This study explores the impact of firm productivity on the probability of default for the Indian high-defaulting industries for the period 2015–2024.
The study uses the Cobb–Douglas productivity function (CD) to capture two measures of productivity: assets and employees. The default risk has been computed using Altman’s (2000) Z-score and Byström’s (2006) distance-to-default (DD) metrics. The relationship between productivity and default risk for the dataset comprising 251 companies has been examined using panel data regression analysis.
The results indicate that asset productivity (ASS.PROD) positively drives firms’ financial health and contributes to mitigating their default risk, whereas employee productivity (E.PROD) does not impact firms’ likelihood of default.
The managers and investors from an emerging market can benefit from the study’s findings. Managers must focus on improving firms’ asset productivity to mitigate the rising default risk of Indian high-defaulting industries. Investors should rely on productivity measures to assess default risk before evaluating and arriving at investment decisions.
This is among the first few studies in India to apply both CD production functions and dual default risk measures across multiple high default industries to investigate the relationship between productivity and default risk. This study, therefore, fills the research gap from an Indian perspective.
1. Introduction
Productivity serves as a key driver of a firm’s growth and performance. The productivity of firm signifies the effectiveness with which an organization converts its resources into outputs, serving as an indicator of its operational capability. Industries with greater production levels have less probability of facing defaults because they are better at withstanding economic and financial downturns (Wu et al., 2024). On the contrary, less productive firms have a higher risk of default owing to their poor financial performance. Hence, productivity affects the financial well-being of the firms (Parnes, 2023).
The earlier estimation of default risk relied heavily on accounting-based metrics (Altman et al., 2015; Dalwai and Salehi, 2021). However, this assessment lacked the sensitivity in determining operational inefficiencies that are caused by inefficient utilization of inputs employed. Productivity, on contrary, which is measured using the interaction between input and output factors provides the indication of firms’ operational excellence and the probable financial vulnerability. Only a few research studies have highlighted that a decline in productivity reflected in reduced output levels can serve as a precursor to default risk (Bryan et al., 2013).
However, this relationship is not investigated in the Indian context where the majority of research is dominated by conventional accounting indicators, namely leverage, solvency and liquidity ratios for corporate default prediction (Bonelli, 2023; Khurana and Sharma, 2024; Shetty and Vincent, 2024). While productivity is widely regarded as a vital measure of operational efficiency, its potential contribution to lowering default risk remains under-examined in Indian academic research. This limited empirical investigation into the role of productivity improvements in reducing default probability highlights a significant gap in the existing literature. In several Indian industries, firms operate with large labour forces [1] and substantial fixed assets, rendering employee costs and asset utilization important drivers of financial health. When lower employee costs stem from productivity gains, they help lower operating expenses and stabilize operating margins (Parnes, 2023). Simultaneously, in asset-intensive sectors, efficient deployment of capital assets enables generating adequate operating cash flows to meet high fixed financing commitments (Becchetti and Sierra, 2003). These improvements in employee and asset productivity directly contribute to strengthening firms’ financial health. Hence, India’s comparatively low labour costs and asset-intensive industries, such as metals and mining, construction and real estate, provide a unique context to study how firms’ productivity can influence their probability of default. This study investigates the impact of productivity on default risk for the period 2015–2024 by using data of 251 non-financial firms listed on the national stock exchange (NSE). The study has applied Cobb–Douglas (CD) production function to measure productivity by using two approaches, namely assets productivity (Wu et al., 2024; Konrad and Mangel, 2000) and employee productivity (Khatun and Afroze, 2016; Parnes, 2023; Wu et al., 2024) by using different combination of inputs and outputs. The CD function has been employed because it allows researchers to derive a complete indication of the company’s operational efficiency, growth prospects and financial susceptibility (Hossain et al., 2013; Rani et al., 2023).
The default risk is measured using Altman’s Z-score (2000) and Byström’s (2006) distance-to-default (DD) model. The hypotheses were formulated and tested using panel data regression models. The results of regression analysis indicated that asset productivity (ASS.PROD) is positively related with both the default risk measures. It implies that higher productivity leads to a higher value of financial health scores such as Z-score and DD metrics which tends to reduce firms’ default risk. The study found no impact of employee productivity (E.PROD) in predicting corporate failure.
The key motivation behind this study arises from the information about the rise in default risk in certain industries, namely textiles, food products, metals and mining, construction and real estate development. The Credit Rating Information Services of India Limited (CRISIL) reports that these industries have consistently been found to be in a distressed state over the last 36 years (crisil-ratings-annual-default-and-ratings-transition-study-fy-2024.pdf). Although their contribution to Gross Domestic Product (GDP) has been documented in economic reports, increased default probabilities raise concerns about their long-run viabilities. This study has targeted this population set to investigate if higher productivity induces their financial solvency and can contribute in reducing their default risk.
Secondly, evaluations of default risk in India continue to be anchored in financial ratios and credit ratings, while firm-level productivity is rarely factored in as an indicator of resilience. Such reliance often channels credit toward less productive firms (George et al., 2022). It thereby elevates the default risk and overlooks opportunities to strengthen more efficient and competitive enterprises. Hence, another motivation arises from the need to examine whether productivity can serve as an indicator of default risk, thereby facilitating the channelling of credit toward more productive and financially resilient firms.
