Journal of Accounting and Investment Vol. 24 No. 2, May 2023 Article Type: Research Paper Do dividends still matter? The role of investment opportunities on the ability of dividends to predict future earnings Sila Ninin Wisnantiasri Abstract Research aims: This study investigates the role of firm characteristics explained by the investment opportunity (IO) on the ability of dividends to predict future earnings. Design/Methodology/Approach: This study performed an empirical study on firms listed in the consumer goods sub-sector on the Indonesian Stock Exchange, divided into companies with strong and weak IO categories to clearly see the role of IO by comparing the variable dividend coefficients of the two sample categories. Through purposive sampling, the researcher determined the research sample, totaling 42 firm samples for the weak IO category and 48 firm samples for the strong IO category. Then, the multiple regression analysis utilizing IBM SPSS Statistic Version 23 was employed to analyze the relationship between variables. Research findings: Surprisingly, companies with weak IO showed a more remarkable ability to predict future earnings than companies with strong IO because the dividend coefficient of companies with weak IO was higher than that of strong IO, denoting that the number could explain the strength of ability. Theoretical contribution/Originality: The result provides alternative explanations to the previous inconsistent results from the dividend's ability to predict future earnings. The result also supports the argument that the companies with weak IO may use dividends to convey information signals and compensate the investor for unsatisfied performance, which is called counter-signal when strong IO refrain from doing so and rely on additional information. Practitioner/Policy implication: Investors should notice companies' characteristics, such as investment opportunities, while considering dividends as a signal for future performance to make an investment decision. Research limitation/Implication: The research did not fully capture all companies in Indonesian Stock Exchange, but specifically for the companies’ sub-sector that aggressively paid the dividend. Thus, future research is hoped to provide empirical studies for other sector companies listed on Indonesia Stock Exchange to enrich alternative explanations. Keywords: Dividend; Dividend signaling; Investment opportunities; Future earnings Introduction In classical dividend theory, the information provided by dividends increases with the investment opportunities (IO) (Rock & Miller, 1985). However, recent research has uncovered differences in signaling impact between weak and strong IO companies. Dividends can signal stronger for AFFILIATION: Accounting Study Program, Faculty of Economics, Universitas Terbuka, Banten, Indonesia CORRESPONDENCE: sila.wisnantiasri@ecampus.ut.ac.id DOI: 10.18196/jai.v24i2.17016 CITATION: Wisnantiasri, S. N. (2023). Do dividends still matter? The role of investment opportunities on the ability of dividends to predict future earnings. Journal of Accounting and Investment, 24(2), 519-532. ARTICLE HISTORY Received: 01 Dec 2022 Revised: 26 Jan 2023 02 Feb 2023 Accepted: 22 Feb 2023 This work is licensed under a Creative Commons Attribution-Non-Commercial- No Derivatives 4.0 International License JAI Website: https://scholar.google.co.id/citations?user=doeUMi8AAAAJ&hl=en https://fe.ut.ac.id/program-studi-akuntansi/ https://fe.ut.ac.id/program-studi-akuntansi/ https://fe.ut.ac.id/program-studi-akuntansi/ mailto:sila.wisnantiasri@ecampus.ut.ac.id http://dx.doi.org/10.18196/jai.v24i2.17016 https://crossmark.crossref.org/dialog/?doi=10.18196/jai.v24i2.17016&domain=pdf https://journal.umy.ac.id/index.php/ai/article/view/17016 Wisnantiastri Do dividends still matter? The role of investment opportunities … Journal of Accounting and Investment, 2023 | 520 companies with weak IO (Kaplan & Pérez-Cavazos, 2022); on the other hand, companies with strong IO rarely share dividends. Companies with weak IO also tend to require less cash so that dividends can be distributed in a larger number and proportion of profits. Meanwhile, companies with strong IO prefer to reserve their profits (Kaplan & Pérez-Cavazos, 2022). Berkshire Hathway, led by Warren Buffet, chose not to distribute dividends because they prefer to reinvest the company’s net income and believe that in the long term, investors will be more profitable even if they do not receive dividends (Wijayanti, 2022), increasing value of the company. In assessing investment opportunities in various financial literature, Tobin’s Q proxy (Tobin, 1969) based on market value has been widely used (Erickson & Whited, 2011). Moreover, leading technology companies like Google, Facebook, and Amazon are predicted to share dividends in the next few years. However, they had an accumulative cash value of USD 290 billion in 2020 (Said, 2021). They used their capital to invest in long- term growth. In this case, the value of cash holdings becomes one of the proxies in investment opportunities because it reflects the company’s ability to maintain cash used to finance investments (Kaplan & Pérez-Cavazos, 2022). On the other hand, in Indonesia, several companies continue to distribute dividends even though they are in