The Effect of Liquidity, Leverage, and Profitability on Financial Distress with Audit Committee as a Moderating Variable

This study examines the effect of liquidity, leverage, and profitability on financial distress with the audit committee as a moderating variable. This study used secondary data from the annual reports of manufacturing companies listed on the Indonesia Stock Exchange (IDX) from 2016 to 2019. The research sample was selected using purposive sampling, and 33 companies were obtained as the study samples. The data were then analyzed using the logistic linear regression method with SPSS ver 26 software. The study results found that liquidity and profitability had a negative effect on financial distress, whereas leverage had a positive effect on financial distress. In addition, the study also found that the audit committee enhanced the effect of liquidity and profitability on financial distress. In contrast, the audit committee reduced the effect of leverage on financial distress.

only support the results of some previous empirical evidence that did not explain the causal effect relationship. (3) In Liang et al.
(2020) study, they used machine learning methods through a model prediction approach and provided empirical evidence of two primary indicators (FRs and CGIs). However, the prediction results did not reveal the best indicators (FRs or CGIs); rather, an explanation of the combination of the two indicators. In the authors' best knowledge, it is inconsistent with the purpose in the preliminary explanation of the study.

LITERATURE STUDY 2.1 Signaling theory
Signaling theory suggests that a company provides signals to users of financial statements. The theory is derived from pragmatic accounting theory, which emphasizes the influence of information on changes in information user behavior (Apriada and Suardikha, 2016). According to Morris (1987), signal theory shows how information asymmetry in the market can be minimized by signaling more information to other parties. Information asymmetry in the capital market occurs because a company (management) has more information than outsiders (investors). Brigham and Houston (2018) assert that a signal is an action performed by a company's management that provides instructions for investors about how management in its capacity views and takes action when it sees prospects for the company. Companies with profitable prospects will strive to avoid selling shares, even attempting to defend them and seek new capital in other ways such as using debt (Nurmayasari et al, 2021).
The management is required to be transparent in presenting the company's financial statements. Financial statements are prepared based on the activities of the company at a certain time. From financial statements, one is able to recognize whether the company is in a healthy condition or experiencing financial distress (Baskoro et al, 2021). A company in healthy condition is indicated with the profit gain in a relatively long period of time. In addition, it can also be observed from the value of the company's cash flow. High cash flow over a long period indicates the company is able to pay debts to creditors (Agustini and Wirawati, 2019).

Financial distress
Financial distress is a condition in which a company is encountering financial difficulties (Rainingtyas et al, 2021). According to Nabawi and Efendi (2020), financial distress occurs because a company cannot manage and maintain its financial performance stability, which stems from the failure to promote the products it manufactures. Thus, such a situation causes a decrease in sales and income from at least sales. It leads the company to experience operating losses and net losses for the current year. Furthermore, the losses will result in a capital deficiency due to a decrease in the value of retained earnings used to make dividend payments. Thus, such a situation leads to a deficiency in total equity. If the condition persists, the corporate's total liabilities may exceed the total assets it owns. As such, a company is associated with the experience of financial distress, which eventually, if it is unable to overcome the above conditions, it will lead to bankruptcy (Nabawi and Efendi, 2020).

Liquidity
Liquidity is an indicator of a company's ability to meet obligations as soon as it falls due. The company with high liquidity indicates its high ability to meet maturing obligations. The liquidity ratio acts to determine the extent to which the company is able to pay its short-term obligations that are due soon (Kadim et al., 2020). In this study, a company's liquidity is assumed to be a predictor of the condition of a corporate financial distress and is measured by the current ratio (CR).

Leverage
Based on some of the previous empirical evidence results, such as the research conducted by Pawitri and Alteza (2020), it revealed that leverage positively affected financial distress. However, these aforementioned results are not in line with those of Yuliani and Sulpadli (2020), which stated that leverage had a negative effect on financial distress. However, the results of these two studies contradicted the results of those conducted by Stephanie et al. (2020) and Maulidia and Asyik (2020), which found that leverage had no effect on financial distress conditions.

