are a measure to determine the extent of data dispersion from each other or relative to the mean. One of the most important dispersion parameters is the standard deviation. Among the research variables, Earnings Stability (ES) has the highest standard deviation, and Information Asymmetry (ASY) has the lowest standard deviation.
No. | Variable | Scale | Minimum | Maximum | Sum | Range | Mean | Standard Error Mean | Standard Deviation | Variance | Skewness | Kurtosis | Median | Mode |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
1 | ERC | Continuous | -12.876 | 13.264 | -1325.141 | 26.140 | -0.866 | 0.076 | 2.963 | 8.782 | 0.187 | 2.063 | -0.820 | 0.000 |
2 | EPS | Continuous | -1.171 | 1.568 | 78.078 | 2.739 | 0.051 | 0.003 | 0.101 | 0.010 | -0.905 | 69.631 | 0.050 | 0.000 |
3 | GO | Continuous | 0.682 | 76.022 | 4443.330 | 75.340 | 2.904 | 0.086 | 3.380 | 11.425 | 9.319 | 158.203 | 1.912 | 0.682¹ |
4 | ROA | Continuous | -0.563 | 0.830 | 232.021 | 1.393 | 0.152 | 0.004 | 0.151 | 0.023 | 0.524 | 1.079 | 0.128 | -0.105¹ |
5 | CFO | Continuous | -1.311 | 0.727 | 185.601 | 2.038 | 0.121 | 0.004 | 0.145 | 0.021 | -0.142 | 7.112 | 0.101 | -0.111¹ |
6 | LEV | Continuous | 0.031 | 1.505 | 820.520 | 1.474 | 0.536 | 0.005 | 0.207 | 0.043 | 0.146 | 0.173 | 0.537 | 0.267¹ |
7 | SG | Continuous | -3.195 | 2.953 | 401.103 | 6.148 | 0.262 | 0.009 | 0.360 | 0.129 | -0.528 | 10.911 | 0.269 | 0.000 |
8 | OP | Continuous | -1.393 | 36.698 | 365.668 | 38.091 | 0.239 | 0.024 | 0.958 | 0.918 | 36.060 | 1372.543 | 0.187 | 0.045¹ |
9 | ASY | Continuous | 0.000 | 0.122 | 56.037 | 0.122 | 0.037 | 0.000 | 0.011 | 0.000 | -0.136 | 3.468 | 0.038 | 0.000 |
10 | BETA | Continuous | -3.430 | 7.304 | 1151.515 | 10.734 | 0.753 | 0.024 | 0.930 | 0.865 | 0.546 | 3.595 | 0.722 | -3.430¹ |
11 | SYNCH | Continuous | -7.935 | 5.887 | -814.329 | 13.822 | -0.532 | 0.043 | 1.672 | 2.796 | -0.271 | 1.238 | -0.433 | 4.595 |
12 | DELAY | Continuous | 0.009 | 1.000 | 1036.699 | 0.991 | 0.678 | 0.007 | 0.267 | 0.071 | -0.595 | -0.732 | 0.720 | 0.009¹ |
13 | MR | Continuous | -0.019 | 0.092 | 51.348 | 0.111 | 0.034 | 0.001 | 0.035 | 0.001 | 0.236 | -1.159 | 0.025 | -0.019¹ |
14 | IF | Continuous | 0.083 | 0.652 | 466.865 | 0.569 | 0.305 | 0.005 | 0.213 | 0.045 | 0.616 | -1.222 | 0.238 | 0.083¹ |
15 | FF | Continuous | 0.000 | 1.000 | 390.690 | 1.000 | 0.255 | 0.004 | 0.167 | 0.028 | 1.047 | 1.380 | 0.229 | 0.498 |
16 | STV | Continuous | 0.000 | 5.202 | 936.485 | 5.202 | 0.612 | 0.018 | 0.717 | 0.514 | 2.145 | 5.865 | 0.362 | 0.000¹ |