a loss position, such as PT Saratoga Investama (Wareza, 2019), PT Indika Energy Tbk (Ulfah, 2020), and PT Suya Toto Indonesia Tbk (Rahmawati, 2021). They shared dividends from retained earnings in previous years or generated income from other entities. In recent years, many companies have also postponed dividend payments due to the impact of the pandemic. Dividends themselves are considerations in choosing an investment depending on the type of investor. Particularly, short-term investors expect large dividends (Melani, 2021). Meanwhile, the survey revealed that Indonesian people invested in stocks for long-term investment (Lidwina, 2021). These investors do not mind smaller dividends if the company keeps expanding. The dividend yield itself is also one of the investor’s considerations since it provides a large rate of return. In companies with high dividend yields, company value is attributable to dividend growth (Kaplan & Pérez- Cavazos, 2022). In Indonesia, companies with the highest dividends are also indexed in the “IDX High Dividend 20.” This index measures the price performance of 20 companies that have distributed cash dividends for the last three years and have high dividend yields. At the end of 2019, 20 companies were included in this index, and 25% came from the consumer goods sector, as shown in Table 1, which are companies with stock codes UNVR, GGRM, HMSP, KLBF, and INDF. It indicates that companies in this sector are the most aggressive in paying dividends. The number of Indonesian investors has proliferated since 2018 (PT Kustodian Sentral Efek Indonesia, 2021). As of February 2021, the number of investors has reached more than 4 million. Indonesian investors are currently dominated by local investors (Olavia, 2022). Unfortunately, the behavior of individual investors in the Indonesian capital markets tends to be irrational in investing (Sumani et al., 2017). The first ranking variable consideration in investment decisions is news or reviews in the media, while dividends are ranked 21 of 33 criteria (Christanti & Mahastanti, 2011). It raises questions about Wisnantiastri Do dividends still matter? The role of investment opportunities … Journal of Accounting and Investment, 2023 | 521 whether Indonesian investors correctly take the company’s signals. Additionally, research on dividends in Indonesia still focused on the effect of dividend policy or changes in dividends on profits or changes in profits with inconsistent results (Anggara, 2020; Gultom, 2018; Widsatrya & Subroto, 2014), but no one provided an explanation of the company’s characteristics based on investment opportunities. Therefore, this research seeks to provide alternative explanations for these inconsistencies. This study used Tobin’s Q as the primary measurement of investment opportunities because it has been widely used in the financial literature (Kaplan & Pérez-Cavazos, 2022). Table 1 IDX High Dividend 20 Code Earnings Dividends Payout (%) FY15 FY16 FY17 FY18 FY19 FY15 FY16 FY17 FY18 FY19 ITMG 63 131 253 262 129 100 100 100 100 80 PTBA 2,036 2,006 4,476 5,024 4,057 30 30 75 75 75 LPPF 1,781 2,020 1,907 1,097 1,367 70 70 70 85 0 ADRO 152 335 483 418 404 50 30 53 48 48 TLKM 15,489 19,352 22,145 18,032 60 70 75 90 UNTR 3,853 5,002 7,403 11,126 11,312 66 40 45 40 40 INDF 2,968 4,145 4,168 4,166 4,908 50 50 50 50 50 HMSP 10,363 12,762 12,671 13,538 13,722 100 98 99 101 100 GGRM 6,436 6,677 7,754 7,792 10,881 78 75 65 64 46 BBRI 25,398 26,228 28,997 32,351 34,373 30 40 45 50 60 BMRI 20,335 13,807 20,640 25,015 27,482 30 45 45 45 60 INTP 4,357 3,870 1,860 1,146 1,835 35 88 139 177 80 UNVR 5,852 6,391 7,005 9,109 7,393 100 100 100 99 100 ASII 14,464 15,156 18,881 21,673 21,707 50 45 40 40 40 CPIN 1,837 2,213 2,465 4,6002 26 42 37 42 TOWR 2,958 3,040 2,100 2,200 0 23 57 55 PGAS 5,256 304 1,990 3,454 939 42 45 39 40 40 BBNI 9,067 11,339 13,616 15,015 15,384 25 35 35 25 25 KLBF 2,004 2300 2,404 2,457 2,507 44 45 49 50 50 BBCA 18,019 20,606 23,310 25,855 28,565 22 24 27 32 20 Source: Bloomberg, IDX, MiraeAsset Sekuritas Indonesia Research Based on the background of the problems, the problem can be formulated on how the influence of investment opportunities on the ability of dividends to predict earnings levels by using investment opportunities proxies, Tobin’s Q (Kaplan & Pérez-Cavazos, 2022). Further, this research was conducted to provide benefit as a literature that describes the role of investment opportunities in the ability of dividends to predict future earnings levels. It allows stakeholders, especially investors, to map the company’s condition based on investment opportunities. Then, they can conclude whether dividends are the right signal to predict company performance in the future so that the decision-making process increases accuracy due to relevant instruments. The researcher focused on consumer goods companies since they aggressively pay dividends, as presented in Table 1. Wisnantiastri Do dividends still matter? The role of investment opportunities … Journal of Accounting and Investment, 2023 | 522 Literature Review and Hypotheses Development In the early study, Miller and Modigliani (1961) found that dividend policies are important factors in determining stock prices, so companies are expected to manage dividend policy properly. Besides, the