The vanishing interest income of Chinese banks as an indicator of loan quality problems
The growth of the Chinese economy was already slowing down before the coronavirus crisis, posing challenges to the financial system. International bodies, such as the IMF, have been concerned about the state of the banking sector. The true quality of bank assets is unknown; not all observers take the reliability of banks’ official accounting figures for granted.
Like the Chinese economy, the Chinese banking system has grown fast, but unlike the real economy, growth in the banking system shows no signs of abatement. The Chinese banking system has become the largest in the world (see Figure 1), surpassing the total balance sheet size of the Euro area banking sector in 2016 (Cerutti and Zhou 2018). According to China Banking and Insurance Regulatory Commission statistics, the aggregate balance sheet was ¥290,000 billion i.e. about €37,100 billion, as of December 2019, whereas the aggregate balance sheet total of Euro area banks was €32,400 bill.
Most Chinese banks are government owned, and public officials can use bank lending as a policy tool. It seems that the banks have no tendency to favour well performing and creditworthy firms in lending decisions (Bailey et al. 2011). In contrast, expectations regarding banks’ tendency to favour public enterprises affect government-owned companies’ behaviour (Megginson et al. 2014). Against this backdrop, it may not be surprising that chronic loan quality problems are pervasive in Chinese banks.
Figure 1 Balance sheet total of Chinese depositary corporations, billions of CNY
A novel econometric technique based on a simple idea
The ratio of non-performing loans (NPLs) to total loans was officially about 1.8% at the end of 2018 (CBIRC 2019). However, according to FitchRatings (2016), around 20% of loans may actually be de facto non-performing. The IMF (2016) has also suggested that the amount of doubtful loans is larger than that officially reported.
Using a unique data set compiled at the Bank of Finland Institute for Economies in Transition (BOFIT), I investigate signs of manipulation in Chinese banks’ NPL data (Kauko 2020).1 I propose a novel method to derive an indicator of hidden loan quality problems. It starts with the simple observation that, all else equal, the interest income of a bank diminishes if claims on its customers turn into NPLs, regardless of whether the bank reports them or not. Therefore, the existence of hidden NPLs is more likely if a bank earns less interest revenue than comparable banks with similar loan portfolios. Hence, the gap between actual and potential interest income reflects the quality of the loan portfolio.
The econometric analysis was carried out in two stages. First, the stochastic frontier analysis was used to estimate rates of return on healthy loan portfolios which constitute the efficient frontier. The difference between the rate of return at the efficient frontier and the actual return of a bank is then used as a proxy for the bank’s hidden NPLs. Although the indicator may be affected by other factors, such as lending at favorable rates to some stakeholder groups, I do find that whenever a bank discloses more NPLs, this difference decreases, as would be expected.
Second, I test which factors explain developments in this indicator. For instance, the gap between the actual rate of return and its potential value remained broadly constant in 2013-2015 but surged in 2016 (see Chart 2). This coincides with slowdown of China’s economic growth rate around that time (Kerola 2019).
Figure 2 Shocks to hidden NPLs (measured on the vertical axis by the estimated coefficients of annual effects in the Stochastic Frontier Analysis)
Banks may have an incentive to hide the true amount of NPLs in order to keep wholesale financiers calm. Such an incentive may arise if a bank is concerned about its own creditworthiness and would be stronger if the bank is heavily dependent on wholesale funding. This suggests the following three hypotheses.
- Banks that rely heavily on interbank funding are more likely to hide NPLs.
- Banks with weak capital adequacy are more likely to hide NPLs.
- Banks with poor profitability (measured by ROE) are more likely to hide NPLs.
I test these hypotheses using GMM estimations based on data on 131 Chinese banks from 2013-2018. I find that banks that are heavily dependent on wholesale funding tend to have higher values of the hidden NPL indicator. Interestingly, when one controls for openly disclosed NPLs, wholesale funding no longer explains differences between banks. This is consistent with the view that interbank liabilities affect the proxy for hidden NPLs only because they affect banks’ incentives to openly disclose loan quality problems.
Instead, the evidence on the impact of capital adequacy and ROE on hidden NPLs proved to be weak and elusive. Surprisingly, strong capital adequacy has a positive correlation with the amount of hidden NPLs. This may indicate that strongly capitalised banks do not pass high-risk receivables on to the shadow banking sector.
Perhaps surprisingly, big banks are more likely to have credit quality problems, i.e. high values of the hidden NPL indicator. This may be related to the finding by Fungáčová et al (2020) that the biggest banks in China tend to be persistently inefficient.
Controls related to asset liquidity, dependence on short-term funding, lending to corporates, and the relative size of the deposit base are not statistically significant.
Overall, my findings are consistent with the view that a number of Chinese banks do not openly disclose credit quality problems. The situation seems to have worsened during the recent slowdown in macroeconomic development. While authorities may have reasons to encourage banks to grant loans to troubled companies, such practices come at the cost of good credit risk management.
Allen, F, X Gu, J Qian and Y Qian (2017), “Implicit Guarantee and Shadow Banking: the Case of Trust Products”, Imperial College London Working Paper.
Bailey, W, W Huang and Z Yang (2011), “Bank loans with Chinese characteristics: some evidence on inside debt in a state-controlled banking system”, Journal of Financial and Quantitative Analysis 46: 1795-1830.
CBIRC (2019), “China Banking and Insurance Regulatory Commission”, CBIRC Releases Supervisory Statistics of the Banking Sector — 2018 Q4.
Central Baking (2020), “Chinese banks probably concealing NPLs – Finnish paper”, 2 March 2020.
Cerutti, E and H Zhou (2018), “The Chinese banking system: Much more than a domestic giant”, VoxEU.org, 9 Feb 2018.
Chen, H, R Li and P Tillman (2019), “Pushing on a string: state-owned enterprises and monetary policy transmission in China”, China Economic Review 54: 26-40.
FitchRatings (2016), “China: Multi-Year Resolution of Problem Credit – Leverage Weighs on Asset Quality”, Special Report, 28 June 2016.
Fungáčová, Z, P-O Klein and L Weill (2020), “Persistent and transient efficiency: Explaining the low efficiency of Chinese big banks”, China Economic Review 59.
IMF (2016), Global Financial Stability Report, April 2016: Potent Policies for a Successful Normalization
IMF (2019) Global Financial Stability Report, October 201: Lower for longer
Kauko, K (2019), “Benford’s law and Chinese banks’ non-performing loans”, BOFIT Discussion Paper.
Kauko, K (2020), “The vanishing interest income of Chinese banks”, BOFIT Discussion Paper.
Kerola, E (2019), “In search of fluctuations: Another look at China’s incredibly stable GDP growth rates”, Comparative Economic Studies 61: 359-380.
Megginson, W L, B Ullah and Z Wei (2014), “State ownership, soft-budget constraints and cash holdings: Evidence from China’s privatized firms”, Journal of Banking & Finance 48: 276-291.
1 The results are also briefly reviewed in Central Banking (2020).