Chinese banks put to the test of RMB8tn of Covid-19 problematic loans

July 03 2020

Executive summary

  • In China, financial and monetary conditions are clearly easing, as measured by our credit impulse index: as of May 2020, it stood at 6.0pp, up from 1.9pp at the end of 2019. However, the PBOC’s stance appears to be much more prudent compared to other major economies. Monetary stimulus in China currently is one third of that in 2009-10. According to our forecasts, more than 3/4 of monetary easing for this year has already been implemented. As the economic recovery is on track, the PBOC is likely shifting from a proactive to a wait-and-see stance.
  • The ongoing credit easing amid the Covid-19 shock will worsen banks’ asset quality. We estimate that by the end of 2021, ‘problematic’ loans  could increase by +124% (+RMB7.7tn) compared to the end of 2019. We find that recapitalization tools available to banks (e.g. perpetual bonds, central government special bonds, profit concession) are enough to cover potential losses incurred by these problematic loans. The financial de-risking campaign launched in 2017 managed to improve China’s overall bank asset quality and reduce systemic risk, making a systemic event unlikely in the short term. That being said, individual non-systemic failures among smaller city-level and rural banks, like in 2019, are still likely.
  • What does this mean for companies? While we expect GDP to come back to its pre-crisis level in early 2021, and GDP growth at +7.6% for the full year, we forecast insolvencies to increase by +40% over 2020-21 v. 2019. Deteriorating credit quality will significantly push the number of zombie companies higher, mainly in the automobile, textile and agrifood sectors. These sectors have seen declining profitability and rising indebtedness since 2015.

Monetary stimulus is one third of that in 2009-10

The Covid-19 pandemic and its related lockdowns have resulted in an unprecedented shock to the global economy. Central banks across the world have been quick to react, increasing funding and injecting liquidity. While the global response has been synchronised (differing from previous crisis reactions in recent history), the intensities differ. In China, financial and monetary conditions are clearly easing, but the central bank’s stance appears to be much more prudent compared to other major economies. More precisely, we expect the balance sheet of the People’s Bank of China (PBOC) to expand by +5% of GDP in 2020, compared with +14% for the U.S. Federal Reserve and the European Central Bank.

In the U.S. and the Eurozone, the yearly expansion of central banks’ balance sheets is the fastest observed in recent history. The situation is different in China. We have built a proprietary credit impulse index to better gauge the efficiency of monetary policy, and compare it with previous rounds of stimuli. This index is defined as the change in new credit issued as a share of nominal GDP. For credit data, we use the monthly aggregate financing data published by the PBOC (in particular bank loans, shadow-bank financing, bonds and stocks) and government bonds data. Our credit impulse index has a predictive power on cyclical economic indicators, exhibiting in particular a 12-month lead on the official manufacturing PMI.

The PBOC used the many tools at its disposal to ease the monetary policy (see Table 1). As of May 2020, our credit impulse index stood at 6.0pp, up from 1.9pp at the end of 2019 (see Figure 1). This improvement was driven by an acceleration in bank loans, stocks and corporate and public bonds issuance. After two years of decline, driven by regulatory tightening, shadow-bank financing is now rising again, but at a slow pace.

Table 1: Main monetary policy measures in 2020

 Table 1: Main monetary policy measures in 2020
Sources: PBOC, Euler Hermes, Allianz Research
Going forward, we do not expect the credit impulse index to increase much further. This result is based on our expectations of nominal GDP for the rest of the year, and on guidance given by policymakers. In mid-June, PBOC Governor Yi Gang said that total outstanding loans would increase by about RMB20tn, and total outstanding credit by at least RMB30tn. As measured by our credit impulse index, this would mean that more than 3/4 of monetary easing for this year has already been implemented. This would imply that the peak of monetary stimulus this time round would be around one third of that of the 2009-10 stimulus, and two thirds of those in 2013 and 2016.

Figure 1: Credit impulse and breakdown (in pp)
Figure 1: Credit impulse and breakdown (in pp)
Sources: Wind, Euler Hermes, Allianz Research
This result, along with economic data turning more encouraging recently (e.g. PMI surveys surprising positively in June), suggest that the PBOC is likely to shift from a proactive to a wait-and-see stance. The pause in the policy rate (1-year loan prime rate) since May, after 30bp worth of cuts since the beginning of the year, is in line with this prudent approach. In terms of liquidity injections, as of May 2020, we estimate that the PBOC has released this year a total of RMB2.43tn (2.5% of 2019 nominal GDP) through lending facilities and open-market operations (RMB676bn – see Figure 2), and cuts in the reserve requirement ratios (RMB1.75tn).

Further cuts in the policy rate should be limited (to avoid weighing further on banks’ margins), but liquidity facilities are likely to be used further, notably for banks to support the forceful fiscal stimulus and public investment projects. In total, authorities plan to issue RMB8.5tn worth of government bonds, of which just RMB3.1tn has been done as of May 2020. Banks are indeed the main owners of government bonds, holding 72% of total government bonds, and 87% of local government bonds, as of May 2020.

Figure 2: PBOC net liquidity injections to the banking system (RMB bn)
Figure 2: PBOC net liquidity injections to the banking system (RMB bn)
Sources: PBOC, Wind, Euler Hermes, Allianz Research
On top of broad-based monetary easing measures, policymakers have made clear that banks need to particularly focus on areas and sectors most in need, namely the private companies and small firms. We have also created a credit impulse index focused on the private sector to gauge the efficiency of such targeted easing of the monetary policy. We find that the private sector credit impulse index has rebounded in a synchronised way with the overall credit impulse index (see Figure 3). That was not the case in the previous round of stimulus (in 2016) and probably means that support targeted to the private sector has been effective this time.