Profitability
Based on the empirical evidence from previous study results, such as Nabawi and Efendi (2020), profitability had a negative effect on financial distress. Nevertheless, these study results are not in line with those of Hosea et al. (2020), which found that profitability had a positive effect on financial distress. However, the results of these two studies contradict the research of Asmarani

Size of the audit committee
The existence of an audit committee enhances the effectiveness of the oversight function of the company's board of directors and commissioners (Damayanti et al., 2017). Large size of the audit committee is expected to have increased resources and the improved quality of supervision, especially in the company's liquidity (Ardiyansyah and Paramita, 2020). Therefore, an audit committee with large members will have more resources to monitor the corporate financial performance in its liquidity; thus, it can reduce the potential for financial difficulties.

RESEARCH METHOD 3.1 Research type and description of the research population (object)
In this study, the population is public companies in the manufacturing and construction sectors listed on the Indonesia Stock Exchange (IDX) from 2016 -2019. Manufacturing industry companies in the construction sector were chosen since they constituted the most considerable number of companies listed. The study was focused on determining the effect of liquidity, leverage, and profitability on financial distress with the size of the audit committee as a moderating variable in manufacturing companies listed on the Indonesia Stock Exchange.
Furthermore, the observation period was carried out for four years, starting from 2016 to 2019 since it could provide an overview of the current state of the corporate's financial changes. Meanwhile, the sample selection from the manufacturing industry population used purposive sampling techniques, such as: 1) basic and chemical industries, 2) miscellaneous manufacturing, and 3) consumer goods industries. After grouping the samples, the companies engaged in basic and chemical industries, followed by goods and consumption industries, were dominated by companies operating in miscellaneous industries.

Techniques of data collection
This research used secondary data, which were obtained indirectly through intermediaries (recorded by other parties). The data consisted of annual reports and audited financial statements for December 3.
Manufacturing and construction companies never publishing restatement for consecutive years from 2016 to 2019 (1) 4. The number of sample companies 33 5.
Total research data (33 x 4 years) 132 (Source: data collection) Based on Table 1 above, the number of data used in this study accounts for 33 companies with a total observation for 4 years, totaling 132 sample companies.

Variables and measurements
Variables measurement is a process of values or attributes assignment to an object or research instrument used to be processed and produces an absolute value, from which, conclusions can be drawn. In this study, the authors used a measurement scale with the type of ratio for the five variables (liquidity, leverage, profitability, and the size of the audit committee) and a nominal of measurement scale for the financial distress variable. The following measurements are used to measure the five variables used in this study: a) Liquidity In this study, the liquidity variable is measured using the proxy of the current ratio, i.e., total current assets divided by total current liabilities of a company (Ariawan, 2017). The formulation of the current ratio is as follows (Ariawan, 2017):

b) Leverage
In this study, leverage is measured using the proxy of debt to total assets (DTA), which is total liabilities divided by a company's equity. DTA ratio calculated using the following formula (Putri and Erinos, 2020) (Putri dan Erinos, 2020):

c) Profitability
In this study, the return on assets (ROA) ratio is a proxy for measuring profitability variables, i.e., the amount of net profit earned divided by each rupiah of funds converted in total assets. In other words, the ROA ratio is measured based on the total net income divided by the total assets owned by the company. The ROA ratio is determined using the equation as follows (Liang et

e) Financial distress
In this study, the variable of financial distress is measured using the Altman Z-Score method. A company is assigned value (1) if it has a Z score > 2.90, indicating not in a bankruptcy state. Meanwhile, the value (0) is assigned if the company has a Z value < 1.20, indicating that the company is in a bankruptcy state. The formulation of financial distress using a proxy from the Altman Z-Score method is as follows (Rohiman dan Damayanti, 2019): : Keterangan: Z : Total score of Altman Z-score; WCTA : Ratio of working capital to total assets; RETA : Ratio of retained earnings to total assets; EBITTA : Ratio of earnings before interest and tax to total assets; MVEBVD : Ratio of the market value of equity to book value of total debt; STA : Ratio of sales to total assets.