Moreover, there is scant literature that utilizes the productive capacities of employees and assets in examining their long-term impact on financial solvency of Indian firms. A few studies that exist fall short in several respects, including not utilizing the CD function (Datta, 2013; Kanoujiya and Rastogi, 2024), using considerably older research timeframe (Datta, 2013), and a different sampling population (George et al., 2022). On the contrary, this study applies CD production function to examine productivity effects and incorporates two complementary measures of default risk, Altman’s Z-score and the Byström distance-to-default model for the recent ten-year period (2015–2024). It further adopts a sampling framework focused on high-defaulting industries, enabling a more detailed and context-specific analysis of the productivity and default risk relationship.
This study makes the following contributions to the existing research. First, productivity metrics derived from the input and output combinations are frequently overlooked in accounting research for credit-risk assessment. The current research advances the development of default prediction models by embedding indicators of operational efficiency, thereby enhancing the robustness and accuracy of credit risk assessment. Second, it yields valuable insights for policymakers and financial institutions seeking to foster financial stability through productivity-oriented policy interventions and optimized resource allocation. This research adds value by highlighting the predictive role of asset productivity in assessing default risk. The findings equip banks with insights to enhance credit allocation by directing resources more effectively toward firms that exhibit stronger asset productivity. Third, to best of our knowledge, existing literature in Indian context has not examined impact of productivity on financial health and stability using current methodology. This study uniquely utilizes the CD functional model to compute productivity measured by asset and employee productivity. Fourth, the study has specifically targeted industries with high default risk and investigates whether firms that demonstrate higher productivity can forestall their default risk. There is a lack of literature in Indian scenario that exclusively targets current sampling set and timeframe for estimating this relationship.
The rest of the study is organized as follows. Section 2 presents a review of the literature in three sub-sections. This is followed by Section 3 explaining the research design and description of variables. Section 4 discusses the findings of the study. The implications and conclusion of the study are provided in Section 5.
2. Literature review and hypotheses development
This section gives description about past literature in three sub-sections. Section 2.1 provides a literature review on papers related to default risk. The literature on productivity is reviewed in Section 2.2. The association between productivity and financial stability is summarized in Section 2.3.
2.1 Default risk
The existing literature focuses on predicting a company’s distress before it enters the stage of insolvency (Beaver, 1966). One of the most popular and reliable models for predicting corporate defaults is Z-score developed by Altman in 1968. It utilized a multivariate discriminate analysis technique to distinguish a distressed firm from a non-distressed firm (Altman, 1968). However, this original Z-score, based on U.S. manufacturing firms, lost its accuracy over time and evolving financial conditions. Altman (2000) revised this model by updating coefficients and accounting for differences across sectors and firm types. The model was equipped with more recent data to enhance its relevance and accuracy across varied financial contexts. The application of the Altman Z-score (2000) model has been widely supported by researchers in today’s economy (Bryan et al., 2013; Dalwai and Salehi, 2021; Khurana and Sharma, 2024).
Bryan et al. (2013) computed the default risk of USA non-financial firms to investigate the relationship between productivity and financial distress. Khurana and Sharma (2024) applied this model to capture the default risk of Indian high default industries, whereas Dalwai and Salehi (2021) used this model to measure financial distress of Oman’s firms. Festa et al. (2021) computed default risk of Indian top pharmaceutical firms through Z-score (2000) model. The results highlighted the soundness of this model to capture financial health of Indian-based firms. This study applies this revised Z-score for carrying out the primary analysis of this study.
Although Altman Z-score has long served as a core model for default risk prediction through accounting-based ratios, advances in financial modelling have brought in market-based approaches to offer further analytical perspectives. In response to the need for market-based evaluation, default prediction studies have emphasized contingent claim theory to predict the financial failure of firms. Merton’s (1974) distance-to-default model which is based on contingent claim theory is the fundamental model applied in accounting and finance research for evaluating credit risk of corporations. The DD metric utilizes firms’ market value of equity and debt information to forecast corporate failure. Byström (2006) argued that Merton’s DD model premise of constant leverage proportions lacks empirical evidence in the literature. Nevertheless, the modified DD metric which underlines the importance of debt ratio and stock volatility can be computed for any company, independent of its capital structure, or capital volatility, thereby favouring more realistic assumptions. Byström (2006) DD metric, which is easy to compute, also produces results comparable to the conventional Merton (1974) model (Agrawal and Maheshwari, 2016; Byström, 2006). Based on previous studies (Chaturvedi and Singh, 2024; Kabir et al., 2020), the modified DD metric has been employed to measure default risk as part of the robustness of the results.
2.2 Productivity
Productivity is widely regarded as a key measure of a firm’s operational performance (Sharma and Sharma, 2024a, b). Research highlights productivity as a proactive indicator, capable of revealing early signs of financial health decline before they are reflected in standard financial ratios. Within this context, asset productivity and employee productivity stand out as vital elements. Asset productivity evaluates how effectively a company employs its capital resources to generate income, whereas employee productivity captures the contribution of human resources to output. When examined together, these metrics provide a crucial understanding of firms’ operational efficiency.
Numerous studies have highlighted significance of asset productivity in assessing a firm’s overall performance (Rani et al., 2023; Sharma and Sharma, 2024a, b). Sharma and Sharma (2024) indicated that firms demonstrating higher asset turnover rates consistently achieve stronger operational efficiency and financial outcomes. Similarly, Amoa-Gyarteng (2021) found that the extreme sales declines are usually associated with poor asset usage, which can be a precursor of profitability and productivity concerns.
The way asset productivity enhances profitability of firms, there is evidence of employee productivity contributing to financial profitability and resiliency of firms (Konrad and Mangel, 2000; Sharma and Sharma, 2024a, b). Mok (2002) found that higher employee productivity increases profitability which in turn reduces default risk. Using evidence from Greek firms, Parnes (2023) showed that employee productivity enhances operational efficiency that contributes to firms’ financial stability. Similar results are supported by Dabla-Norris et al. (2012), who found that both employee and asset productivity significantly influence firms’ overall growth and development.
