*3.6. Profitability*

The next important characteristic of the company—profitability—is an important determinant of capital structure. According to the pecking order theory, a firm first instances its activity from retained earnings. If this source is not sufficient, a company tries to obtain external sources of capital (Myers and Majluf 1984). From this perspective, higher financial leverage does not always imply or correlate with higher profitability.

Another implication may be derived from the trade-off theory, which assumes a state of balance between equity and debt capital, while the cost of debt capital is lower. More profitable companies usually have sufficient financial resources necessary to pursue their investment plans. However, more profitable companies may find a tax shield to be a decisive argument for increasing financial leverage (Bouallegui 2006), which is especially important for companies from countries where the tax rate is high. The theory of free cash flows also posits that more profitable companies should indebt themselves because it provides a self-control mechanism. It forces management to transfer free cash flows as dividends to their shareholders instead of investing in less profitable investment projects (Izdihar 2019).

Highly profitable companies have much easier access to external financing at a much lower cost (Cassar and Holmes 2003). This is also supported by the substitutive theory, which posits that less risky and more profitable companies are much more able to finance their activity from external sources, especially debt. High profitability also minimizes the risk of bankruptcy, and for this reason, the capacity of indebtedness is increased (Ramli et al. 2019). Highly profitable companies, which finance their activity from internal sources, are not required to disclose detailed information on their operations (Li and Islam 2019). Internal sources of finance (retained earnings) and increased indebtedness may be attractive for investors since a firm's shareholding is not diluted (Karacaer et al. 2016). On the basis of the above discussion, it may seem that the impact of profitability on capital structure is ambiguous (Degryse et al. 2012). However, from the perspective of NTBFs, we can suppose that more profitable companies would have much better credit standing and better access to debt. Therefore, we treat the firm's profitability parameter as a control variable.

#### *3.7. Growth Opportunities*

Growth opportunities are an important firm characteristic influencing capital structure in the high-tech sector. Most often, high-tech companies tend to use their own equity funds because of innate higher risk and the necessity of more costly supervision of this type of company (Myers 1977). High growth opportunities, on the one hand, create the chance of development, but on the other hand, pave the way for new risk. Usually, enormous growth opportunities accompany low equity values which are necessary to finance important investment projects. Fortunately, these companies, even when dealing with severe financial problems, don't have problems with raising equity capital. Indebtedness may put pressure and discipline on the management and enforce a more efficient decision-making process. The valuation process of high-tech companies is based on their future potential (option), which is heavily burdened with risk. Therefore, the market valuation is under the threat of impairment. This is especially important considering that the asset is in substantial part intangible and, as a result, cannot serve as collateral (Karacaer et al. 2016). Thus, some researchers (Rajan and Zingales 1995) hypothesize an inverse relationship between growth opportunities and financial leverage. This relationship is also implied by the pecking order theory, which posits that a firm tends to finance its activity by internal funds and, afterward, look for external ones. Agency costs theory provides similar implications for high-tech companies. Additional monitoring costs related to management supervision may be substantial, especially when growth opportunities do exist, which supposedly will lead to an increased cost of debt. High-tech companies will be discouraged from taking on more debt in their balance sheet in order to minimize potential conflict between shareholders and creditors (Ramli et al. 2019). The implication of the substitution theory also confirms

that relationship, because high-tech companies are more prone to the risk of financial situation deterioration. Therefore, we treat the firm's growth opportunities as a control variable.

## **4. Sample Characteristics, Research Design, and Results**

The study sample consisted of 31 companies listed on the Warsaw Stock Exchange classified as high-tech firms in sectors like biotechnology, R&D in physics, natural sciences, engineering, biology, medical laboratories, computer software, e-commerce, marketing analysis, etc. We decided to use data derived from firms listed on the stock exchange because of a higher quality of accounting data. These companies, under the scrutiny of stock market institutions and the public, are obliged to meet higher standards of transparency and are audited. Companies may also be classified as NTBFs because the oldest firm in the study period is 17 years old, and the average age is around six years. The initial sample consists of 155 firm-year observations covering the period of 2014–2018. The final sample is limited to only 102 firm-year observations due to the missing data.

Our main object of interest is capital structure, and as a dependent variable, we use the leverage ratio calculated as total liabilities to total assets. As a proxy for the innovation generated internally, we use a ratio of the sum of R&D expenses recorded in the P&L statement and year-to-year change in R&D outlays recorded in the balance sheet, deflated by the total assets. In our opinion, this is the only possible way to measure R&D outlays based on information derived from a financial statement. As the proxy for the innovation acquired externally, we use a year-to-year change of intangibles extracted from the balance sheet, excluding R&D expenses recognized. We also use a set of control variables such as profitability (ROE) and growth opportunities. In order to avoid the influence of outliers, all data were winsorized. Table 1 presents the characteristics of the main variables used in the model.