Data analysis techniques 3.4.1 Logistics regression test
Logistic regression is used if a dependent variable in the causal relationship being tested is a dummy variable whose nominal scale variables with two or more categories are represented by the numbers 0 (zero), 1 (one), and 2 (two). The logistic regression data analysis technique primarily aims to see the level of prediction of category with the more dominant type (Ghozali, 2016).

Overall fit model
In this study, the likelihood value of L model is the probability that the hypothesized model describes the input data. A decrease in likelihood (-2LogL) indicates a better regression model, or in other words, the hypothesized model fits the data (Ghozali, 2016). Therefore, the hypothesis testing in this study was carried out using an equation based on a moderation regression analysis model. According to (Ghozali, 2016), the moderation regression analysis model uses an analytical approach that maintains sample integrity and provides a basis for controlling the influence of moderating variables. To use this method, it is necessary to compare three regression equations to determine the type of moderating variable. Three equations are formulated in this study as follows: : Interaction between the size of audit committee and liquidity; UKA_LEV : Interaction between the size of audit committee and leverage; UKA_PROF : Interaction between the size of audit committee and profitability; ε : Value of error term

Goodness of fit test of the regression model (Hosmer and Lemeshow's GoF test)
The goodness of fit test of regression model was carried out using Hosmer and Lemeshow's goodness of fit test, which is a hypothesis test used to determine the goodness or fitness between predictions in the logistic regression model compared to data from observations. Therefore, the model can be regarded to be fit (Ghozali, 2016).
Hosmer and Lemeshow's goodness of fit test is used to measure whether the predicted probabilities match the observed probabilities. The basis for decision-making is based on the statistical value of Hosmer and Lemeshow's goodness of fit test of 0.05. In this sense, H0 is rejected because there is a significant difference between the model and the observation value so that the goodness of fit model is not reliable and the model cannot predict the observed value. Suppose the statistical value of Hosmer and Lemeshow's goodness of fit test is 0.05, it means that H0 is accepted and the model is able to predict the observed value (Ghozali, 2016).

Coefficient of determination (Nagelkereke R Square)
Cox and Snell's R square is based on the likelihood estimation technique with a maximum value of less than one (1); thus, it is difficult to interpret. This model attempts to imitate the size of R square in multiple regression. Nagelkereke R square is used to obtain a coefficient of determination that can be interpreted, such as the R 2 value in the multiple regression model (Ghozali, 2016).
Cox and Snell's R square is adjusted to Nagelkereke R square to ensure that the value varies from 0 to 1. It can be achieved by dividing the value of Cox and Snell's R square with its maximum value. A value close to 1 means that the independent variables provide almost all the information required to predict the dependent variables. Meanwhile, a smaller value (away from 1) means that the independent variables have limited ability to describe the dependent variable (Ghozali, 2016).

RESULTS AND DISCUSSIONS 4.1 Descriptive statistics
Based on descriptive statistical test results, we obtained data from the multiplication of the number of company samples. It accounts for 33 companies with a total observation period of 4 years, resulting in 132 research data. Descriptive statistics are presented to explain the data description of all variables taken into account in the research. Table 4 presents the descriptive statistics of the variables used in this study. Based on Table 2 above, we can observe that the number of objects under examination (N) from 2016 to 2019 is 132 data. According to Table 4, we could also observe each variable's minimum, maximum, mean, and standard deviation values. The table is used to assist in the determination of the deviation for each affecting variable to one another.