The previous literature has widely applied CD function to measure asset and employee productivity, by expressing output as a function of two core inputs: asset and employees (Hossain et al., 2013; Rani et al., 2023). Hossain et al. (2013) examined efficiency of various production functions and found that CD function was the best fit to measure the productivity in developing economies. Using data of Indian manufacturing firms, Rani et al. (2023) demonstrated the superiority of CD production function in outperforming deep learning models. Based on the above studies, this study employs employee and asset-based indicators to measure productivity using CD production function.
2.3 Productivity and default risk
Resource-Based theory posits that firm’s operational efficiency is governed by how well its input resources are utilized to produce desired outputs (Gharfalkar et al., 2018). The improvement in outputs is essential for sustaining long-term competitive advantage and industrial resilience. Krugman (1990) argues that productivity may not be everything, but it is the most crucial of everything in the long run. During the economic downturn in the US in the 1970s, productivity reduction was extensively attributed to the financial crash (Lichtenberg, 1992). As a result, in the 1980s, corporations prioritized productivity and cost management as the primary tactics, which contributed to the revival of the US economy at the time.
The earlier research suggests that productivity improves through effective asset utilization that upholds revenues and competitiveness, while poor utilization often denotes profitability challenges and greater financial vulnerability (Amoa-Gyarteng, 2021; Gambhir and Sharma, 2015). By managing asset productivity effectively, firms can generate sufficient returns to cover replacement, obsolescence, market, and other uncertainty costs. This, in turn, reinforces financial resilience while mitigating the risk of default. Further, higher employee productivity plays a vital role in sustaining growth and enhancing a firm’s creditworthiness. As highlighted by Parnes (2023), firms with low employee productivity are unable to support increasing growth rates, making it difficult to secure external financing. In contrast, higher productivity enables firms to grow consistently, improve profitability, and strengthen the confidence of lenders and investors. It thereby lowers the perceptions of financial distress among capital contributors. In the Indian context, employee productivity fosters cost competitiveness due to relatively lower labour costs [1]. Lower labour costs across Indian industries facilitate better firm performance by improving cost efficiency and sustaining healthier operating margins. Similarly, asset productivity is expected to enhance operational efficiency because firms operating in capital-intensive and high-default industries are closely monitored by lenders and governed by regulatory mechanisms such as the insolvency and bankruptcy code, which restrict inefficient use of capital. Improved asset productivity enables firms to extract higher operating cash flows from their capital base, increase capacity utilization, and reduce idle investments. These effects strengthen financial resilience and lower the probability of default, implying that productivity gains in the Indian context enhance financial stability. Accordingly, it is expected that both employee and asset productivity significantly contribute to lowering default risk.
Based on these arguments, this study formulates the following hypotheses for this study. The conceptual framework for the same is provided in Figure 1 (see Figure 1 for conceptual framework).
The conceptual path diagram shows relationships among productivity measures and default risk using both solid and dashed arrows. On the left, a rectangular box labeled “Productivity (Independent Variable)” has two solid arrows extending to the right. One solid arrow points upward right to a box labeled “Employee Productivity (I V 1) single asterisk”. Another solid arrow points downward right to a box labeled “Asset Productivity (I V 2) double asterisk”. Above the employee productivity box, a dashed vertical arrow points upward to a box labeled “Inputs: Number of employees; Capital employed. Output: Sales”. From the employee productivity box, a solid diagonal arrow labeled “H 1” points upward right to a box labeled “Default Risk (D V) triple asterisk”. Below the asset productivity box, a dashed vertical arrow points downward to a box labeled “Inputs: Total Assets. Output: Sales”. From the asset productivity box, a solid diagonal arrow labeled “H 2” points downward right to another box labeled “Default Risk (D V)”. Between the two default risk boxes, a central box labeled “Controlled By: Leverage; Size; Liquidity; Loss” appears. A dashed vertical arrow points upward from this control box to the upper “Default Risk (D V) triple asterisk” box, and another dashed vertical arrow points downward from the control box to the lower “Default Risk (D V)” box.Conceptual framework. Note: Author’s own. (IV 1)* denotes employee productivity as first independent variable, (IV 2)** denotes asset productivity as second independent variable and (DV)*** denotes default risk as dependent variable
The conceptual path diagram shows relationships among productivity measures and default risk using both solid and dashed arrows. On the left, a rectangular box labeled “Productivity (Independent Variable)” has two solid arrows extending to the right. One solid arrow points upward right to a box labeled “Employee Productivity (I V 1) single asterisk”. Another solid arrow points downward right to a box labeled “Asset Productivity (I V 2) double asterisk”. Above the employee productivity box, a dashed vertical arrow points upward to a box labeled “Inputs: Number of employees; Capital employed. Output: Sales”. From the employee productivity box, a solid diagonal arrow labeled “H 1” points upward right to a box labeled “Default Risk (D V) triple asterisk”. Below the asset productivity box, a dashed vertical arrow points downward to a box labeled “Inputs: Total Assets. Output: Sales”. From the asset productivity box, a solid diagonal arrow labeled “H 2” points downward right to another box labeled “Default Risk (D V)”. Between the two default risk boxes, a central box labeled “Controlled By: Leverage; Size; Liquidity; Loss” appears. A dashed vertical arrow points upward from this control box to the upper “Default Risk (D V) triple asterisk” box, and another dashed vertical arrow points downward from the control box to the lower “Default Risk (D V)” box.Conceptual framework. Note: Author’s own. (IV 1)* denotes employee productivity as first independent variable, (IV 2)** denotes asset productivity as second independent variable and (DV)*** denotes default risk as dependent variable
Firms with higher employee productivity exhibit a lower probability of default.
