Financial Performance in Predicting Stock Return: Study on Non-Financial Companies Listed on IDX

Accounting and financial information are usually given to investors to help predicting the stock returns because the data has the value and historical. This study aims to examine how the ability of financial performance that represent leverage consisting of debt to asset ratio (DAR), debt to equity ratio (DER), and profitability including return on asset (ROA), return on equity (ROE), and net profit margin (NPM) in predicting stock returns. The research applies a final sample of 163 companies of non-financial sector registered in IDX in 2015-2019. This research applies analysis of multiple linear regression. The study results prove that DAR, ROE, NPM have no significant influence toward the stock returns. DER and ROA give such a significant and also negative influence on stock returns. These results mean that investors can consider DER and ROA in predicting stock returns as signals from accounting and financial data.


INTRODUCTION
Stock return is known as things expected from investors for their investment in a company for future profits. Investors need both financial and non-financial information about the company in considering their investment decisions. Investors will decide to purchase stock with the expectation of getting profits in the form of dividends and capital with the belief that their investment is guaranteed even though it is inseparable from risks (Alexander & Destriana, 2013). Investors will get a maximum return along with high risk or receive a minimal return with a small risk (Carlo, 2014).
Information related to company performance is needed to grab the investors to invest their capital (Alexander & Destriana, 2013). Contrary to the "old" hypothesis of efficient market that the returns are sometimes being unpredictable over time, but now it is commonly understood that returns on equity are claimed to be more predictable and also vary over time (Bollerslev et al., 2015). Financial data or accounting information can be implemented in realtime in the context of predictability of stock returns (Martin & Wagner, 2019). This means that investors can take advantage of accounting information or company financial information to predict the return on investment.
Predicting future returns is able to be the influencing factors of investors to make the right decisions of their investment. In line with the signaling theory proposed by Wolk et al. (2000), the reason for a company to share financial information with investors is due to information asymmetry. This is a situation when the management has more knowledge about the actual condition of the company than investors.
Predicting stock returns is interesting for academics and practitioners as it tends to have such an fundamental impacts in various aspects which are known as to help managing the asset price and also management of risk (Wang et al., 2021). Investors usually tend to use historical information of finance in order to help them predicting the future returns of stock through fundamental process of analysis. This analysis relates to the factors that company has such as the income and the expenses, their assets, the liabilities, managing experience, the profits, and the dynamics of business industry (Iqbal et al., 2013).
Basic financial ratios tend to show some predictive power for stock returns in the short and long term (Jiang & Lee, 2012). Some variables from accounting data used by other studies to predict stock returns are a debt to asset ratio or DAR, or DER, called ROA, the return on equity or also ROE, and also the net profit margin or named NPM. These financial ratios are considered factors affecting stock returns and are part of the leverage ratio (DAR and DER) and profitability (ROA, ROE, and NPM) so investors need to know how these variables can predict stock returns.
Many previous studies have discussed predictors of stock returns including DAR, DER, ROA, ROE, and NPM, for example, Andriani & Winedar (2020), Ghale (2015), Tumonggor et al. (2017), Jais et al. (2012), Mahardika & Artini (2017), Öztürk & Karabulut (2018), Tikasari & Surjandari (2020), Ristyawan (2019), Rohpika & Fhitri (2020), Simorangkir (2019), Hasanudin et al. (2020), and Yuliaratih & Artini (2018). They tested the influence of some variables including DAR, DER, ROA, ROE, and NPM on stock returns and showed inconsistent results. In addition, previous studies only used company data from certain sectors. This study will be different from previous research because it uses data from non-financial companies covering various industrial sectors on the stock exchange. According to Fjesme (2020), non-financial companies have a highly concentrated investment portfolio where this concentration plays a role in increasing risk-based returns. Researchers see the growth of investors in Indonesia as significant, so that by analyzing nonfinancial companies will better describe the condition of the industry and is expected to avoid investment decision mistakes for investors.