signaling theory was first introduced by Spence (1973) to explain the job-market signaling model. Then, the idea was developed around the idea that profits convey information and companies incur costs to convey that information to the market. Companies with strong profits can bear these costs, whereas less strong companies cannot (Bhattacharya, 1979; DeAngelo & DeAngelo, 2006). Based on signaling theory, researchers (e.g., Bhattacharya, 1979; John & Williams, 1985; Kaplan & Pérez-Cavazos, 2022; Rock & Miller, 1985) support the notion that dividend policy allows firms to signal to the market the true type of firms based on their future perspectives. The announcement of an increase (decrease) in the dividend rate will mean that the company is making improvements (decrease) in the company’s prospects. As a result, the market usually is sensitive to this decision by revising stock prices. However, the idea that markets react to dividends because of earnings information is rejected by researchers in favor of alternative explanations (Allen & Michaely, 2003; DeAngelo & DeAngelo, 2006; Farre-Mensa et al., 2014). The rejections of this point of view stem from two reasons: several studies have concluded that dividends do not convey information about the earnings (Grullon et al., 2005), and the results of studies on less mature companies are increasingly facing signal costs (John & Williams, 1985). Specifically, companies with weak investment opportunities need to hold less cash to manage dividends more aggressively and deliver a larger portion of permanent income to the market. Conversely, companies with strong investment opportunities prefer to save a portion of permanent income that could otherwise be used to pay dividends. From an informational perspective, this conservative dividend policy limits investors' inferences about dividend earnings. In Indonesia, the findings of dividend role in predicting future earnings still bring inconsistent results. Dividend changes are related to the company’s earnings changes in the future (Gultom, 2018; Prasetyanta, 2014) and stock prices (Riyani & Andriana, 2019; Widsatrya & Subroto, 2014). Rejecting roles in predicting future earnings have also been found (Anggara, 2020). Even though the result also rejects dividends payment in financial performance, companies that offer low dividend rates positively impact the company's financial performance (Nuriksani & Sari, 2022). These findings suggest another explanation for inconsistent results from the characteristics of companies. Because of inconsistent results of dividend roles, the researcher attempted to find an alternative explanation, as Kaplan and Pérez-Cavazos (2022) promoted, that differentiates companies from investment opportunities and compare the result from different group companies. The researcher also considered the COVID-19 pandemic influencing firms' earnings. Through the development of these hypotheses, the following are research frameworks depicted in Figure 1. Wisnantiastri Do dividends still matter? The role of investment opportunities … Journal of Accounting and Investment, 2023 | 523 Figure 1 Research Frameworks The research equation is described in the equation: 𝐹𝐸𝑡 = 𝛼0 + 𝛼1𝐷1 + 𝛼2𝐸𝑡 + 𝛼3𝑁𝑒𝑔𝐸𝑡 + 𝛼3𝐴𝐶𝑡 + 𝑎4𝐶𝑜𝑣𝑡 + ∈ (1) The explanation for each component of the equation is as follows: FE stands for future earning which is based on next year's net income (t+1) where dividends payment and earnings in the corresponding year (t) are denoted by the symbols D and E. NegE is a dummy variable for earnings where AC is an accrual derived from the difference between net income and cash flow from operations and Cov is the COVID-19 pandemic. Based on the description, the hypothesis in this study was put forward as follows: H1: Companies with weak investment opportunities have a stronger predictive ability on earnings levels than those with strong investment opportunities. H0: Companies with weak investment opportunities have no stronger predictive ability on earnings levels than those with strong investment opportunities. Research Method Population, Samples, and Data The population of this study was all companies in the consumer goods subsector listed on the Indonesia Stock Exchange during 2018-2020 and the most aggressive companies in sharing dividends. In addition, Indonesia was chosen as the population because of the specific characteristics of Indonesian investors that tend to be irrational when investing (Sumani et al., 2017). It will provide different explanations for the enrichment of alternative elucidations. Data resources were supported by the Indonesian Capital Market Institute (TICMI). Wisnantiastri Do dividends still matter? The role of investment opportunities … Journal of Accounting and Investment, 2023 | 524 Research Model and Variables This study used the quantitative method to investigate the role of investment opportunities in the relationship between dividends ability and future earnings level from each firm category. Therefore, the researcher compared the dividend coefficient of each category to draw conclusions about which category had a stronger ability. First, the researcher categorized firms into strong IO and weak IO through the value of Tobin’s Q in each firm. Second, the researcher run hypothesis tests for model 1 