17 | Liq | Continuous | 0.089 | 0.710 | 287.943 | 0.621 | 0.188 | 0.001 | 0.050 | 0.002 | 2.137 | 11.611 | 0.179 | 0.089¹ |
18 | STT | Continuous | 2.833 | 16.013 | 16281.393 | 13.180 | 10.641 | 0.046 | 1.815 | 3.295 | -0.563 | 0.625 | 10.884 | 3.296¹ |
19 | GOW | Continuous | 0.000 | 72.800 | 1243.790 | 72.800 | 0.813 | 0.142 | 5.551 | 30.811 | 9.181 | 92.377 | 0.000 | 0.000 |
20 | IS | Continuous | 0.000 | 0.987 | 1084.156 | 0.987 | 0.709 | 0.005 | 0.210 | 0.044 | -1.402 | 1.930 | 0.751 | 0.000 |
21 | ACFEX | Continuous | 0.000 | 1.000 | 721.783 | 1.000 | 0.472 | 0.009 | 0.366 | 0.134 | -0.054 | -1.352 | 0.667 | 0.667 |
22 | AQ | Nominal | 0.000 | 1.000 | -- | -- | -- | -- | -- | -- | -- | -- | 0.000 | |
23 | BI | Continuous | 0.000 | 1.000 | 947.605 | 1.000 | 0.619 | 0.006 | 0.251 | 0.063 | -0.788 | 0.526 | 0.600 | 0.600 |
24 | DUAL | Nominal | 0.000 | 1.000 | -- | -- | -- | -- | -- | -- | -- | -- | 0.000 | |
25 | PQ | Continuous | 0.431 | 2.128 | 1549.317 | 1.697 | 1.013 | 0.003 | 0.127 | 0.016 | 1.717 | 9.578 | 0.990 | 0.990 |
26 | FRT | Continuous | -0.500 | 0.050 | 11.127 | 0.550 | 0.007 | 0.001 | 0.026 | 0.001 | -14.221 | 241.050 | 0.009 | 0.008 |
27 | AO | Nominal | 0.000 | 1.000 | -- | -- | -- | -- | -- | -- | -- | -- | 0.000 | |
28 | ES | Continuous | -98.140 | 698.244 | 2441.692 | 796.384 | 1.596 | 0.539 | 21.070 | 443.958 | 25.136 | 797.142 | 0.253 | 0.000 |
29 | RS | Nominal | 0.000 | 1.000 | -- | -- | -- | -- | -- | -- | -- | -- | 0.000 | |
30 | CC | Continuous | -45.110 | 34.370 | 100.587 | 79.480 | 0.066 | 0.041 | 1.620 | 2.623 | -8.065 | 528.421 | 0.049 | -45.110¹ |
31 | RSI | Continuous | 0.000 | 300.157 | 5346.892 | 300.157 | 3.495 | 0.342 | 13.379 | 178.994 | 15.617 | 292.243 | 1.239 | 0.000 |
32 | PSY | Continuous | 0.000 | 1.000 | 715.385 | 1.000 | 0.468 | 0.003 | 0.137 | 0.019 | 0.097 | 4.308 | 0.473 | 0.500 |
33 | ATR | Continuous | -81.274 | 74.020 | 725.876 | 155.294 | 0.474 | 0.167 | 6.541 | 42.784 | 0.145 | 58.838 | 0.478 | 0.000 |
34 | BSI | Continuous | -1.000 | 1.000 | 247.571 | 2.000 | 0.162 | 0.006 | 0.240 | 0.058 | -0.650 | 5.257 | 0.178 | 1.000 |
The results of calculating the Pearson correlation coefficients between the independent variables and the dependent variable (Earnings Response Coefficient (ERC)) and their intensity based on the output from Clementine 12 software are presented in Table 3. In Clementine 12 software, correlation coefficients are classified into three groups: weak, moderate, and strong.