The research tries to focus on companies of non-financial sector registered on the IDX from 2015 to 2019 as samples because financial companies usually show different types and cultures from other companies because they focus on financial activities so they have differences in financial performance analysis (Marwati & Yulianti, 2015). The objective of this study is obtaining the empirical evidence of the influence of DAR, DER, ROA, ROE, and also NPM on stock returns in non-financial companies. This study is done on hope to help contributing to the development of literature and financial accounting related to the ability of accounting data history informations or ratios of finance in order to help predicting stock returns, especially in companies of non-financial sector registered on the IDX. This research will also make a practical contribution to shareholders and potential investors who will start to invest in the capital market by considering accounting information or financial performance. The results of this study are also expected to give recommendations for management or companies to pay more attention in provide the good and right financial information which can be a signal for investment decisions for investors and potential investors.
Financial performance analyzed through fundamental analysis will provide information to investors and potential investors in the form of good or bad signals of the financial condition of a company. A positive signal will affect how stock prices can increase. The increase in stock prices is an expectation for investors as a return from the shares invested. Signals from financial performance to predict stock returns are explained by signaling theory. The signaling theory states that decisions and communication of financial information are signals directed from management to investors to minimize information asymmetry and facilitate decisions of rational investment (Goetzmann et al., 2014). These signals are the basis of financial communication policy. Companies provide financial information to investors to solve the information asymmetry problems between management and investors because management knows more information about the company than investors (Wolk et al., 2000).
The signaling theory is based on the premise that the availability of information is not evenly distributed to all stakeholders at the same time. Managers usually are privy to consistent, accurate, and relevant personal information that informs their decisions which can ultimately signal to the market (Otieno & Ochieng, 2015). As the investors usually tend to maximize their stock returns, investors will invest only in companies that are predicted to have stable future performance (Quiry et al., 2022).x DAR is known as a part of the leverage ratio that is useful in providing information related to the level of guarantee for creditors on the return of their investment. A low debt-to-asset ratio indicates that the company can increase profits to make guarantees for loan repayments for creditors (Hadi, 2015). The higher the DAR level, the greater the risk, and investors are also expecting high returns. It is linier with the research by Andriani & Winedar (2020) revealed that DAR gives negative and also significant influence on the variable of stock returns. When the DAR is higher, the stock return is lower. However, Ristyawan (2019), reveals that DAR gives such a significant and also positive influence toward the stock returns. Meanwhile, Rohpika & Fhitri (2020) indicates that DAR has no significant influence toward the stock returns. The explanation above becomes the basis for researchers to formulate the first hypothesis for this study, namely. H1: Debt to asset ratio (DAR) has a negative effect on stock returns.
The leverage ratio can also be seen from DER. Investors can use this ratio to see how their invested equity becomes a guarantee for company loans (Tumonggor et al., 2017). A high DER indicates that the company uses debt as a source of capital and it shows a riskier investment so that investors prefer to avoid investing in the company and resulting in a fall in stock prices (Jais et al., 2012). This is in line with Jais et al. (2012), Hasanudin et al. (2020), and Rohpika & Fhitri (2020) that DER gives negative and also significant influence toward the stock returns. The higher the DER, the lower the stock return. On the other hand, Rachmawati & Rahayu (2017) revealed that DER gives a significant and also positive influence toward the stock returns. Supriadi (2015) and Tumonggor et al. (2017) show that DER has no effect on stock returns. The explanation above becomes the basis for researchers to formulate the second hypothesis for this study, namely. H2: Debt to equity ratio (DER) gives a negative influence toward the stock returns.
Looking from the profitability ratios is the return on assets (ROA). Investors prefer to invest in companies with high ROA values. High ROA indicates a high management's ability to earn profits through high asset values (Heikal et al., 2014). In line with Heikal et al. (2014) and Yuliaratih & Artini (2018), ROA gives such a positive and also significant influence toward the stock returns. When the ROA is high, the stock return is also getting high. But in the other side, Simorangkir (2019) shows that it is found a significant but negative influence of ROA toward the stock returns. Putro (2020) reveals that ROA gives no significant influence toward the stock returns. Those explanation then becomes the basis for researchers to formulate the third hypothesis for this study, entitle. H3: Return on assets (ROA) gives a positive influence toward the stock returns.