with strong IO company samples and model 2 with weak IO samples. Data Collection Technique Then, this study employed the purposive sampling criteria, including (1) consumer goods companies listed on the Indonesia Stock Exchange (IDX) in the study period (2018-2020) and (2) the companies with complete information regarding this study’s requirement for analysis. Based on the criteria above, the samples obtained were from 48 firm years for companies with strong IO and 42 firm years for companies with weak IO. Using the quartile method, the firms were then categorized into three groups based on Tobin’s Q value: weak IO, medium IO, and strong IO. The researcher also used only two groups, strong IO and weak IO, to analyze. Operational Definition and Measurement of Variables Operational variables used in this study were as utilized in (Kaplan & Pérez-Cavazos, 2022) and modified with COVID-19 pandemic variables as presented in Table 2. Data Analysis This study utilized the IBM SPSS Statistic Version 23. SPSS is a computer package specializing in quantitative data analysis and is widely used by market researchers (Mooi & Sarstedt, 2019). Then, the researcher used multiple regression to analyze data because the equation included multiple independent variables. First, regression analysis needed data requirements to determine if regression analysis could be used through assumptions of normality, multicollinearity, heteroscedasticity, and autocorrelation tests. To test the hypothesis, the researcher run tests with regression analysis to predict causalities correlations in model 1 for companies with strong IO and model 2 for companies with weak IO. Finally, the researcher compared the dividends coefficient between model 1 with strong IO samples and model 2 with weak IO samples to find whether the criteria had a stronger ability to predict future earnings. The steps taken in the regression analysis (Mooi & Sarstedt, 2019) are presented in Figure 2. Wisnantiastri Do dividends still matter? The role of investment opportunities … Journal of Accounting and Investment, 2023 | 525 Results and Discussion Descriptive Analysis Before running a regression analysis to test the hypothesis, the researcher prepared data requirements, including sample size, to be acceptable. The object of this research was consumer goods sector companies selected based on predetermined criteria. The researcher then divided firm years of consumer sector companies in 2018-2020 into three categories based on investment opportunities (IO) proxies by Tobin’s Q score: strong IO, medium IO, and weak IO. Then, the researcher used only strong IO and weak IO for better comparison. After removing outlier data, the final samples were from two categories: model 1 for strong IO companies amounting to 48 firm years and model 2 for weak IO companies as many as 42 firm years. Table 2 The Operationalization of Variables Variable Measurement Data Resources Category Reference Investment Opportunities (IO) Consisting of Tobin’s Q (Q) proxies: the natural logarithm of 1 + the market of equity plus total liabilities divided by total assets LOG(1+MVE+LT)/AT) The Indonesia Capital Market Institute (TICMI) Dependent Variable Future Earning (Earning levels (Et+1))) Net income in year+1 The Indonesia Capital Market Institute (TICMI) Independence Variable Dividend (D) Dividend payment during the year The Indonesia Capital Market Institute (TICMI) Control Variables Earning (E) Net income The Indonesia Capital Market Institute (TICMI) NegE Indicator variable that assumes the value of 1 if net income is lower than zero Based on net income collected from the Indonesia Capital Market Institute (TICMI) Accruals (AC) Difference between net income and cash flows from operations (NI- OANCF) The Indonesia Capital Market Institute (TICMI) COVID-19 Indicator variable that assumes the value of 1 for the existence of the COVID-19 pandemic and the value of zero for otherwise Based on the year declaration of the COVID- 19 pandemic from the WHO and Indonesia Government In Table 3, the amount value of the dividend paid in both categories obtained a minimum value of 0, meaning that the firms did not pay the dividend in the observed period. In these models, the researcher included dividend 0 payment since the researcher not only compared the magnitude of dividends but also emphasized the existence of dividend signaling. Kaplan and Pérez-Cavazos (2022) also included dividend 0 firms in their study. Wisnantiastri Do dividends still matter? The role of investment opportunities … Journal of Accounting and Investment, 2023 | 526 In this regard, the researcher assumed that not paying dividends was a signal to future earnings. In addition, the maximum dividend value paid was 13.935 billion in strong IO categories. The strong IO companies also had a greater mean of dividends of 1.378 billion. Besides, the standard deviation of strong IO was higher than the weak IO companies, signifying that deviations were broader. Figure 2 Steps Taken in Regression Analysis On the other hand, the earnings in weak IO companies had a minimum value of -203 and a maximum value of 834, smaller than strong IO. It denotes that weak IO had more loss and less profit company. The means of earning and standard deviation of weak IO were also much smaller than