Variable | Symbol | Pearson Correlation Coefficient | Correlation Strength |
|---|---|---|---|
Earnings per Share | EPS | -0.202 | Strong |
Growth Opportunities | GO | 0.074 | Strong |
Return on Assets | ROA | -0.100 | Strong |
Operating Cash Flows | CFO | -0.156 | Strong |
Financial Leverage | LEV | 0.056 | Strong |
Sales Growth | SG | 0.157 | Strong |
Operating Profit Ratio | OP | -0.022 | Weak |
Information Asymmetry | ASY | 0.142 | Strong |
Systematic Risk | BETA | 0.052 | Strong |
Stock Price Synchronicity | SYNCH | 0.020 | Weak |
Delayed Stock Price Response | DELAY | -0.005 | Weak |
Market Index Return | MR | 0.145 | Strong |
Inflation Rate | IF | 0.116 | Strong |
Free Float | FF | 0.162 | Strong |
Stock Turnover Rate | STV | 0.208 | Strong |
Stock Liquidity | Liq | 0.217 | Strong |
Number of Stock Trades | STT | 0.151 | Strong |
Government Ownership | GOW | 0.050 | Moderate |
Institutional Ownership | IS | -0.123 | Strong |
Audit Committee Financial Expertise | ACFEX | -0.047 | Moderate |
Audit Quality | AQ | -0.018 | Weak |
Board Independence | BI | -0.049 | Moderate |
CEO Duality | DUAL | -0.010 | Weak |
Earnings Quality | PQ | -0.082 | Strong |
Timeliness of Reporting | FRT | -0.005 | Weak |
Auditor's Opinion Type | AO | 0.166 | Strong |
Earnings Stability | ES | 0.094 | Strong |
Financial Statement Restatements | RS | -0.042 | Weak |
Earnings Conservatism | CC | -0.009 | Weak |
Relative Strength Index per Share | RSI | 0.085 | Strong |
Psychological Line Index | PSY | 0.039 | Weak |
Adjusted Trading Volume | ATR | -0.029 | Weak |
Trading Imbalance | BSI | 0.018 | Weak |
Given that the objective of the research is to identify and rank the financial, non-financial, and behavioral variables affecting the Earnings Response Coefficient (ERC), and considering that 1530 data points were collected for 153 companies over a 10-year period (2013-2022), the data are considered big data, and data mining techniques can be utilized. Data mining is the process of extracting useful information from large datasets. It involves using statistical techniques and algorithms to discover patterns, relationships, and trends in the data. Data mining can also be used to uncover hidden insights and predict future events. Data mining uses various algorithms and statistical models to identify hidden patterns and relationships that may not be immediately apparent. It is often used to discover valuable information and make predictions based on historical data. In this research, two data mining methods, "stepwise forward regression" and "regression decision tree," were used to predict the impact of financial, non-financial, and behavioral variables on the Earnings Response Coefficient. Therefore, the mentioned models were first modeled in the Clementine 12 software environment. As observed, the data were first loaded from the Excel environment into the software, and the functions related to the descriptive statistics outputs and research variables charts were defined. Then, uninterpretable variables such as company name and fiscal year were filtered out using the Filter function, and the variables were defined based on the type of dependent and independent variables in the Type function. Subsequently, the data were randomly divided into training, testing, and evaluation groups in the Partition function at ratios of 55%, 35%, and 10%, respectively, and entered into the machine learning algorithm. Finally, the data of the three groups were entered into the "stepwise forward regression" and "regression decision tree" algorithms.
After running the data mining model using the stepwise forward regression method, the financial, non-financial, and behavioral factors affecting the Earnings Response Coefficient were ranked by importance, as shown in Table 4.
Factor Name | Symbol | Importance | Impact Rank |
|---|---|---|---|
Auditor's Opinion Type | AO | 0.202 | 1 |
Earnings per Share | EPS | 0.133 | 2 |
Stock Liquidity | Liq | 0.115 | 3 |
Growth Opportunities | GO | 0.107 | 4 |
Earnings Stability | ES | 0.103 | 5 |
Inflation Rate | IF | 0.099 | 6 |
Sales Growth | SG | 0.062 | 7 |
Operating Profit Ratio | OP | 0.052 | 8 |
Relative Strength Index per Share | RSI | 0.047 | 9 |
Market Index Return | MR | 0.038 | 10 |
After running the data mining model using the regression decision tree method, the financial, non-financial, and behavioral factors affecting the Earnings Response Coefficient were ranked by importance, as shown in Table 5.