ROE is claimed as one of the profitability ratios. This ratio is a good predictor of stock returns (Banerjee, 2019). ROE shows that management can manage investor equity to be turned into profit. The high ROE of a company can increase demand for stock of the company and it is going to influence the price of the stock in the market of capital. An increase in prices of stock also makes stock returns increase. It is supported by Ghale (2015), Simorangkir (2019) and Mahardika & Artini (2017) that it is found a significant and also positive effect of ROE and also stock returns. When the ROE is high, the stock return is also becoming high. But it is different with Ristyawan (2019) which revealed that ROE gives such a negative and also significant effect toward the stock returns. Tumonggor et al. (2017) reveal that ROE shows no significant influence toward the stock returns. The explanation above becomes the basis for researchers to formulate the fourth hypothesis for this study, namely. H4: Return on equity (ROE) gives a positive influence toward the stock returns.
The company's profitability is also able to be seen from the net profit margin (NPM). This ratio shows that a high NPM value can be found in companies that have high-profit growth which is obtained from the company's high sales (Heikal et al., 2014). When the company can earn a profit from high sales, it makes investors feel confident in the expected stock return. Öztürk & Karabulut (2018), Yuliaratih & Artini (2018), Simorangkir (2019) and Tikasari & Surjandari (2020) prove that it is found a positive and also significant influence of NPM and stock returns. This indicates that when the NPM is higher so is the stock return. The different result comes from Hasanudin et al. (2020) show that it is revealed a negative and also significant effect of NPM and stock returns. Ristyawan (2019) shows that NPM gives no significant influence toward the stock returns. The explanation above becomes the basis for researchers to formulate the fifth hypothesis, which is. H5: Net profit margin (NPM) gives a positive influence toward the stock returns.

RESEARCH METHODS
This study applies quantitative research that uses secondary data from nonfinancial sector of public companies registered on the IDX. This research also used financial data and company stock data. Data were taken from the OSIRIS database and the website of IDX. The observation period was 5 years from 2015 to 2019. The determination of the observation period was based in the middle of the Covid-19 pandemic since 2020 that caused changes in market conditions. The population was the entire companies regitered on the IDX during the year of 2015-2019. This study involved companies of non-financial sector and banking registered on the IDX in 2015-2019 as samples. This study excludes financial and banking companies because these companies concentrate on financial activities so that they differ in According to Khan et al. (2021) defined the debt to asset ratio as total liabilities is divided by total assets DER Debt to equity ratio ℎ ℎ Source: Jais et al. (2012) According to Jais et al. (2012) defined the debt to equity ratio as total debt is divided by shareholder equity This study covered five independent variables including DAR, DER, ROA, ROE, and NPM and one dependent variable, namely RET. The test of classical assumption was performed to ensure the ideal model in order to help testing the hypothesis. Table 3 disclosed the result of the test of One-Sample Kolmogorov-Smirnov showed a significance value of 0.016 or <0.05. It was also the test of Monte Carlo resulting 0.253 of significance value that it is stated that the data on all variables are distributed in normal way because the assumption of normality has been fulfilled.  Table 4 showed the results of VIF values are 1.242 for DAR, 1.275 for DER, 1.685 for ROA, 1.822 for ROE, and 1.027 for NPM. The value of VIF of all independent variables is revealed to be lower than 10 which indicates that there is not found any multicollinearity symptom in all independent variable data because they have fulfilled the multicollinearity assumption. Looking from the Table 5, Spearman's value of the DAR, DER, ROA, ROE, and NPM variables was higher than 0.05. Thus, the data on the dependent variable are free from heteroscedasticity symptoms.  After confirming that the regression model in the study has fulfilled the criteria for the classical assumption test, the model was ideal for testing the hypothesis. The analysis of multiple linear regression was applied in order to help determining the effect of DAR, DER, ROA, ROE, and also NPM toward the stock returns in non-financial companies registered in IDX. The results is seen at Table  7. Looking from the Table 7, the equations for multiple linear regression is made as follows.