strong IO, showing that strong IO had a greater profit average. It was strengthened with the mean value of NegE greater than strong IO. Hence, more NegE implies that companies had more losses. Furthermore, the mean of strong IO's accruals value was much greater than weak IO, indicating that strong IO companies had positive operating cashflows. In contrast, weak IO companies had negative accruals mean value. Therefore, the future performance of strong IO companies is brighter than weak IO since the mean value of strong IO is much greater than weak IO. According to the regression analysis assumption test described in Table 4, the researcher could conclude that both model 1 in companies with strong IO and model 2 in companies with weak IO had normal data, no collinearity problem, homoscedasticity data, and no autocorrelation. It implies that regression models were ready to interpret. Wisnantiastri Do dividends still matter? The role of investment opportunities … Journal of Accounting and Investment, 2023 | 527 Table 3 Descriptive Statistics for Companies with Strong IO and Weak IO Variables Minimum Maximum Mean Standard Deviation Companies with Strong IO (N=48) Dividend (billions) 0 13.935 1.378 3.263 Earning (billions) (42) 13.722 2.132 3.478 NegE 0 1 0.13 0.33 Accruals (billions) (8.098) 12.450 139 3.130 Future Earning t+1 (billions) (84) 8.581 1.184 2.159 COVID 0 1 0.31 0.46 Companies with Weak IO (N=42) Dividend (billions) 0 551 48 97 Earning (billions) (203) 834 50 155 NegE 0 1 0.36 0.48 Accruals (billions) (657) 1.869 (33) 364 Future Earning t+1 (billions) (509) 1.035 115 279 COVID 0 1 0.33 0.48 Hypothesis Test Results After evaluating the models, the researcher tested the hypothesis by interpreting multiple regression results to answer the research question. Hypothesis testing was carried out on companies with strong IO and then on companies with weak IO. The regression results for testing are demonstrated in Table 5. First, the researcher interpreted the regression model by examining the model fit through F-test. Table 5 reveals that both model 1 and model 2 had p-values of 0.000, clearly lower than 0.05. Then, the researcher interpreted the model’s R2. Model 1 for companies with strong IO showed an R-Square value of 0.572 or 57.2%, and model 2 for companies with weak IO had an R-Square value of 0.483 or 48.3%. Both model 1 and model 2 were moderate but still satisfactory. On companies with strong IO, the significance test on multiple regression by IBM SPPS version 23 was carried out on the effect of dividend (D), earnings (E), NegE, accruals (AC), and COVID (Cov) on future earnings level (FE). It indicates that with a significance level of 5%, the variables that showed a significant relationship among companies with strong IO were dividends (D) with a Sig value of 0.034, earnings (E) with a Sig value of 0.004, and accruals (AC) with a Sig value of 0.012. Furthermore, the regression results on companies with weak IO by IBM SPSS version 23 were also conducted on the effect of dividend (D), earnings (E), NegE, accruals (AC), and COVID (Cov) on future earnings level (FE), as obtained in Table 5. With a significance level of 5%, the variables that uncovered a significant relationship were dividend (D) with a p- value of 0.009 and earnings with a p-value of 0.001. Wisnantiastri Do dividends still matter? The role of investment opportunities … Journal of Accounting and Investment, 2023 | 528 Table 4 Regression Analysis Assumption Test Value Parameter Results Strong IO Weak IO Strong IO Weak IO Normality 0.58 0.117 < 0.05 Normal data Normal data Multicollinearity D E NegE AC Cov (2.627) (2.814) (1.065) (2.123) (1.619) D E Neg E AC Cov (1.125) (1.475) (1.386) (1.085) (1.136) VIF < 10 No collinearity problem No collinearity problem Heteroscedasticity Uniformly distributed Uniformly distributed Scatterplot analysis Homoscedasticity Homoscedasticity Autocorrelation 2.127 1.553 (dU < d < 4 dU) No auto correlation No auto correlation To conclude hypothesis testing, the researcher tested the effect of IO on the ability of dividends to predict future earnings by comparing the dividend coefficient in each model. The comparing results are presented in Table 6. Table 5 The Regression Results Variable Strong IO Weak IO Coefficient Sig Coefficient Sig D 0.224 0.034 0.946 0.009 E 0.250 0.004 0.862 0.001 NegE 7.472E+9 0.991 9.377E+10 0.226 AC -0.252 0.012 -0.130 0.175 Cov 9.025+11 0.114 3.652E+10 0.605 F-test 0.000 0.000 R-Square 0.572 0.483 Based on the comparative coefficient value of the two categories obtained from the coefficient of variable dividend value in each model shown in Table 6, it can be concluded that hypothesis H1 was accepted and H0 was rejected. It is because the dividend coefficient value in companies with weak IO of 0.946 was greater than the coefficient value of companies with strong IO of 0.224. It implies that companies with weak investment opportunities had a stronger effect on predicting future earnings. Table 6 Result of Comparing Dividend Ability Model Companies with strong IO (Model 1) Companies with weak IO (Model 2) Result Dividend coefficient (D) 0.224 0.946 Weak IO > Strong IO H1 => accepted The hypothesis testing result revealed that the ability of dividends to predict future earnings in companies with weak IO was