Overall model fit test results
The overall model fit test was evaluated by comparing the value between -2 log-likelihood (2LL) at the initial (block number = 0) with a value of -2 log-likelihood (-2LL) at the final (block number = 1), and the results were at -2 log-likelihood (-2LL). In Table 3, the test results indicate that the initial -2LL value is 251.959. After processing three independent variables, the final value of -2LL decreased to 34.601. The decrease in the value of -2 log-likelihood can be interpreted that incorporating independent variables into the model can improve the fit model and show a better regression model. In other words, the hypothesized model fits the data.

Equation model of logistics regression test results
The logistic regression model can be formulated from the estimated parameter values in the variables in the equation  The Exp (B) value is also called odds ratio (OR) for each independent variable, and its interactions are as follows: (1) The variable of liquidity (LIK) with the proxy of current ratio (CR), provides more predictions to financial distress by 0.644 times. The coefficient value is -0.440 from the natural logarithm of 0.644. Therefore, the coefficient value is negative, and we can conclude that the current ratio has a negative effect on financial distress. (2) The variable of leverage (LEV) with the debt to total assets (DTA) as a proxy provides more predictions to financial distress by 0.237 times. The value of 1.441 results from the natural logarithm of 0.237. Therefore, given the positive coefficient value, we conclude that debt to total assets positively affects financial distress. (3) The variable of profitability (PROF) with the return on assets (ROA) as a proxy cannot predict financial distress.
The coefficient has a negative value, i.e., -26.313, from which we can conclude that return on assets has a negative effect on financial distress. (4) The interaction of liquidity with the size of audit committee also predicts financial distress by 0.001 times. The coefficient value is 5.274 from the result of the natural logarithm of 0.001. Therefore, the coefficient value is positive, and it can be concluded that the interaction of liquidity with the size of the audit committee has a positive effect on financial distress. (5) The interaction of leverage with the size of the audit committee also provides a financial distress prediction by 0.022 times. Meanwhile, the coefficient value is -1.672, resulting from the natural logarithm of 0.022. Given the negative coefficient value, we can conclude that the interaction between leverage and the size of the audit committee has a negative effect on financial distress. (6) The interaction of profitability with the size of the audit committee cannot predict financial distress. The coefficient value of -14,449 from the positive coefficient can be concluded that the interaction of profitability with the size of the audit committee negatively affects financial distress.

Wald test results (hypotheses)
As shown in Table 4, the Wald test results present 6 (six) hypothesis testings in this study. Therefore, it can be concluded that the results of six hypotheses testing with reference to the one-sided test (one-tiled) can be seen in Table 5 as follows:

The effect of liquidity on financial distress
These study results indicate that liquidity has a negative effect on financial distress. Based on the study results, we can explain that liquidity is used as an indicator that a company can meet its immediately maturing obligations. A company with high liquidity is highly able to meet its maturing obligations. The liquidity ratio has a function to determine the extent to which the company can pay its short-term obligations that will be due soon (Kadim et al., 2020). In this study, the company's liquidity is assumed to be a predictor of its financial distress as measured by the current ratio (CR).

The effect of leverage on financial distress
The study results indicate that leverage has a positive effect on financial distress. Based on the results, we can explain that financial leverage is used to measure the company's ability to meet both short and long-term obligations. Analysis of this ratio is required to measure a company's ability to pay debts (short-term and long-term) if, at some point, it is liquidated or dissolved. A company with high financial leverage indicates it has much debt to outsiders, high financial risk due to financial distress.

The effect of profitability on financial distress
The study results reveal that profitability has a negative effect on financial distress. This study indicates that a company with high profitability has a high profit. Thus, it is less likely to experience financial distress. In this study, a company's profitability is assumed to predict a company's financial distress as measured by the return on assets ratio.

The effect of interaction between the size of the audit committee and liquidity on financial distress
The study results found that the size of the audit committee reduced the effect of liquidity on financial distress. The interaction between the size of the audit committee and liquidity indicates that when the company can overcome its short-term obligations, it also signals to investors its sound financial status and is away from potential financial distress. In addition, the company's condition that has an excellent current ratio is considered a healthy company. On the other hand, a lower current ratio is relatively regarded as more prone to financial distress.