Firms with higher asset productivity exhibit a lower probability of default.
3. Research design
3.1 Sampling selection and data source
The study has relied upon secondary database for collecting data of all variables under the study. Out of all non-financial companies listed on NSE, six industries have experienced highest default rate since last 36 years: textiles, food products, metals and mining, construction and real estate development (crisil-ratings-annual-default-and-ratings-transition-study-fy-2024.pdf). The final sample of 251 companies was finalized after excluding financial services firms owing to difference in their accounting methods, and companies with missing information. The detail of deriving sampling companies is provided in Table 1. Collecting data from Centre for Monitoring Indian Economy (CMIE’s) Prowess database, 2,510 panel observations for the period 2015–2024 were analysed through panel data regression analysis. The classification of industries as per National Industrial Classification (NIC) codes is provided in Table 2 (see Tables 1 and 2 for sampling procedure and NIC codes, respectively).
Procedure of deriving sampling firms under the study
| Criteria | No. of firms |
|---|---|
| Total number of firms listed on NSE as on 31st March 2024 | 1,298 |
| Less: financial companies | (291) |
| Less: non-financial companies which excludes the major defaulting industries list as per CRISIL rating report | (698) |
| Initial number of total non-financial firms in six high defaulting industries as per Prowess database | 309 |
| Less: companies with unavailable data | (58) |
| Final sample of six high defaulting industries | 251 |
Industry classification of firms in the sampling frame
| Two-digit NIC code | Industry’s main product/Service group | No. of firms |
|---|---|---|
| 01 | Dairy products | 4 |
| 10 | Vegetables oils and products, tea, sugar, coffee and other food products | 41 |
| 11 | Alcoholic beverages | 13 |
| 13 | Diversified cotton textile, cotton and blended yarn | 34 |
| 14 | Readymade garments | 8 |
| 20 | Man-made filaments and fibres | 18 |
| 24 | Iron and steel and other metals | 54 |
| 25 | Fabricated metal products | 9 |
| 41 | Real estate construction | 27 |
| 42 | Infrastructural construction | 43 |
| Total firms | 251 |
3.2 Measuring variables under the study
3.2.1 Measures of dependent variable- default risk
The default risk is measured through Altman Z-score (2000). The model classifies a corporation into defaulting or non-defaulting firm based on the Z-scores. The company scoring a Z-score above 2.99 is financially stable, below 1.81 is financially distressed and between 1.81 and 2.99 has a moderate chance of default. The following equation is used for estimating default risk.
where
W.Cap = Net Working Capital
T.ASS = Total Assets
R.E = Retained Earnings
PBIT = Profit before Interest and Tax
SAL = Sales
To ensure the robustness of the results, the Byström (2006) distance-to-default metric was applied to capture default risk. The market-based DD model is simple to compute and suitable for the characteristics of emerging economy (Chaturvedi and Singh, 2024; Kabir et al., 2020). The equation of modified DD metric is as follows:
where
DDBYS = Bystrom’s Distance-to-Default Metric
Leverage = Leverage ratio is calculated as TD/(VE + TD), VE denotes the market value of equity capital and TD is enumerated as summation of long- and short-term liabilities.
e = volatility in daily equity returns
3.2.2 Measures of independent variables – productivity
Financial ratios are frequently overused in default prediction (Ciampi et al., 2021; Laitinen and Muñoz-Izquierdo, 2023). However, firms with similar ratios can exhibit different distress levels due to underlying operational inefficiencies (Gambhir and Sharma, 2015). The CD function measures how well a company employs its inputs to produce output, whose lower values foretell the probability of default much before the decline is observed in the financial ratios. Based on previous studies (Amoa-Gyarteng, 2021; Khatun and Afroze, 2016; Mok, 2002; Dabla-Norris et al., 2012; Wu et al., 2024), the current research measures a firm’s productivity using CD function in two aspects: employee and asset productivity. Both types of productivity are analysed independently because they represent firms’ different resource efficiencies. Asset productivity indicates how efficiently assets generate income, whereas employee productivity reflects the efficiency of human capital. By analysing them separately, firms can identify the origin of default risk, from the employee or asset side.
The following CD production function is applied to measure employee productivity.
where
Y = Output (Sales)
E = Number of Employees
F = Capital Employed
A = Efficiency Parameter
α = Coefficient of number of Employees
β = Coefficient of Capital Employed
The above model is converted into the linear equation by taking the log on both sides. The log transformation is described as follows:
The CD production function for asset productivity is as follows. The details for measurement of variables are provided in Table 3 (see Table 3 for measurement of variables).