stronger than that of dividends to predict future earnings in companies with strong IO. These results indicate similarities to the study results of Kaplan and Pérez-Cavazos (2022). This similar result illustrates consistency with the previous theory that companies with strong (weak) IO are less (more) willing to pledge Wisnantiastri Do dividends still matter? The role of investment opportunities … Journal of Accounting and Investment, 2023 | 529 a larger fraction of their permanent earnings as dividends, which from an information perspective, decreases (increases) the earnings predictability of dividends. Firms with weak investment opportunities tend to respond to earnings shocks by paying dividends, allowing their dividends to convey future earnings information to the market. The stronger signal of dividends is because they have a stronger valuation incentive to communicate earning information. In that case, it is suspected that dividends will be compensated because companies cannot provide an expected signal from profit, which is called a counter-signal (Aghamolla et al., 2021). Otherwise, companies with strong profitability and growth opportunities refrain from doing so (Kaplan & Pérez-Cavazos, 2022). They choose not to signal earnings with dividends but rely on additional information to differentiate them from other types of companies to avoid noisy information signals and shun their standard signal (Feltovich et al., 2002). Nevertheless, these results differ from classical signaling models of dividends. The previous theory argues that there is a stronger signal in companies with stronger investment opportunities because higher cost allows the signal to convey more information (John & Williams, 1985; Rock & Miller, 1985). From the investor characteristic viewpoint, in Indonesia, dividends are variable in investment decisions making (Christanti & Mahastanti, 2011). Dividends can be used as a signal that conveys information to the market (Rock & Miller, 1985) and contains information about future changes (Ham et al., 2020). Hence, the rank of dividends considerations in Indonesia is 22 of 33 criteria, not higher than fundamental criteria, such as financial statements, earnings, and reputation (Christanti & Mahastanti, 2011). It reinforces that investors also rely on additional information besides dividends. The descriptive analysis also showed that companies with strong IO had a greater average profit than weak IO. This data supports by companies’ profile explanations. From this data, it can be implied that companies with strong IO have the flexibility to distribute dividends and bear the signal cost that arises (Bhattacharya, 1979; DeAngelo & DeAngelo, 2006) but chooses not to use this kind of signal. In contrast, companies with weak IO are suspected of having unfavorable fundamental conditions because of the lower Tobin’s Q value. Companies with weak IO are still likely to distribute dividends. Furthermore, Indonesia’s investor profile is dominated by individual investors (Olavia, 2022), which can also explain this result since individual investors have an investment goal in the form of dividend yields. In addition, Indonesian people invest in stocks for long-term investment (Lidwina, 2021). In Indonesia, dividend policy is an essential factor influencing stock prices (Riyani & Andriana, 2019), so companies with good fundamental conditions prioritize dividend distribution and allocating funds for long-term investments. Corporate earnings divided into a permanent component of earnings and transitory earnings will be beneficial in estimating the company's future value, which will be helpful in the decision-making stock return (Putra, 2008). Dividend changes also convey information about earnings changes (Prasetyanta, 2014). However, there are opposite predictions about the role of dividends Wisnantiastri Do dividends still matter? The role of investment opportunities … Journal of Accounting and Investment, 2023 | 530 in conveying future earnings because dividend policy has no positive effect on future earnings (Anggara, 2020). The finding of this research also explains inconsistent results of the effect of dividends by an alternative explanation from the companies’ characteristic viewpoint: investment opportunities. Conclusion In conclusion, companies with weak IO have a stronger dividend signal than those with strong IO. Weak IO prefers to choose the dividend that conveys information about future earnings to the market, whereas strong IO refrains from doing so. This result contributes an alternative explanation of the inconsistency of previous results. Hence, investors investing in the Indonesia Stock Exchange should notice companies' characteristics, i.e., investment opportunities, while considering dividends as a signal for better investment decision-making. As a result of this, investors must better understand and categorize firms from IO characteristics from investment opportunities proxies by Tobin’s Q score. The limitation of this study is that the research samples were too specific based on aggressiveness in paying dividends, and only one firm characteristic was studied. Different company sectors may have another result and explanation. Therefore, future research is expected to provide a sample from other sectors or different