The global financial crisis of 2007-2008, stemming partly from underestimated credit risk in the real estate sector, highlighted the need for a deeper understanding of this risk and its influencing factors. China's real estate market has experienced a significant crisis since 2021, marked by defaults from highly leveraged companies such as Evergrande and Blu-ray. By September 2022, 173 real estate companies had defaulted on credit bonds, totaling nearly 170 billion yuan. This situation poses a substantial threat to China's financial system and economic health, prompting the People's Bank of China to label real estate as a major risk. While existing research focuses on factors like policy adjustments and excessive expansion, the role of political connections – informal government-enterprise relationships – remains under-examined. These connections can increase leverage, particularly for private companies seeking resource advantages. This paper investigates whether political connections impact the credit risk of private real estate enterprises in China, exploring the potential mediating role of excessive debt. The study uses panel data from 123 listed Chinese real estate companies (2008-2021), measuring political connection based on the government or political positions held by senior executives and credit risk using the KMV model.

Extensive research exists on real estate credit risk, spurred by the Asian financial crisis of 1997 and the 2007 US subprime crisis. Studies such as Davis and Zhu (2004) examined the relationship between asset value and credit risk, finding that rising house prices increase financing needs and lending, potentially leading to risk. Kim (2013) modeled default probabilities for commercial real estate loans. Eichholtz (2021) focused on the US real estate credit cycle and default risk. Manz et al. (2021) highlighted the role of regional economics and financing channels. In the Chinese context, Jin (2007) used the CPV model to analyze credit default risk and the impact of debt ratios. Hu et al. (2018) examined real estate credit risk measurement using various indicators. Research on political connections' influence on businesses exists. Faccio (2006) found that politically connected firms have higher asset-liability ratios and default rates. Yu and Pan (2008) showed how political connections can reduce tax burdens. Yu et al. (2012) and Yuan et al. (2015) examined the impacts on financing constraints and innovation, respectively. Recent studies like Li and Jin (2021), Ding et al. (2023), Liu and Zhao (2023), and Brahma et al. (2023) further explored the quantitative impacts of political connections on corporate performance and risk. However, few studies specifically address the relationship between political connections and credit risk within the real estate sector, leaving a research gap this paper aims to fill.

This study employs panel data from 123 A-share listed real estate companies in China from 2008 to 2021, excluding those with ST, *ST, or PT statuses or those with suspended listings. Financial data was sourced from the IFind database and the People's Bank of China. Political connections were manually compiled from annual reports, identifying senior executives who served in government, the People's Congress, or the CPPCC. A four-level scoring system was used, based on the level of political office held (national, provincial, municipal, county). Credit risk (EDF) was measured using the KMV model, which calculates the default probability based on equity value, equity volatility, and a default point derived from liabilities. The main regression model is: EDF = β₀ + β₁PCᵢₜ + β₂controlsᵢₜ + εᵢₜ, where EDF is the default probability, PC represents political connection, and controls include management shareholdings, proportion of independent directors, number of senior executives, operating income growth rate, M2, and tertiary industry GDP. The study employs OLS and fixed-effects models. Robustness checks include alternative variable measures (eliminating irrelevant observations, using a binary political connection variable, and a simplified Merton DD model), and addressing endogeneity through Two-Stage Least Squares (2SLS) using executive education background as an instrument variable, and propensity score matching (PSM) using multiple firm-level control variables to address selection bias. Excessive debt (LEVB) was calculated using a Tobit regression model to determine the difference between the actual debt ratio and a target debt ratio, based on firm characteristics and industry medians. Heterogeneous impact analysis was conducted by comparing private firms in real estate, construction, and manufacturing.

The study's baseline results consistently reveal a significant positive correlation between political connections and credit risk for private real estate listed companies, regardless of the inclusion of control variables or fixed effects. This suggests that higher political connections are associated with higher credit risk. The intermediary analysis indicates that political connections influence credit risk through excessive debt. Higher political connections lead to higher asset-liability ratios and more excessive debt, which in turn increase credit risk. Robustness tests using alternative variable measures, including replacing the independent variable with no political connection and a binary variable, and the dependent variable with a simplified Merton model, produced consistent results. Addressing potential endogeneity using 2SLS and PSM further confirmed the significant positive relationship. The heterogeneous impact analysis showed that this relationship is specific to private real estate companies. State-owned enterprises did not show this correlation. Within the private sector, the construction industry also showed a positive relationship between political connection and credit risk, but the impact through excessive debt was not significant. In contrast, there was no significant relationship between political connections and credit risk in the manufacturing industry. Furthermore, the impact of political connections on credit risk is significantly stronger after 2013, suggesting the availability of more diverse and decentralized financing channels, such as internet finance, increases the likelihood of excessive debt and amplifies the negative effects of political connections. The rise of internet finance since 2013 appears to exacerbate the issue.

The findings address the research question by demonstrating a causal relationship between political connections and credit risk in private Chinese real estate companies, primarily mediated by excessive debt. This is significant because it highlights an often-overlooked factor in understanding the recent crisis in China's real estate sector. The results underscore the potential risks associated with informal government-business relationships, particularly in sectors heavily reliant on government approvals and resource allocation. The sector-specific nature of the findings suggests that policies targeting credit risk management should be tailored to the specific characteristics of the private real estate sector, considering the unique incentives and constraints faced by these firms. The increased influence after 2013 points to the need for regulatory oversight of the rapidly evolving internet finance landscape and its impact on corporate debt levels.

This paper makes a significant contribution by demonstrating the positive link between political connections and credit risk in private Chinese real estate firms, explaining this relationship through the mediating effect of excessive debt. The findings underscore the importance of considering informal institutional factors when assessing financial risk. Future research could expand the definition of political connections, include a broader range of industries and geographical regions, and investigate the role of specific policy interventions in mitigating these risks.

This study has some limitations. The definition of political connection is relatively narrow, focusing primarily on formal ties with government officials. A broader definition encompassing informal connections could provide a more complete picture. The sample is limited to listed Chinese real estate companies, restricting generalizability. Future research could include unlisted firms and international comparisons. The KMV model, while widely used, has limitations in capturing the complexity of credit risk.