Variables used under panel data regression analysis
| Dependent variable | Proxies | Independent variables | Proxies | Control variables | Proxies |
|---|---|---|---|---|---|
| Default risk | Z-score (2000) | Employee productivity | INPUTS – no. of employees, capital employed | Leverage | Ratio of total debt to total assets |
| Distance-to-default (2006) | OUTPUT – sales | Size | Natural logarithm of market capitalization | ||
| Asset productivity | INPUTS – total assets | Liquidity | Ratio of cash and cash equivalents to total assets | ||
| Loss | Dummy variable, which takes the value 1 when a firm incurs loss and 0 otherwise | ||||
| OUTPUT – sales | Volatility | Measured as standard deviation of daily equity returns and Ln.AGE | |||
| Ln.AGE | Computed as natural logarithm of firm’s age |
where
Y = Output (Sales)
F = Total Assets
A = Efficiency Parameter
α = Coefficient of Total Assets
Log Y = Log A + αLogT + Ui
3.2.3 Measures of control variables
The study controls for the following factors that can impact default risk. Leverage (total liabilities/total debt), size (natural logarithm of the market capitalization), liquidity (cash and cash equivalents/total assets) and loss, a dummy variable, that takes the value of 0 when a firm defaults and 1 for the non-default. It means that how loss incurred by the company (provided code = 0) will affect the default risk of the regression analysis. Besides these variables that can impact default risk measured using Altman’s Z-score (2000), return volatility (standard deviation of daily equity returns) and age (natural logarithm of firm’s age) are also included as control variables for impacting default risk under the distance-to-default metric (Chaturvedi and Singh, 2024; Kabir et al., 2020).
4. Research findings
4.1 Summary statistics
The study has performed summary statistics to look at the impact of firms’ productivity characteristics on firms’ default risk. The results of descriptive statistics are provided in Table 4. To deal with outliers, all continuous variables were winsorized at 1% from both tails. The average Z-score (2000) value of 2.725 suggests that firms lie in the grey zone indicating a moderate probability of default. The DD metric also shows that the mean value of firms’ financial stability is low standing at 0.543. The mean value of employee productivity (3.326) suggests that on average firms sustain their financial resilience through employee productivity.
Summary statistics
| Variables | Mean | Std. dev. | Minimum | Median | Maximum | Skewness | Kurtosis |
|---|---|---|---|---|---|---|---|
| Z-score (2000) | 2.725 | 3.775 | −1.840 | 1.641 | 22.692 | 1.140 | 2.672 |
| DDBY | 0.543 | 0.371 | 0.175 | 0.416 | 2.113 | 0.773 | 2.077 |
| E.PROD | 3.326 | 2.561 | 0.140 | 2.794 | 14.263 | 0.590 | 1.980 |
| ASS.PROD | 2.39 | 1.616 | 0.084 | 2.235 | 8.888 | 0.305 | 1.785 |
| Leverage | 0.618 | 0.291 | 0.094 | 0.614 | 1.868 | 0.045 | 1.781 |
| Size | 6.278 | 2.155 | 1.677 | 6.158 | 11.655 | −1.191 | 1.021 |
| Liquidity | 0.022 | 0.048 | −0.053 | 0.008 | 0.309 | 1.367 | 3.037 |
| Loss | 0.248 | 0.432 | 0.000 | 0 | 1 | 0.154 | 0.858 |
| Volatility | 3.19 | 0.89 | 1.343 | 3.173 | 5.979 | −1.221 | 1.128 |
| Ln.AGE | 3.534 | 0.498 | 2.197 | 3.466 | 4.86 | −1.284 | 1.228 |
The standard deviation values for Z-score and employee productivity are captured at 3.775 and 2.561, respectively. It indicates that sectoral differences exist among sampling firms. However, skewness values of all variables appear to be in the accepted ranges (−1.96 to +1.96). Similarly, kurtosis findings also show that values of all variables range between −3 and +3, suggesting that sampling firms are normally distributed. The average of ASS.PROD appears to be 2.39. It shows that firms are productive in terms of asset utilization and converting them into sales. The mean value of liquidity (0.022) which captures firms’ cash holdings to total assets suggests that on average firms suffer from liquidity concerns. The average leverage (0.618) demonstrates that firms are less reliant on external sources of financing. The mean volatility and Ln.AGE appears to be (3.19) and (3.53), respectively (see Table 4 for summary statistics).
4.2 Correlations
Table 5 tabulates the correlation coefficients between variables under the study. There is a significant and positive association (correlation coefficient = 0.24, correlation coefficient = 0.118) between employee productivity and default risk measures: Z-score (2000) and distance-to-default (DDBY), respectively. It indicates that higher employee productivity leads to better financial stability. Similarly, asset productivity has shown a positive association (correlation coefficient = 0.376; correlation coefficient = 0.194) with Z-score (2000) and distance-to-default, respectively, which again suggests that greater asset productivity mitigates the likelihood of default risk.