firm characteristics to enrich the alternative explanations. Acknowledgment The researcher wants to express her gratitude to Lembaga Penelitian dan Pengabdian Masyarakat (LPPM) Universitas Terbuka for fundings this research with contract number B/962/UN31.LPPM/PT.01.03/2022. The researcher also hopes this study can contribute to many parties, both academically and practically. References Aghamolla, C., Corona, C., & Zheng, R. (2021). No reliance on guidance: counter-signaling in management forecasts. The RAND Journal of Economics, 52(1), 207–245. https://doi.org/10.1111/1756-2171.12367 Allen, F., & Michaely, R. (2003). Payout policy. Handbook of the Economics of Finance, 1, 337– 429. https://doi.org/10.1016/S1574-0102(03)01011-2 Anggara, I. W. G. W. P. (2020). Pengaruh kebijakan dividen terhadap laba di masa depan. E- Jurnal Akuntansi, 30(9), 2428. https://doi.org/10.24843/eja.2020.v30.i09.p20 Bhattacharya, S. (1979). Imperfect policy, in the hand " fallacy " the bird. The Bell Journal of Economics, 10(1), 259–270. https://doi.org/10.2307/3003330 Christanti, N., & Mahastanti, L. A. (2011). Faktor-faktor yang dipertimbangkan investor dalam melakukan investasi. Jurnal Manajemen Teori Dan Terapan, 4(3), 37–51. https://doi.org/10.20473/jmtt.v4i3.2424 https://doi.org/10.1111/1756-2171.12367 https://doi.org/10.1016/S1574-0102(03)01011-2 https://doi.org/10.24843/eja.2020.v30.i09.p20 https://doi.org/10.2307/3003330 https://doi.org/10.20473/jmtt.v4i3.2424 Wisnantiastri Do dividends still matter? The role of investment opportunities … Journal of Accounting and Investment, 2023 | 531 DeAngelo, H., & DeAngelo, L. (2006). The irrelevance of the MM dividend irrelevance theorem. Journal of Financial Economics, 79(2), 293–315. https://doi.org/10.1016/j.jfineco.2005.03.003 Erickson, T., & Whited, T. M. (2011). Treating measurement error in Tobin’s q. Review of Financial Studies, 25(4), 1286–1329. https://doi.org/10.1093/rfs/hhr120 Farre-Mensa, J., Michaely, R., & Schmalz, M. (2014). Payout policy. Annual Review of Financial Economics, 6, 75–134. https://doi.org/10.1146/annurev-financial-110613-034259 Feltovich, N., Harbaugh, R., & To, T. (2002). Too Cool for School? Signalling and Countersignalling. The RAND Journal of Economics, 33(4), 630-649. https://doi.org/10.2307/3087478 Grullon, G., Michaely, R., Benartzi, S., & Thaler, R. H. (2005). Dividend changes do not signal changes in future profitability. Journal of Business, 78(5), 1659–1682. https://doi.org/10.1086/431438 Gultom, R. J. (2018). Hubungan perubahan dividen dengan perubahan laba perusahaan terdaftar di bei. Studi Akuntansi Dan Keuangan, 1(1), 72-93. https://doi.org/10.21632/saki.1.1.72-93 Ham, C. G., Kaplan, Z. R., & Leary, M. T. (2020). Do dividends convey information about future earnings? Journal of Financial Economics, 136(2), 547–570. https://doi.org/10.1016/j.jfineco.2019.10.006 John, K., & Williams, J. (1985). Dividends, dilution, and taxes: a signalling equilibrium. The Journal of Finance, 40(4), 1053–1070. https://doi.org/10.1111/j.1540- 6261.1985.tb02363.x Kaplan, Z., & Pérez-Cavazos, G. (2022). Investment as the opportunity cost of dividend signaling. The Accounting Review, 97(3), 279–308. https://doi.org/10.2308/tar-2019- 0313 Lidwina, A. (2021, April 12). Mayoritas orang indonesia berinvestasi saham untuk jangka panjang. Katadata.co.id. https://databoks.katadata.co.id/datapublish/2021/04/12/mayoritas- orang-indonesia-berinvestasi-saham-untuk-jangka-panjang Melani, A. (2021, April 11). Trivia saham: yield dividen jadi pertimbangan investasi di pasar modal, apa itu ?. Liputan6. https://www.liputan6.com/saham/read/4529305/trivia- saham-yield-dividen-jadi-pertimbangan-investasi-di-pasar-modal-apa-itu Miller, M. H., & Modigliani, F. (1961). Dividend policy, growth, and the valuation of shares. The Journal of Business, 34(4), 411–433. https://doi.org/10.1086/294442 Mooi, E., & Sarstedt, M. (2019). A concise guide to market research: the process, data, and methods using ibm spss statistic. In the process, data, and methods using ibm spss statistics. Springer. https://doi.org/10.1007/978-3-642-12541-6 Nuriksani, P., & Sari, V. P. (2022). Pengaruh pembayaran dividen terhadap kinerja keuangan perusahaan pada perusahaan manufaktur di BEI. Jurnal Ekonomi, 27(1), 94–108. https://doi.org/10.24912/je.v27i1.860 Olavia, L. (2022, July 11). Masih didominasi milenial dan gen z, jumlah investor saham tembus 4 juta di semester i. Investror.id. https://investor.id/market-and-corporate/299355/masih- didominasi-milenial-dan-gen-z-jumlah-investor-saham-tembus-4-juta-di-semester-i Prasetyanta, A. (2014). Pengaruh perubahan dividen terhadap profitabilitas masa yang akan datang (future profitability). Jurnal Ekonomi Dan Bisnis, 17(2), 129-148. https://doi.org/10.24914/jeb.v17i2.278 PT Kustodian Sentral Efek Indonesia. (2021). Statistik pasar modal indonesia februari 2021. Kustodian Sentral Efek Indonesia, 1–6. https://www.ksei.co.id/files/Statistik_Publik_Januari_2021.pdf https://doi.org/10.1016/j.jfineco.2005.03.003 https://doi.org/10.1093/rfs/hhr120 https://doi.org/10.1146/annurev-financial-110613-034259 https://doi.org/10.2307/3087478 https://doi.org/10.1086/431438 https://doi.org/10.21632/saki.1.1.72-93 https://doi.org/10.1016/j.jfineco.2019.10.006 https://doi.org/10.1111/j.1540-6261.1985.tb02363.x