Figure 3: Credit impulse, overall vs. private sector (in pp)
Figure 3: Credit impulse, overall vs. private sector (in pp)
Sources: Wind, Euler Hermes, Allianz Research

Pockets of failures in the Chinese banking system still likely

Banks are playing and will play a central role in supporting the economic recovery post-Covid-19. Chinese authorities have asked them to both increase new lending and extend existing loans at reduced costs. This brings up concerns over the banking system’s asset quality, and is likely to be one of the reasons why policymakers are not implementing more aggressive broad-based credit easing.

The Chinese economy recovering more visibly in H2 2020 and into 2021 (as we expect) does not mean that we will see the end of the impact from the Covid-19 crisis. Loan and payment failures could occur later on as emergency relief measures are removed. Notably, policies were announced in June to facilitate the disposal of non-performing loans and temporary forbearance measures would also mean that loans would start to fail only from next year. Authorities have indeed recently suggested that Non-Performing Loans (NPLs) will increase in 2021, and asked banks to prepare with more provisions. We would also watch out for ‘special mention loans’, a category in China’s loan classification that includes loans that are not yet non-performing but may become problematic in the future.

The historical relationship between NPLs and GDP since 2010 suggests that our forecasted growth profile for China in 2020 could result in RMB7.7tn worth of loans becoming problematic (see Figure 4). We use this expression to include both NPLs and ‘special mention loans’ , a category in China’s loan classification that includes loans that are not yet non-performing but may become problematic in the future. This implies a +124% increase from the end of 2019 (RMB6.2tn in total then, of which RMB2.4tn NPLs and RMB3.8tn special mention loans).

Figure 4: Problematic loans (RMB bn)

Figure 4: Problematic loans (RMB bn)
Sources: Wind, Euler Hermes, Allianz Research
The reassuring news is that the Chinese banking system is now less fragile than it was a few years ago. Indeed, a de-risking campaign of the financial system had started in 2017. It managed to improve overall bank asset quality, and reduce interdependence within the banking system (and thus systemic risk) – see Table 1. Note in particular that medium- and small-sized banks experienced larger changes, with inter-bank funding as a share of total funding declining from 11.7% to 8.2%.

Table 2: Impact of financial sector de-risking campaign started in 2017
 Table 2: Impact of financial sector de-risking campaign started in 2017
Sources: Wind, Euler Hermes, Allianz Research
That being said, an overall more robust situation does not mean individual failures of small banks could not still happen. In 2019, prior to the Covid-19 crisis, Baoshang Bank was taken over by the state, Bank of Jinzhou was bailed out and Hengfeng Bank was recapitalised. Small banks (city and rural commercial banks) continue to be the vulnerable section of the Chinese banking system, exhibiting higher NPL ratios but lower provision ratios (see Figure 5). At the end of 2019, the PBOC was warning that 13.4% of banks in China presented a “high risk”, many of which were rural banks. In particular, c.40% of rural financial institutions were rated “high risk”. Given that rural banks represent 13.3% of financial sector assets in China, this would mean that c.5% of total financial sector assets are “high risk”. As such, based on the PBOC data and given that there are enough options to recapitalise the banking system if necessary (see Figure 6), financial risk overall should be manageable.

Figure 5: Asset quality comparison by types of banks
Figure 5: Asset quality comparison by types of banks
Note 1: Bubble size represents relative magnitude of non-performing loans.
Note 2: Regulation requires a provision coverage ratio of 120-150%.
Sources: CBIRC, Euler Hermes, Allianz Research
Figure 6: Recapitalization options and potential new problematic loans (RMB bn)
Figure 6: Recapitalization options and potential new problematic loans (RMB bn)
* These are bonds issued by the central government that banks can purchase and pledge as collateral in exchange for liquidity from the PBOC. In this chart, we have not taken into account a multiplier for liquidity leverage. The government announced a quota of RMB1tn of central government special bonds to be issued in 2020.
** The State Council called in mid-June for Chinese banks to make profit concessions to support economic growth.
Sources: Euler Hermes, Allianz Research

What does this mean for companies?

In the short term, the historical relationship between activity data and our credit impulse index suggests that the Chinese economy’s recovery into 2021 will be supported (see Figure 7). We expect China’s GDP to come back to pre-crisis level early in 2021, while growth for the full year should reach +7.6%, after +1.5% in 2020 (and +6.1% in 2019). As explained previously, macroeconomic data rebounding does not mean that loan and payment failures won’t take place. We expect business insolvencies to increase by +40% between 2019 and 2021.

Figure 7: Official manufacturing PMI and credit impulse

Figure 7: Official manufacturing PMI and credit impulse
Sources: National Bureau of Statistics of China, Wind, Euler Hermes, Allianz Research
In a medium-run perspective, accelerating credit raises the risk of maintaining zombie companies afloat. We look at indebtedness and profitability to find out which sectors are most at risk of zombification. The automotive, textiles, agrifood and machinery & equipment sectors were already exhibiting vulnerabilities prior to the Covid-19 crisis, with indebtedness already increasing and margins declining between 2015 and 2019. Most of these sectors (apart from machinery & equipment) have further been hit hard by the Covid-19 crisis (see Figure 8), exacerbating the risk that zombie companies might emerge in these areas of the economy.

Figure 8: Change in margin ratio and liability-to-assets ratio, 2019-end to May 2020 (red dots are sectors that already exhibited increase in indebtedness and decline in margins between 2015 and 2019)
Figure 8: Change in margin ratio and liability-to-assets ratio, 2019-end to May 2020 (red dots are sectors that already exhibited increase in indebtedness and decline in margins between 2015 and 2019)
Sources: National Bureau of Statistics of China, Euler Hermes, Allianz Research
Françoise Huang
Senior Economist for APAC
francoise.huang@eulerhermes.com