Pairwise correlations
| Variables | (1) | (2) | (3) | (4) | (5) | (6) | (7) | (8) | (9) | (10) |
|---|---|---|---|---|---|---|---|---|---|---|
| (1) Z-score (2000) | 1.000 | |||||||||
| (2) DDBY | 0.808 | 1.000 | ||||||||
| (0.000) | ||||||||||
| (3) E.PROD | 0.240 | 0.118 | 1.000 | |||||||
| (0.000) | (0.000) | |||||||||
| (4) ASS.PROD | 0.376 | 0.194 | 0.844 | 1.000 | ||||||
| (0.000) | (0.000) | (0.000) | ||||||||
| (5) Leverage | −0.509 | −0.521 | 0.090 | −0.031 | 1.000 | |||||
| (0.000) | (0.000) | (0.000) | (0.123) | |||||||
| (6) Size | 0.429 | 0.593 | 0.138 | 0.188 | −0.386 | 1.000 | ||||
| (0.000) | (0.000) | (0.000) | (0.000) | (0.000) | ||||||
| (7) Liquidity | 0.203 | 0.230 | 0.011 | −0.041 | −0.195 | 0.118 | 1.000 | |||
| (0.000) | (0.000) | (0.594) | (0.038) | (0.000) | (0.000) | |||||
| (8) Loss | −0.329 | −0.315 | −0.178 | −0.255 | 0.469 | −0.378 | −0.139 | 1.000 | ||
| (0.000) | (0.000) | (0.000) | (0.000) | (0.000) | (0.000) | (0.000) | ||||
| (9) Volatility | −0.368 | −0.713 | −0.116 | −0.125 | 0.414 | −0.567 | −0.130 | 0.303 | 1.000 | |
| (0.000) | (0.000) | (0.000) | (0.000) | (0.000) | (0.000) | (0.000) | (0.000) | |||
| (10) Ln.AGE | −0.021 | 0.042 | −0.078 | −0.083 | −0.024 | 0.181 | 0.019 | −0.001 | −0.103 | 1.000 |
| (0.282) | (0.033) | (0.000) | (0.000) | (0.223) | (0.000) | (0.349) | (0.951) | (0.000) |
The control variables have also shown expected associations. The leverage (correlation coefficient = −0.509) and loss (correlation coefficient = −0.329) are negatively correlated with Z-score (2000). The positive correlation coefficient of size (correlation coefficient = 0.429) with Z-score (2000) implies that large-sized firms are less prone to default risk. The liquidity coefficient (correlation coefficient = 0.203) is also positively correlated with Z-score (2000). It indicates that firms with higher liquid assets show more financial resiliency and lower default risk. The volatility (correlation coefficient = −0.368) appears to be negatively correlated with Z-score (2000), which suggests that firms with more variable returns tend to have a higher likelihood of default (see Table 5 for correlation results).
4.3 Panel data regression analysis
The relationship between firms’ productivity and default risk has been investigated using panel data regression analysis. The Breusch-Pagan Lagrangian multiplier test was applied to choose between pooled ordinary least squares (OLS) model and random-effect model. The p-value less than 0.05 indicated that random-effect model is appropriate fit for the regression model. The next step was to choose between random-effect and fixed-effect model. For this, the study applied Durbin–Wu–Hausman capital (DWH) test. The p-value of this test was observed at less than 0.05, which indicated that fixed-effect model is the best fit for the results. The results of the Hausman test are provided in Table 6. The following equation investigates the relationship between firm productivity and default risk.
where
Z-Scoreit is the default risk measure for a company i in year t computed using Altman’s Z-score (2000). E.PRODit and ASS.PRODit are the productivity information obtained from the CD functions. Sizeit, Liquidityit and Lossit represent control variables for the regression analysis. The auto-correlation and heteroscedasticity concerns were identified by Durbin–Watson and White test, respectively. Therefore, clustering by company method was applied to generate standard errors as the remedial measures for these concerns.
4.4 Relationship between firms’ productivity and default risk
The study examines the impact of firms’ productivity on their financial health using panel data regression analysis. The results estimated using Eq. (1) are reported in Table 7. Intuitively and appearing from results, the coefficient of asset productivity captured by ASS.PROD (coefficient = 0.616; t-statistics = 6.40) is positive and significant at one percent level of significance. The positive relationship between Z-score (2000) and ASS.PROD indicates that a greater value of asset productivity increases firms’ financial solvency and resultantly lower their probability of default. These results demonstrate that managers are making strategic decisions that are enhancing the firms’ asset productivity levels, strengthening the resilience of companies’ in allocating the production factors against probability of defaults.
Relationship between default risk and productivity
| Z-Score(2000)it = β0 + β1 × E.PRODit + β2 × ASS.PRODit + β3 × Leverageit + β4 × Sizeit + β5 × Liquidityit + β6 × Lossit + εit | ||||
|---|---|---|---|---|
| Dependent variable | Altman’s Z-score (2000) | |||
| Variable | Coefficients | t-Statistics | p-value | Significance |
| E.PROD | −0.051 | −1.08 | 0.282 | |
| ASS.PROD | 0.616 | 6.40 | 0.000 | *** |
| Leverage | −2.974 | −6.01 | 0.000 | *** |
| Size | 0.732 | 7.55 | 0.000 | *** |
| Liquidity | 2.398 | 1.60 | 0.111 | |
| Loss | −0.006 | −0.04 | 0.966 | |
| Constant | −1.389 | −2.00 | 0.046 | ** |
| Mean dependent variance | 2.725 | SD dependent variance | 3.775 | |
| R-squared | 0.286 | Number of obs. | 2,510 | |
| F-test | 51.216 | Prob > F | 0.000 | |
| Akaike crit. (AIC) | 9269.397 | Bayesian crit. (BIC) | 9304.365 | |
It lends support to the findings of Becchetti and Sierra (2003) and Bryan et al. (2013) who demonstrated that productivity of a firm has an essential role to play in reducing firms’ default risk. This study’s results further suggest that firms that efficiently handle their tangible and intangible assets can enhance operational efficiency, leading to higher firm performance. If managers fail to enforce efficiency in asset utilization, enterprises become vulnerable to lower productivity. It will pave the way for corporations to fall into deep financial difficulties and increase their level of financial distress. These results also confirm the findings of Amoa-Gyarteng (2021) and Wu et al. (2024).
The regression results in Table 7 show that the coefficient of asset productivity is captured at 0.616. It suggests that a rise in one per cent of asset productivity will cause a 0.616% rise in financial solvency. In other words, firms excelling at asset productivity have lower probability of facing defaults. It lends support to the second hypothesis that productivity accounts for a significant aspect in improving firms’ financial health.