https://doi.org/10.1111/j.1540-6261.1985.tb02363.x https://doi.org/10.2308/tar-2019-0313 https://doi.org/10.2308/tar-2019-0313 https://databoks.katadata.co.id/datapublish/2021/04/12/mayoritas-orang-indonesia-berinvestasi-saham-untuk-jangka-panjang https://databoks.katadata.co.id/datapublish/2021/04/12/mayoritas-orang-indonesia-berinvestasi-saham-untuk-jangka-panjang https://www.liputan6.com/saham/read/4529305/trivia-saham-yield-dividen-jadi-pertimbangan-investasi-di-pasar-modal-apa-itu https://www.liputan6.com/saham/read/4529305/trivia-saham-yield-dividen-jadi-pertimbangan-investasi-di-pasar-modal-apa-itu https://doi.org/10.1086/294442 https://doi.org/10.1007/978-3-642-12541-6 https://doi.org/10.24912/je.v27i1.860 https://investor.id/market-and-corporate/299355/masih-didominasi-milenial-dan-gen-z-jumlah-investor-saham-tembus-4-juta-di-semester-i https://investor.id/market-and-corporate/299355/masih-didominasi-milenial-dan-gen-z-jumlah-investor-saham-tembus-4-juta-di-semester-i https://doi.org/10.24914/jeb.v17i2.278 https://www.ksei.co.id/files/Statistik_Publik_Januari_2021.pdf Wisnantiastri Do dividends still matter? The role of investment opportunities … Journal of Accounting and Investment, 2023 | 532 Putra, W. M. (2008). Pengukuran permanent earnings pada hubungan dengan kebijakan dividend perusahaan: review penelitian. Jurnal Akuntansi Dan Investasi, 9(2), 189–196. https://journal.umy.ac.id/index.php/ai/article/view/1107/1185 Rahmawati, W. T. (2021, May 28). Surya Toto tetap membagikan dividen dengan yield 2,38% meski rugi, ini jadwalnya. Kontan.co.id. https://investasi.kontan.co.id/news/surya-toto-tetap- membagikan-dividen-dengan-yield-238-meski-rugi-ini-jadwalnya Riyani, Y., & Andriana, S. (2019). Pengaruh Pengumuman Kebijakan Dividen terhadap Volatilitas Harga Saham. Jurnal Riset Terapan Akuntansi, 3(1), 59–67. https://doi.org/10.5281/zenodo.3837944 Rock, K., & Miller, M. H. (1985). Dividend policy under asymmetric information. The Journal of Finance, 40(4), 1031–1051. https://doi.org/10.1111/j.1540-6261.1985.tb02362.x Said, A. A. (2021). Meski untung segunung, faktanya google, facebook dan amazon tidak pernah membagikan dividen. Liputan6. https://www.liputan6.com/bisnis/read/4539134/meski- untung-segunung-faktanya-google-facebook-dan-amazon-tidak-pernah-bagikan- dividen Spence, M. (1973). Job market signalling. The Quarterly Journal of Economics, 87(3), 355–374. https://doi.org/10.1055/s-2004-820924 Sumani, S., Sandroto, C. W., & Mula, I. (2017). Perilaku investor di pasar modal indonesia. EKUITAS (Jurnal Ekonomi Dan Keuangan), 17(2), 211. https://doi.org/10.24034/j25485024.y2013.v17.i2.2250 Tobin, J. (1969). A equilibrium approach to monetary theory. Journal of Money, Credit and Banking, 1(1), 15–29. https://doi.org/10.2307/1991374 Ulfah, F. U. (2020, April 23). Meski rugi, indika (INDY) tetap bagikan dividen tunai US$ 30juta. Bisnis.com. https://market.bisnis.com/read/20200423/192/1231511/meski-rugi- indika-indy-tetap-bagikan-dividen-tunaius30juta Wareza, M. (2019, May 22). Rugi Rp 6,2 T, perusahaan sandi tetap bagi dividen Rp 298 M. CNBC Indonesia. https://www.cnbcindonesia.com/market/20190522130943-17- 74283/rugi-rp-62-t-perusahaan-sandi-tetap-bagi-dividen-rp-298-m Widsatrya, I. W. K., & Subroto, B. (2014). Pengaruh perubahan pengumuman dividen terhadap harga saham. Jurnal Ilmiah Mahasiswa FEB Universitas Brawijaya, 3(2), 1–23. https://jimfeb.ub.ac.id/index.php/jimfeb/article/view/2299 Wijayanti, R. I. (2022, May 7). Ini alasan suatu emiten tidak membagikan dividen, setuju atau tidak ?. IDXChannel.com. https://www.idxchannel.com/market-news/ini-alasan- suatu-emiten-tidak-membagikan-dividen-setuju-atau-tidak/2 https://journal.umy.ac.id/index.php/ai/article/view/1107/1185 https://investasi.kontan.co.id/news/surya-toto-tetap-membagikan-dividen-dengan-yield-238-meski-rugi-ini-jadwalnya https://investasi.kontan.co.id/news/surya-toto-tetap-membagikan-dividen-dengan-yield-238-meski-rugi-ini-jadwalnya https://doi.org/10.5281/zenodo.3837944 https://doi.org/10.1111/j.1540-6261.1985.tb02362.x https://www.liputan6.com/bisnis/read/4539134/meski-untung-segunung-faktanya-google-facebook-dan-amazon-tidak-pernah-bagikan-dividen https://www.liputan6.com/bisnis/read/4539134/meski-untung-segunung-faktanya-google-facebook-dan-amazon-tidak-pernah-bagikan-dividen https://www.liputan6.com/bisnis/read/4539134/meski-untung-segunung-faktanya-google-facebook-dan-amazon-tidak-pernah-bagikan-dividen https://doi.org/10.1055/s-2004-820924 https://doi.org/10.24034/j25485024.y2013.v17.i2.2250 https://doi.org/10.2307/1991374 https://market.bisnis.com/read/20200423/192/1231511/meski-rugi-indika-indy-tetap-bagikan-dividen-tunaius30juta https://market.bisnis.com/read/20200423/192/1231511/meski-rugi-indika-indy-tetap-bagikan-dividen-tunaius30juta https://www.cnbcindonesia.com/market/20190522130943-17-74283/rugi-rp-62-t-perusahaan-sandi-tetap-bagi-dividen-rp-298-m https://www.cnbcindonesia.com/market/20190522130943-17-74283/rugi-rp-62-t-perusahaan-sandi-tetap-bagi-dividen-rp-298-m https://jimfeb.ub.ac.id/index.php/jimfeb/article/view/2299 https://www.idxchannel.com/market-news/ini-alasan-suatu-emiten-tidak-membagikan-dividen-setuju-atau-tidak/2 https://www.idxchannel.com/market-news/ini-alasan-suatu-emiten-tidak-membagikan-dividen-setuju-atau-tidak/2