Factor Name | Symbol | Importance | Impact Rank |
|---|---|---|---|
Earnings Stability | ES | 0.7443 | 1 |
Stock Liquidity | Liq | 0.1432 | 2 |
Earnings per Share | EPS | 0.0595 | 3 |
Stock Price Synchronicity | SYNCH | 0.0038 | 4 |
Trading Imbalance | BSI | 0.0017 | 5 |
Adjusted Trading Volume | ATR | 0.0017 | 6 |
Psychological Line Index | PSY | 0.0017 | 7 |
Relative Strength Index per Share | RSI | 0.0017 | 8 |
Earnings Conservatism | CC | 0.0017 | 9 |
Financial Statement Restatements | RS | 0.0017 | 10 |
Subsequently, the ranks of the factors identified as influential on the Earnings Response Coefficient in at least one of the two methods (stepwise forward regression and regression decision tree) were statistically compared. Given that the ranks of two dependent samples (factors in the two methods) were compared, the non-parametric Wilcoxon signed-rank test and SPSS software were used. In the Wilcoxon test, the null hypothesis claims that the ranking of factors by the two methods is equal. Therefore, if the test's significance level is less than 0.05, the null hypothesis is rejected at a 95% confidence level, indicating that the ranking of factors by the two methods is significantly different. In 9 out of 10 cases, the ranking of factors is greater based on the stepwise forward regression method than the regression decision tree method, and in one case, it is smaller, with a mean rank of 5.83 for the larger and 2.50 for the smaller. The Wilcoxon test statistic is -2.567, and its significance level is less than the error level of 0.05. Therefore, at a 95% confidence level, the null hypothesis of the Wilcoxon test is rejected, leading to the conclusion that the rankings of the stepwise forward regression and regression decision tree methods for the factors affecting the Earnings Response Coefficient are not the same.
Moreover, the mean absolute error (MAE) of the stepwise forward regression method model in the evaluation stage is 1.798, and based on Table 6, the mean absolute error (MAE) of the regression decision tree method model in the evaluation stage is 1.57. Therefore, it can be concluded that the regression decision tree method identifies and ranks the factors affecting the Earnings Response Coefficient more accurately.
Discussion and Conclusion
The response of stock prices to the disclosure of profit changes at the time of profit announcements by companies is a finding that has long been widely claimed regarding the relevance of profits for stock pricing based on positive accounting theory. There is significant consensus in the theoretical foundations that profit increases lead to stock price increases and profit decreases lead to stock price decreases.
One of the tools that can be used to measure investor reactions to accounting profit information is the Earnings Response Coefficient (ERC). The ERC estimates changes in a company's stock price as a result of the company's profit information announced to the market. The ERC is another measure for abnormal returns observed in response to unexpected elements of the company's reported profit. In other words, the ERC measures the sensitivity of stock markets to profit reports through the regression slope coefficient between abnormal returns and unexpected earnings. The ERC is very useful in fundamental analysis for calculating the intrinsic value of stocks using the company's financial data, which can serve as the basis for investor evaluations to determine market reactions to profit information in the company's stock price. Each company has a unique ERC based on how reported information affects investor reactions and stock prices. In other words, company stock prices show different reactions to one unit of unexpected earnings because the ERC is influenced by various factors related to the company's characteristics. Company characteristics are private or specific information that distinguishes the company from others, impacting the ERC.
Research on the ERC aims to identify and explain different market reactions to profit information from various companies. Various factors, including earnings quality and financial reporting, company status and performance, capital market performance, investment environment, corporate governance framework, and investor behavioral factors, cause different market reactions to profits. In this research, we attempted to identify and rank the financial, non-financial, and behavioral factors affecting the ERC of companies using data mining techniques, employing two data mining methods: stepwise forward regression and regression decision tree.
The analysis results based on the data mining technique using the stepwise forward regression method showed that among the factors related to the company's financial status and performance, the variables "Earnings per Share," "Sales Growth," "Growth Opportunities," and "Operating Profit Ratio" and among the factors related to the company's capital market status and performance, the variable "Stock Liquidity" influence the ERC. Additionally, among the factors related to the company's financial reporting characteristics, the variables "Auditor's Opinion Type" and "Earnings Stability" influence the ERC. Among the factors related to the company's investment environment, the variables "Market Index Return" and "Inflation Rate" and among the financial and behavioral factors of investors, the variable "Relative Strength Index per Share" influence the ERC. Based on the data mining technique using the stepwise forward regression method, the Auditor's Opinion Type variable has the greatest impact, and the Market Index Return variable has the least impact on the ERC.
