*3.1. Empirical Model*

This study examines if family firms are financially healthier than their counterpart firms. This led us to two specific factors that determine the financial health of the firm as shown in Sections 3.1.1 and 3.1.2 below.

## 3.1.1. Leverage

"We use a panel regression analysis to evaluate whether a family owned company has a different level of leverage to a similar non-family" (Ntoung et al. 2017) to examine the level of a firm's leverage (debt/EBITDA ratio and interest coverage ratio). One reason for excluding the usual ratio of debt/capital is because it is influenced not only by the choice of debt level the company makes, but also by the level of equity the company has. Also, the debt/capital ratio induced the market perception of the firm into the equation, thus it wouldn't provide aclean estimate of the leverage choices made by the firm.

We run five regression equations regarding our two dependent variables for leverage.

The regression equation is illustrated as follows:

$$\begin{aligned} \text{Leverage} &= \alpha\_1 \times (\text{Families dummy}) + \alpha\_2 \times \text{Age} + \alpha\_3 \times \text{Size measure} + \alpha\_4 \times \text{Industry Damage} \\ &+ \alpha\_5 \times \text{Profitability measure} + \alpha\_6 \times \text{Interactive Variable} \end{aligned} \tag{1}$$

where,

Leverage: Debt/EBITDA and Interest Coverage Ratio.

Family firm takes: dummy equals 1 when a firm is a family firm or zero otherwise

Profitability measure: refers to return of equity, return on assets, EBITDA margin, netincome margin Size measure: refers to number of employees, total revenues, total assets

Age: calculate based on the company date of establishment.

Industry dummy: equaling 1 as dummy for each IAC classification code,

Year dummy: equals 1 for each year considered in the analysis.

Furthermore, to correct the presence of heteroskedasticity and serial correlation in the data, we employ the Huber–White sandwich estimator for variance (Ntoung et al. 2017).

3.1.2. Risk Exposure

"To critically analyze the risk profile of family businesses as opposed to their non-family peers, it is vital to examine beyond leverage, factors that reflect overall risk. We further evaluate these factors by employing the Altman Z-score, a predictive model developed to determine acompany's probability of filing for bankruptcy in the next subsequence of years and to measure the overall financial health of a company" (Ntoung et al. 2016b).

Altman Z-score: We consider the Altman Z-score as a dependent variable. "The choice of variable regarding a company risk's of survival was based on four balance sheet and income statement variables, namely profitability, leverage, solvency, liquidity, and activity. The result of the combination of ratios gives rise to a discriminantscore, otherwise called the Z-Score. The ratios are X<sup>1</sup> = working capital/total assets, X<sup>2</sup> = retained earnings/total assets, X<sup>3</sup> = earnings before interest and taxes/total assets, X<sup>4</sup> = market value of equity/book value of total debt, and X<sup>5</sup> = sales/total assets. In 1998, Altman redefined

his model by excluding X<sup>5</sup> (sales/total assets) to forecast the corporate risk of survival for manufacturing firms in Mexico. The weighted coefficients thus have different values" (Ntoung et al. 2017).

$$
\Delta Z'' = 6.5\chi\_1 + 3.2\chi\_2 + 6.72\chi\_3 + 1.0\chi\_4 \tag{2}
$$

Source: Altman et al. (1998, p. 3).

The analysis of family business risk characteristics using Altman Z-score provides information about risk exposure the company is willing to take on, hypothesizing that family firms would have a tendency to be more risk averse, if all things being equal.

We run five regression equations regarding our two dependent variables for leverage. The regression equation is illustrated as follows:

Altman Z-Score = α<sup>1</sup> × (Family dummy) + α<sup>2</sup> × Age + α<sup>3</sup> × Size measure + α<sup>4</sup> × Industry Dummy <sup>+</sup> <sup>α</sup><sup>5</sup> <sup>×</sup> Profitability measure <sup>+</sup> <sup>α</sup><sup>6</sup> <sup>×</sup> Interactive Variables (3)

where,

Altman Z-Score: 6.5X1 + 3.2X2 + 6.72X3 + 1.0X4

Family firm takes: dummy equals 1 when a firm is a family firm or zero otherwise Profitability measure: refers to return of equity, return on assets, EBITDA margin, netincome margin Size measure: refers to number of employees, total revenues, total assets Age: calculated based on the company date of establishment. Industry dummy: equaling 1 as dummy for each IAC classification code,

Year dummy: equals 1 for each year considered in the analysis.

Furthermore, to correct the presence of heteroskedasticity and serial correlation in the data, we employ the Huber–White Sandwich estimator for variance (Ntoung et al. 2017).