**Table 1.** Sample statistics.

Source: our own elaboration based on the data from financial statements.

In order to avoid intercorrelated variables in the model, we performed a correlation analysis, the results of which are presented in Table 2. The highest correlations, however moderate, are between a firm's age and profitability (ROE), financial leverage, and size. The results are logical and correspond to the conclusions of the literature review section. The older a company is, the higher its profitability. Similarly, the older the firm is, the more able it is to indebt itself. Finally, bigger companies tend to be more profitable. The results show that variables INNOV\_INT and INNOV\_EXT are weakly correlated. The rest of the correlation coefficients of independent variables are at a low or moderate level, so including them in the model is not controversial.


**Table 2.** Correlation between variables.

Source: our own elaboration based on the data from financial statements.

To test the hypotheses formulated in the previous section, we used the following model:

LEVi,t = INTANGIBILITYi,t + INNOV\_INTi,t + INNOV\_EXTi,t + CUR\_RATIOi,t + SIZEi,t + AGEi,t + ROEi,t + SALES\_TRi, (1)

where:

LEVi,t—financial leverage (total liabilities/total assets) of the i-company in t-year

INTANGIBILITYi,t—the ratio of intangibles to total assets of the i-company in t-year

INNOV\_INT1i,t—the ratio of internally generated intangibles to total assets of the i-company in t-year INNOV\_EXT2i,t—the ratio of externally acquired intangibles to total assets of the i-company in t-year CUR\_RATIOi,t—liquidity of the company measured as a current ratio (current asset/current liabilities) SIZEi,t—the size of the i-company in t-year as a logarithm of total assets

AGEi,t—age of the i-company in t-year

ROEi,t—profitability of the i-company in t-year measured as a return on equity

SALES\_TRi,t—sales trend of the i-company in t-year calculated as year-to-year change of sales (sales from the t-year minus sales from the t−1 year, and deflated by the sales from t−1 year)

We ran a regression with a robust option in order to obtain robust coefficients. It allows us to avoid many problems with the specification of the model.

We performed an extensive post-estimation diagnosis to test our model. We tested the model for multicollinearity using the variance inflation factor and detected none. We ran a Shapiro-Wilk test for residuals, and we couldn't reject the null hypothesis which states that they are normally distributed. Finally, we used the Ramsey RESET to test for the specification of the model; results (0.048) are in the borderline and may suggest that there are some problems with the specification of the model. The model is better at detecting influence on the dependent variable and should not be treated as a predictive model. The model detects some critical links between variables and has acceptable predicting power (adj. R = 0.54). First of all, we found a strong influence of the firm's age on financial leverage, which suggests that the older the firm is, the more leveraged it is. The results fit the theory and results of other studies. The second important conclusion is that the more liquid the company is, the less leveraged it is. The implication of that result may be that younger companies that are usually less leveraged tend to maintain a safe cash position and hold more cash within the company. Bigger companies may allow themselves to keep a relatively lower level of liquidity because they are able to raise cash faster if needed through the bank system. Therefore, we provide empirical evidence supporting our fourth and sixth hypotheses.

From our perspective, the most crucial results refer to the variables INNOV\_INT and INNOVE\_EXT. The p-value of those variables is at a low (10%), yet still statistically significant (see Table 3). Firstly, INNOV\_INT has a negative coefficient, which suggests that the more a company invests in an innovative in-house project, the less willing a bank sector is to finance it with debt. This provides empirical evidence supporting our second hypothesis and may be explained by the higher informational asymmetry generated by the R&D project, which probably translates to a higher cost of debt. Secondly, INNOV\_EXT has a positive coefficient, which implies that the bank sector is willing to provide more external funds to companies acquiring innovation externally. We ascribe that result to the fact that

external acquisition of technology/invention is perceived to be less risky and the final output more predictable. Again, we provide an argument supporting the third hypothesis. The results must be interpreted with caution, and the hypotheses need to be tested on high-tech companies from other emerging markets.



\* significance at 10% level; \*\* significance at 5% level; \*\*\* significance at 1% level. Source: our own elaboration based on the data from financial statements.

Unfortunately, we find no empirical evidence supporting the first and the fifth hypothesis. With regard to the firm's size, this may be explained by the fact that the majority of companies are of moderate size. In the case of the intangibility parameter, we suppose that this parameter would be more important for companies in sectors other than high-tech. In our opinion, this matter needs further investigation.
