The financial safety net consists of a variety of measures set by governments to deal with financial risks, such as the deposit insurance system and various types of implicit government guarantees or bailouts (Walter & Weinberg, 2001; Demirgüç-Kunt & Huizinga, 2004; Schich, 2008; Scott, 2004; Hoggarth, Jackson, & Nier, 2005). As bailouts increase the probability of future bailouts (Black & Hazelwood, 2013; Duchin & Sosyura, 2014; Kashyap, Rajan, & Stein, 2008) and may lead to the moral hazard problem, a no-bailout event is essential for establishing a proper financial safety net. The government guarantee of financial safety net is an important channel linking banks and governments (Dell'Ariccia et al., 2018; Leonello, 2018).

Bailout / no-bailout events are often considered as quasi-natural experiments due to the uncertainty associated with related legislation and enforcement (Roman, 2019). The takeover of Baoshang Bank and the partial default on negotiable certificates of deposit (NCDs) in 2019 is a landmark event in China. On May 24, 2019, the People's Bank of China (PBoC) announced the takeover of Baoshang Bank by the regulatory authorities due to serious credit risks. On June 2, 2019, the PBoC announced that creditors with over RMB 50 million would be covered, while the coverage ratio was expected to be about 90%. The 2021 Financial Stability Report of China by the PBoC shows that claims below RMB 50 million were fully guaranteed, and about 90% of the claims above RMB 50 million were guaranteed. As bank debts have rigid redemption before the event, a partial default on the NCDs of Baoshang Bank has essential impact on the Chinese banking sector.

Although Baoshang Bank made two announcements on the delayed disclosure of annual report in the first half of 2018, the market did not have strong default expectation before the takeover. It was a strong medium-sized bank in China and NCDs were unsecured debts with high priority, i.e., see the trend of NCD spreads in Fig. 1. It was not until the announcement date for the takeover that the NCD issuance spreads of local and national banks1 began to diverge, which indicates that the takeover and NCDs defaults surprise the market expectations to some extent.

NCDs as uninsured interbank debt instrument, have a market interest rate, large issuance size, transparent information disclosure, and active trading. These characteristics make NCDs suitable for market monitoring. On December 8, 2013, the PBoC issued the Interim Measures for the Management of NCDs, while 10 banks were granted issuance qualifications. On December 12, 2013, five banks issued the first batch of NCDs in the country, i.e., China Development Bank, Industrial and Commercial Bank of China, China Construction Bank, Agricultural Bank of China and Bank of China. According to the Notice on Regulating Interbank Business of Financial Institutions (No. 127) issued by the China Banking Regulatory Commission (CBRC) in 2014, the interbank liabilities of banks cannot exceed 33% of total liabilities, while NCDs were temporarily excluded in the calculation, which enables banks to actively manage their liabilities. As Basel III capital regulatory reform tightened the liquidity requirements, small and medium-sized banks had incentives to issue NCDs as the risk weighting for interbank claims in the calculation of capital adequacy ratio was only 20% or 25%.

Fig. 2 shows that NCDs have witnessed a rapid growth since 2014. In the first quarter of 2018, the PBoC include NCDs with maturities less than one year in the interbank liability of Macro Prudential Assessment (MPA) for banks with assets no less than RMB 500 billion, which triggers the growth to slow down. By the end of 2021, the balance of NCDs reached RMB 14.8 trillion, which accounted for 5.7% of banks' total liabilities. However, it varies a lot over bank types, over 2% for the “big 6” banks, and around 10% for joint stock banks and city commercial banks in Fig. 3.

Since the Global Financial Crisis, the regulatory reform has focused on weakening the bailout expectation of systematically important banks (SIBs), which aims to strengthen the market discipline. It requires a higher tolerance for the occurrence of no-bailout events to increase the no-bailout credibility. With the Bank Recovery and Resolution Directive (BRRD) coming into effect in 2015, bail-ins became an integral part of bank risk resolution in the Euro area afterwards (Zenios, 2016). In 2017, bail-in bondholders of Banco Popular in Spain and three Italian banks suffered losses, which established the no-bailout credibility of “too-big-to-fail” institutions. Recent studies focus on the impact of no-bailout events on the market discipline (e.g., Schäfer, Schnabel, & Weder di Mauro, 2016; Lewrick, Serena, & Turner, 2019; Crespi, Giacomini, & Mascia, 2019; Tölö, Jokivuolle, & Viren, 2021; Mo et al., 2021; Liu, Wang, & Zhou, 2022), i.e., the effect of no-bailout events on the monitoring incentives of bondholders and banks' risk-taking.

A main obstacle of market discipline is the protection of financial safety net (Bliss, 2015), with the latter as an important channel of bank-government nexus (Dell'Ariccia et al., 2018). Extant studies mainly focus on the risk transfer between banks and governments induced by bailouts or no-bailouts of banks (Acharya, Drechsler, & Schnabl, 2014; Ejsing & Lemke, 2011; Feld, Kalb, Moessinger, & Osterloh, 2017; Lamers, Present, Soenen, & Vander Vennet, 2023). China has a market-preserving federalism both at the central and local government level (Jin, Qian, & Weingast, 2005；Montinola, Qian, & Weingast, 1995; Weingast, 2009), which is a type of multi-level fiscal federalism with economic devolution and limited political devolution. China's fiscal system has five levels of government, i.e., central, provincial, city (prefecture), county, and township (Shen & Zou, 2015). However, the existing literature rarely studies the sub-national politics in shaping the incentives of decision makers, especially in developing countries (Wibbels, 2005). There is a lack of a clear assignment of responsibilities among various levels of governments in China (Lin, Tao, & Liu, 2006). Due to the multi-level nature of bureaucracy, sub-national governments may respond to the no-bailouts events in a heterogeneous way. Thus, we focus on the changes of implicit guarantee induced by the no-bailout event and the risk transfer between banks and different levels of sub-national governments.

The contributions of this paper are threefold. First, we investigate the bank-government nexus in China from the fiscal federation perspective, which provides a framework for understanding the impact of financial safety net on bank-government nexus in similar fiscal federations. Second, we examine the impact of a landmark no-bailout event on the nexus between banks' financing cost and local governments' fiscal strength in a typical mega emerging market, which can cast light on the improvement of financial safety net in similar economies with widespread implicit guarantees. Third, we examine the impact of a no-bailout event on the implicit guarantee at various levels of sub-national governments in a multi-level fiscal federalism with economic devolution and limited political devolution, which has policy implications for similar economies to improve the governance of local banks.

The remainder of the paper proceeds as follows. Section 2 surveys the literature and develops the hypotheses. Section 3 presents the methodology and descriptive statistics. Section 4 shows the results. Section 5 conducts robustness tests. Section 6 add some extensions. Section 7 concludes the paper.