The analysis results based on the data mining technique using the regression decision tree method showed that among the factors related to the company's financial status and performance, the variable "Earnings per Share" and among the factors related to the company's capital market status and performance, the variable "Stock Liquidity" influence the ERC. Additionally, among the factors related to the company's financial reporting characteristics, the variables "Earnings Stability," "Earnings Conservatism," and "Financial Statement Restatements" influence the ERC. Among the factors related to the company's investment environment, the variable "Stock Price Synchronicity" and among the financial and behavioral factors of investors, the variables "Trading Imbalance," "Adjusted Trading Volume," "Psychological Line Index," and "Relative Strength Index per Share" influence the ERC. Based on the data mining technique using the regression decision tree method, the variable Earnings Stability has the greatest impact, and the variable Financial Statement Restatements has the least impact on the ERC.
The results of the Wilcoxon test also showed that the rankings of the factors affecting the ERC using the stepwise forward regression method and the regression decision tree method are not the same. Furthermore, comparing the mean absolute error of the two methods indicated that the regression decision tree method identifies and ranks the factors affecting the ERC more accurately. However, the results of both methods confirm the impact of the variables Earnings per Share, Stock Liquidity, Earnings Stability, and Relative Strength Index per Share on the ERC of companies.
Therefore, in general, it can be concluded that in their reactions to the pricing of stocks based on the unexpected earnings announced by the company, investors consider Earnings per Share as an indicator of investment returns in stocks and the stability of earnings as a predictor of future earnings, making decisions based on the relevance and reliability of profits. On the other hand, stock liquidity also acts as a criterion of relevance in buy and sell decisions and the balance between supply and demand in the stock, affecting investor reactions to unexpected earnings. Additionally, positive investor sentiment towards the monthly return trend of stock prices, measured by the Relative Strength Index per Share, impacts investor reactions to unexpected earnings.
Therefore, current and potential investors are advised to pay attention to Earnings per Share and its stability, stock liquidity, and market sentiment towards the stock based on monthly returns when predicting stock pricing for buy and sell decisions at the time of profit announcements. Managers are also advised to manage real profits and smooth out Earnings per Share and create earnings stability to prevent fluctuations in Earnings per Share that lead to a high ERC for the company and increase its risk.
All authors have contributed significantly to the research process and the development of the manuscript.
Declaration
In order to correct and improve the academic writing of our paper, we have used the language model ChatGPT.
Transparency Statement
Data are available for research purposes upon reasonable request to the corresponding author.
Acknowledgments
We would like to express our gratitude to all individuals helped us to do the project.
Declaration of Interest
The authors report no conflict of interest.
Funding
According to the authors, this article has no financial support.
Ethical Considerations
In this research, ethical standards including obtaining informed consent, ensuring privacy and confidentiality were observed.