RET = 0,019 -0,050DAR -0,006DER -0,307ROA +0,052ROE -0,001NPM + Based on Table 3, the Sig. F value was 0.025 or <0.05 which means that the variables DAR, DER, ROA, ROE, and NPM gives a simultaneous influence toward the dependent variable, called as RET or stock return. Then, the effect between the independent and dependent variable as follows: (1) The DAR significance value was shown 0.258 or > 0.05 and the t-test result of -1.131. it can be stated that H1 is getting rejected and H0 is getting accepted or the DAR gives no significant and also negative toward the RET. (2) The DER gives a significance value of 0.019 or <0.05 and the t-test is -2.356. it is stated that H0 is getting rejected and the H2 is getting accepted or the DER gives a significant and negative influence toward the RET or the stock returns. (3) The ROA gives a significance value 0.016 or <0.05 and the t-test is -2.425. it is revealed that H3 is getting accepted and H0 is getting rejected so the ROA gives a significant but negative influence toward the RET. (4) The ROE gives a significance value that is 0.074 or > 0.05 with a t -test result of 1.791. it can be said that H0 is getting accepted and H4 is getting rejected so that the ROE gives no significant and also positive influence toward the RET. (5) The NPM gives a significance value that is 0.138 or > 0.05 with a t-test result of -1.485. This shows that H0 is getting received and H5 is getting rejected or the NPM gives no significant but positive influence toward the RET.
The results show that H1 is rejected or DAR gives no significant and also negative influence toward the stock returns in non-financial companies registered in IDX in 2015-2019. This result is contrary to signaling theory or high DAR does not necessarily indicate bad news from stock returns. This result contradicts Andriani & Winedar (2020) and Ristyawan (2019). But, the results is supported by Rohpika & Fhitri (2020) that DAR has no significant effect on stock returns. Rohpika & Fhitri (2020) explain that the debt owned by the company is able to consider as the investors' advantage in the future through the addition of infrastructure financed by external parties. This is because the high of debt can make management to use the financed infrastructure to increase company profits. Thus, investors do not view debt as a threat or an accounting signal to predict stock returns. In other words, DAR gives no effect on the stock returns so DAR is not a factor or predictor of stock returns.
The results showed that H2 was accepted. Thus, DER gives such a significant but also negative influence toward the stock returns in non-financial companies registered in IDX in 2015-2019. This result is in line with signaling theory or high DER can be bad news for stock returns. This result contradicts with Rachmawati & Rahayu (2017), Supriadi (2015), and Tumonggor et al., (2017). But, this study support Jais et al. (2012), Hasanudin et al. (2020), and Rohpika & Fhitri (2020). Jais et al. (2012) explain that a better stock return can be generated from a mature group of companies than from a new and stable group of companies. Another possibility is that mature company are more financially stable than newly founded companies. This means that a mature company can provide higher stock returns. Whereas immature companies tend to choose debt as a source of financing instead of relying solely on shareholder investment. Besides, companies with high debt will be seen as a risk for investors because companies tend to settle their debtrelated obligations first. The research shows that DER can be an accounting signal for investors and a predictor of stock returns of non-financial companies. This research revealed that the lower the DER, the stock return is getting higher, and also vice versa.
The results also revealed the discovery that H3 was getting accepted. The results showed that ROA had a significant negative effect on stock returns in nonfinancial companies registered in IDX in 2015-2019. Although not in accordance with the initial assumption regarding the direction of its influence, the results of this study are in line with signaling theory where ROA as one of the financial performance information can be news that affects stock returns. The results of this study are not in line with Heikal et al. (2014) and Yuliaratih & Artini (2018) that ROA has a significant positive effect on stock returns. However, this study supports Simorangkir (2019) indicates that ROA gives such a significant and also negative influence toward the stock returns.