The E.PROD indicates the employees’ productivity contribution to the firm’s financial health. It is expected to be positively linked with financial solvency based on the following notion: greater the employees’ productivity, lower is the likelihood of default. However, the regression analysis shows an insignificant impact of E.PROD on firm’s default risk. It can be argued that employees when forced to exceed production levels may experience stress and increased attrition. Overtime, it raises recruiting expenses, diminishing business performance and negatively impacting firms’ financial health. These findings are supported by the results of Zhenjing et al. (2022).
With regards to control variables, the regression coefficients appear to be in the expected direction. The leverage coefficient (coefficient = −2.974; t-statistics = −6.01) is negative and significant. The negative relationship between leverage and Z-score (2000) indicates that higher leverage proportions on balance sheets can lead companies to the stage of default. This relationship matches the results of Bryan et al. (2013) and Dalwai and Salehi (2020), who indicated the negative effects of debts on firms’ health and stability. Similarly, loss has shown (coefficient = −0.006; t-statistics = −0.04) a negative relationship with default risk. However, the relationship appears to be insignificant for this regression model. The value of size coefficient (coefficient = 0.818; t-statistics = 8.70) is positive and significant. It suggests that big-sized companies have a lower likelihood of facing defaults which support the similar findings of Bryan et al. (2013) and Dalwai and Salehi (2021) (see Table 7 for regression results).
4.5 Robustness tests
The study has carried out robustness checks to validate the sensitivity of primary results. Byström’s (2006) distance-to-default, a widely used market-based model metric, has been used to compute default risk. The model is independent of a firm’s capital structure and asset volatility and is also suitable for measuring default risk under the characteristics of an emerging economy. Hence, an additional empirical analysis was performed to investigate the impact of productivity on default risk. The findings of robustness evaluation are tabulated in Table 8. In line with our expectations, ASS.PROD has shown (coefficient = 0.015; t-statistics = 2.29) a positive and significant relationship under this alternate regression results. It demonstrates that financial health is positively impacted by asset productivity. The adjusted R2 has also notably improved from 0.286 to 0.593 indicating that 59% of variations in default risk are now accounted for independent predictors.
Relation between default risk and productivity
| DD(BY)it = β0 + β1 × E.PRODit + β2 × ASS.PRODit + β3 × Leverageit + β4 × Sizeit + β5 × Liquidityit + β6 × Lossit + β7 × Volatilityit + β8 × Ln.AGEit + εit | ||||
|---|---|---|---|---|
| Dependent variable | Distance-to-default (2006) | |||
| Variable | Coefficients | t-Statistics | p-Value | Significance |
| E.PROD | −0.008 | −2.23 | 0.026 | ** |
| ASS.PROD | 0.015 | 2.29 | 0.022 | ** |
| Leverage | −0.112 | −3.06 | 0.002 | *** |
| Size | 0.060 | 13.33 | 0.000 | *** |
| Liquidity | 0.251 | 1.75 | 0.079 | * |
| Loss | 0.019 | 2.01 | 0.045 | ** |
| Volatility | −0.161 | −18.95 | 0.000 | *** |
| Ln.AGE | −0.025 | −1.11 | 0.266 | |
| Constant | 0.818 | 8.23 | 0.000 | *** |
| Mean dependent variance | 0.543 | SD dependent variance | 0.371 | |
| Overall r-squared | 0.593 | Number of obs. | 2,510 | |
| Chi-square | 456.652 | Prob > χ2 | 0.000 | |
| R-squared within | 0.490 | R-squared between | 0.620 | |
The E.PROD which was insignificant under the previous regression model has now shown a significant impact on default risk. However, the association appears to be negatively impacting Z-score (2000). It can be argued that under financial distress, companies frequently incorporate rigorous performance supervision and stricter accountability norms. This can help employees in contributing to firms’ output formation, while becoming highly productive in achieving goals and improving productivity. However, the extensive financial distress is not mitigated by the corresponding gain in employee productivity. These results are supported by the study of García Martín and Herrero (2025). The results, thus, fail to support the relationship between employee productivity and default risk.
The control variables have also shown significant and expected relationships. The coefficient of leverage (coefficient = −0.112; t-statistics = −3.06) is negative and significant. The size coefficient (coefficient = 0.015; t-statistics = 2.29) has also shown a positive and significant relationship with DD metric. The negative coefficient of volatility (coefficient = −0.161; t-statistics = −18.95) suggests that firms with higher volatile returns have a greater probability of facing default. Liquidity and loss which were insignificant under primary results have now shown a significant impact on default risk. The study, therefore, fails to validate the impact of these variables on financial health. However, the significance and coefficients’ direction of other variables validate the robustness of the primary results (see Table 8 for robustness checks regression results).
4.6 Diagnostics check results
System generalized method of moments (SGMM) can generate consistent results by addressing a number of regression problems such as heteroscedasticity, autocorrelation and endogeneity. SGMM works by utilizing the company’s historical data to correct for econometric problems and maintains that regression findings stay intact. SGMM is specifically developed for a small number of years and a large number of firms, making it a suitable regression for the diagnostic investigation. The dynamic nature of the relationship between variables is proven when lag of dependent variable is used as a regressor in the regression analysis. To investigate the same, we regressed two-year lag values of default risk. The results tabulated in Table 9 capture that the first year lagged value of default risk is significant, indicating that SGMM is a suitable fit to investigate the endogeneity issues.