References
- Anvari, A. (2021). The Integration of LARG Supply Chain Paradigms and Supply Chain Sustainable Performance (A Case Study of Iran). Production & Manufacturing Research, 9(1), 157-177. doi 10.1080/21693277.2021.1963349
- Beamon, B. M. (1999). Measuring Supply Chain Performance. International Journal of Operations & Production Management, 19(3), 275-292. doi 10.1108/01443579910249714
- Eldin, N. A., Aidaros, H., Khalaf-Allah, S. S., Diab, M. M., & Bahgy, H. E. E. (2023). An Application Study for Monitoring and Evaluation of the Hygienic Status of Poultry Farms. Benha Veterinary Medical Journal, 43(2), 51-57. doi 10.21608/bvmj.2022.165311.1597
- Fathelrahman, E., Awad, A. I. E., Mohamed, A. I., Eltahir, Y. M., Hassanin, H. H., Mohamed, M. E., & Hoag, D. L. (2020). Biosecurity Preparedness Analysis for Poultry Large and Small Farms in the United Arab Emirates. Agriculture, 10(10), 426. doi 10.3390/agriculture10100426
- Golden, C. E., Rothrock, M. J., & Mishra, A. (2021). Mapping Foodborne Pathogen Contamination Throughout the Conventional and Alternative Poultry Supply Chains. Poultry Science, 100(7), 101157. doi 10.1016/j.psj.2021.101157
- Hafez, H. M., & Attia, Y. A. (2020). Challenges to the Poultry Industry: Current Perspectives and Strategic Future After the COVID-19 Outbreak. Frontiers in veterinary science, 7. doi 10.3389/fvets.2020.00516
- Hakeem, M. J., & Lu, X. (2021). Survival and Control of Campylobacter in Poultry Production Environment. Frontiers in Cellular and Infection Microbiology, 10. doi 10.3389/fcimb.2020.615049
- Hamed, A., & Bohari, A. M. (2022). Adoption of Big Data Analytics in Medium-Large Supply Chain Firms in Saudi Arabia. Knowledge and Performance Management, 6(1), 62-74. doi 10.21511/kpm.06(1).2022.06
- Kler, R., Gangurde, R., Elmirzaev, S., Hossain, S., Vo, N. T. M., Nguyen, V. T., & P, N. K. (2022). Optimization of Meat and Poultry Farm Inventory Stock Using Data Analytics for Green Supply Chain Network. Discrete Dynamics in Nature and Society, 2022, 1-8. doi 10.1155/2022/8970549
- Leuschner, R., Rogers, D. S., & Charvet, F. F. (2013). A Meta‐Analysis of Supply Chain Integration and Firm Performance. Journal of Supply Chain Management, 49(2), 34-57. doi 10.1111/jscm.12013
- Li, N., Ren, Z., Li, D., & Zeng, L. (2020). Review: Automated Techniques for Monitoring the Behaviour and Welfare of Broilers and Laying Hens: Towards the Goal of Precision Livestock Farming. Animal, 14(3), 617-625. doi 10.1017/s1751731119002155
- Liu, K. P., Chiu, W., Chu, J., & Zheng, L. J. (2022). The Impact of Digitalization on Supply Chain Integration and Performance. Journal of Global Information Management, 30(1), 1-20. doi 10.4018/jgim.311450
- Masud, A. A., Rousham, E. K., Islam, M. A., Alam, M.-U., Rahman, M., Mamun, A. A., Sarker, S., Asaduzzaman, M., & Unicomb, L. (2020). Drivers of Antibiotic Use in Poultry Production in Bangladesh: Dependencies and Dynamics of a Patron-Client Relationship. Frontiers in veterinary science, 7. doi 10.3389/fvets.2020.00078
- Poudel, A., Sharma, S., Dhital, K., Bhandari, S., Napit, R., Puri, D., & Karmacharya, D. (2023). Poor Biosafety and Biosecurity Practices and Haphazard Antibiotics Usage in Poultry Farms in Nepal Hindering Antimicrobial Stewardship. doi 10.1101/2023.04.17.536518
- Samaranayake, P., & Laosirihongthong, T. (2016). Configuration of Supply Chain Integration and Delivery Performance. Journal of Modelling in Management, 11(1), 43-74. doi 10.1108/jm2-01-2014-0005
- Sezen, B. (2008). Relative Effects of Design, Integration and Information Sharing on Supply Chain Performance. Supply Chain Management an International Journal, 13(3), 233-240. doi 10.1108/13598540810871271
- Sibanda, N., McKenna, A., Richmond, A., Ricke, S. C., Callaway, T. R., Stratakos, A. C., Gundogdu, O., & Corcionivoschi, N. (2018). A Review of the Effect of Management Practices on Campylobacter Prevalence in Poultry Farms. Frontiers in microbiology, 9. doi 10.3389/fmicb.2018.02002
- Ucenic, C. I., & Ratiu, C. (2017). Improving Performance in Supply Chain. Matec Web of Conferences, 137, 01018. doi 10.1051/matecconf/201713701018
- Wilujeng, S. R., Sarwoko, E., & Nikmah, F. (2022). Triple-a Strategy: For Supply Chain Performance of Indonesian SMEs. Uncertain Supply Chain Management, 10(1), 95-100. doi 10.5267/j.uscm.2021.10.007