Simorangkir (2019) explains that ROA gives a negative coefficient as the company cannot show the equity invested by investors for all assets to earn profits. This shows that the company is less effective to help utilizing the assets to earn a advantage which ultimately declines the company's income resulting in a decline in stock returns. These results indicate that ROA can be an accounting signal or predictor of stock returns. When the ROA is lower, the stock return is higher and vice versa.
The other results also show that H4 was rejected. Thus, ROE gives no significant but positive influence toward the stock returns in non-financial companies registered in IDX in 2015-2019. This result is contradicts to signaling theory or high ROE does not necessarily indicate good news from stock returns. The study reveals the different result with the Ghale (2015), Simorangkir (2019), Mahardika & Artini (2017), and Ristyawan (2019), but support the findings of Tumonggor et al. (2017) discovers that ROE gives no significant influence toward the stock return. Tumonggor et al. (2017) explain that when the company's ROE is low, the company cannot guarantee investors' equity. ROE of newly established companies tends to increase rapidly as profits grow. The difference in the company's growth period may be an insignificant result for ROE toward the stock returns. This makes high ROE uncertain be a good news as signaling theory states. This research reveals that ROE is not seen as an accounting signal to predict stock returns, so ROE cannot be a factor in predicting stock returns.
The results showed that H5 was rejected. Thus, NPM gives no significant influence toward the stock returns in non-financial companies registered in IDX in 2015-2019. This result is contrary to signaling theory or high NPM does not necessarily indicate good news from stock returns. The results of this study contrast with Öztürk & Karabulut (2018), Yuliaratih & Artini (2018), Simorangkir (2019), Tikasari & Surjandari (2020), and Hasanudin et al. (2020). However, this study supports the findings of Ristyawan (2019) that NPM gives no significant effect on stock return. Ristyawan (2019) explains that the NPM does not influence the stock returns because the high NPM value in the sample companies is not in line with high stock returns. This happens because the net profit between periods of the company can be influenced by the volatility of market conditions and other factors. This makes NPM unable to become a benchmark as good news for stock returns as expressed in signaling theory. The study indicates that NPM is not one of significant factors to help predicting the stock returns. Thus, the value of NPM does not make changes in stock returns and NPM cannot be a predictor of stock returns.

CONCLUSION
This study involves 163 non-financial sector of companies registered on IDX in 5 years of observation from 2015 to 2019 as samples. The study indicates that the entire independent variables gives silmultanously a significant influence toward the stock returns. Partially, DER gives a significant negative influence toward the stock returns in non-financial sector companies in 2015-2019. This results are in line with signaling theory and indicates that DER can be a factor as an accounting and financial signal in predicting stock returns. The lower the DER value, the stock return is higher and vice versa. The results also revealed that ROA shows such a significant but also negative influence toward the variable of stock returns in nonfinancial sector companies in 2015-2019. However, this study proves that the company's ineffzectiveness to help utilizing the assets to earn a profit can be an accounting and financial signal in predicting stock returns as signaling theory stated. Thus, the lower the ROA, the higher the stock return and vice versa. Meanwhile, DAR, ROE, and NPM give no significant influence toward the stock returns in non-financial sector companies in 2015-2019. This concerns the importance of classifying certain characteristics of the sample companies in the test in order to obtain significant results in finding predictors of stock returns.
For shareholders, before deciding on their investment, they are expected to conduct a thorough analysis through fundamental analysis using information on financial performance ratios that have been proven in the study to predict stock returns. This is useful to ensure that the expectations of the investment are right on target. Future researchers are expected to include other variables that may give impact to the stock returns by classifying the characteristics of the company to be tested, such as the companies' growth phase or the size of them in the test.