Relation between default risk and productivity
| Z-Score(2000)it = β0 + β1 × Z-Score(2000)it-1 + β2 × E.PRODit + β3 × ASS.PRODit + β4 × Leverageit + β5 × Sizeit + β6 × Liquidityit + β7 × Lossit + εit | ||||
|---|---|---|---|---|
| Dependent variable | Altman’s Z-score (2000) | |||
| Variable | Coefficients | t-Statistics | p-Value | Significance |
| L1 | 0.414 | 8.04 | 0.000 | *** |
| L2 | −0.053 | −0.73 | 0.466 | |
| E.PROD | 0.068 | 0.86 | 0.388 | |
| ASS.PROD | 0.616 | 4.07 | 0.000 | *** |
| Leverage | −6.632 | −5.58 | 0.000 | *** |
| Size | 0.673 | 4.20 | 0.000 | *** |
| Liquidity | 2.619 | 1.21 | 0.228 | |
| Loss | −0.292 | −1.58 | 0.115 | |
| Volatility | 0.021 | 0.02 | 0.987 | |
| Ln.AGE | 0.414 | 8.04 | 0.000 | *** |
| Constant | −0.053 | −0.73 | 0.466 | |
| Mean dependent variance 2.841 | ||||
The findings provided in Table 9 support the previously observed relationship between productivity and default risk. Intuitively and as captured by these findings, firms with higher levels of asset productivity (coefficient = 0.616; t-statistics = 4.07) have a lower probability of default. The coefficient of E.PROD which was insignificant under earlier findings appears to deliver a similar relationship under dynamic regression results. The results, thus, support the second hypothesis that firms that utilize their assets productively can reduce their likelihood of default. The control variables also capture similar relationships as reported in the main analysis. Leverage is negatively correlated (coefficient = −6.632; t-statistics = −5.58) with Z-score (2000), whereas size (coefficient = 0.673; t-statistics = 4.20) appears to show positive impact on financial stability. The liquidity and loss which were insignificant under primary results also exhibited the same relationship under this regression. Overall, SGMM results ensure no endogeneity concerns and demonstrate that primary regression findings remain intact (see Table 9 for SGMM regression results).
5. Policy implications
The results of this study, which show a positive relationship between asset productivity and financial stability, have important implications for regulators, policymakers, investors and creditors. The regulatory authorities while executing the role of a risk assessor should view low asset productivity as a sign of underlying operational inefficiency and integrate it for closer regulatory supervision. The CD production function employed in the present study demonstrates robustness in capturing default probability, reflected through lower levels of output. Authorities can more accurately identify underperforming firms that appear financially stable through traditional financial ratios but are inefficient in their use of capital resources when assessed through CD productivity scores. Therefore, findings suggest that regulators could utilize the CD framework to evaluate productivity’s impact on default risk in India’s financially distressed sectors. Further, the positive association between asset productivity and firms’ financial stability underscores the importance of incorporating productivity indicators into the early warning frameworks employed by financial regulators for evaluating default risk.
Given that government schemes such as the Production Linked Incentive [2] and Technology Upgradation Fund Scheme [3] already reward firms that demonstrate higher efficiency and output growth, integrating asset-based productivity indicators into these frameworks could further strengthen their effectiveness. By explicitly linking incentives to higher CD productivity scores, policymakers can enhance firm-level financial stability across Indian industries with high default rates.
Our study provides significant insights for both external stakeholders, such as shareholders, lenders and investors, and internal stakeholders, including managers and strategists. Since all these stakeholders are concerned with evaluating default risk, understanding the role of asset productivity in reducing such risk is essential. For external stakeholders, investing in firms with higher asset productivity ensures more efficient capital allocation, while internal stakeholders can find these insights expedient to enhance productivity enhancements for mitigating default risk.
6. Conclusions and research limitations
Corporate default prediction has advanced significantly in the last few decades. Default risk studies have particularly attracted the attention of researchers in the wake of the global crisis (2008) when large conglomerates, such as General Motors (GM) and Lehman Brothers, filed for bankruptcy. In India, financial distress concerns creditors and the government because some non-financial industries have persistently been found in defaulting state over the last three decades. Prior literature exhibits that productivity has played a crucial role in combating the default likelihood of firms (Bryan et al., 2013). Nevertheless, the association between productivity and financial solvency has been inadequately investigated in the Indian context. This study examines this relationship and demonstrates how productivity leads to an enhancement in firms’ financial solvency.
Using a sample of 251 companies listed on the NSE, the findings demonstrate that higher asset productivity reduces the firms’ likelihood of default. The benefits of asset productivity extend from generating higher revenues to achieving competitiveness and cost efficiency. Productive enterprises are efficient at enduring financial shocks and volatilities. They possess greater flexibility in adjusting the level of production, managing inventories effectively and responding to changes in market demands. Hence, their resiliency lowers the risk of default and avoids the state of financial distress during financial downturns. The results, therefore, suggest that managers should incorporate asset utilization indicators, such as productivity derived through CD function into their strategic and financial planning efforts. To validate the robustness of the results, this study alternatively computes default risk from the market-based distance-to-default metric. The SGMM was further applied to ensure that results are free from endogeneity issues. The results from these regression analyses confirm that firms successfully achieving asset productivity reduce their probability of facing default risk.
The findings of this study can be interpreted based on the following constraints. First, sampling data is confined to industries with significant default rates. Researchers can broaden the sampling group to include other financial and non-financial industries. The sectors’ comparison can be conducted to analyse how industries with different productivity levels affect firms’ default probabilities. Second, the widely applied CD function to compute productivity suffers from reliability issues associated with the assumption of constant returns to scale (CRS). Many sectors have varied returns to scale, particularly during early development or mature phases. Therefore, CRS can affect the computation of productivity measurement and analysis. Third, the study has investigated the relationship between default risk and productivity in the Indian context. Future research studies can explore the dynamic relationship between bankruptcy and productivity in other parts of the world.
