Make no mistake, there is a problem in the Chinese banking system

This report summarises a number of the key insights and analyses we use to support our strong hypothesis that there is a problem in the Chinese banking system and credit quality is deteriorating rapidly.

We make the following key observations:

  • Credit quality is clearly deteriorating – particularly in the corporate sector.
  • The deterioration in credit quality appears to increase in severity the smaller the bank becomes.
  • We note that banks are typically making use of the “special mention loan” classification to mask the true extent of the growth in non-performing loans.
  • There has been a clear build up in “guaranteed loans” – again, particularly in the smaller banks. Perversely, these loans are exhibiting some of the worst deteriorations in credit quality – and we believe most guarantees made on such loans are not credible.
  • We believe the large run-up in securitized loans (wealth management products) serves to: (i) understate capital requirements; (ii) overstate credit quality; and (iii) results in a large liquidity risk to banks which are operating such trust structures – and indeed the broader system.

Special thanks to UBS and Bernstein for supporting charts.


  • As illustrated by the chart below, the Chinese banking sector comprises the big five SOE banks; 12 joint-stock banks; a large number of smaller (and unlisted) city commercial and rural commercial banks; Chinese policy banks (such as China Development Bank); and a number of other banks including the Postal Savings Bank of China (PSBC) and many thousands of small rural banks.

Screen Shot 2015-08-20 at 4.56.38 pm

  • For Montaka’s purposes, only the big five SOE banks and a number of the JSBs are investable – based on size, liquidity and stock exchange constraints.
  • The key distinction between the two is that:
    • The big five SOE banks are the primary source of big-ticket financing to China’s largest SOEs and has the greatest access to retail deposits; while
    • The JSBs are relatively more exposed to China’s SME and micro-enterprise corporate sector. They have historically lacked the same access to retail deposits (versus the major banks) and have relied more heavily on inter-bank financing.


  • There is no question that credit quality is deteriorating – particularly in the smaller banks (the joint-stock banks, city commercial banks and retail commercial banks).
  • For these banks, the rough scorecard is as follows:
    • In 2014, reported NPLs were up +50%YoY;
    • Yet, if we include all overdue loans in the equation, these were up +100%YoY; while
    • If we just focus on the corporate loans (as opposed to retail loans), these have blown out by large multiples!
  • The increase in overdue corporate loans is illustrated by the chart below (in RMB millions). And this is for the big five SOE banks whose loans are relatively safer credits.

Screen Shot 2015-08-20 at 5.02.40 pm

  • At the smaller end of town, the rate of deterioration is far more dramatic.

Screen Shot 2015-08-20 at 5.03.07 pmNPL RATIOS CAN BE DECEIVING

  • Much more interesting that reported non-performing loans – which we believe the banks will do anything to keep at low levels – are “special mention loans” or SMLs.
  • SMLs include loans that are overdue for less than 90 days; or loans for which the borrower is current, but “there may be some potential issues that could adversely impact future payments,” as China Minsheng Bank (one of the 12 joint stock banks) puts it in its annual report. Examples of such issues include: the borrower has been loss-making for two consecutive accounting periods; or the sector of the borrower is facing well-documented financial distress.
  • Our favourite example of SMLs diverging from NPLs is Ping An Bank (another joint stock bank). While the reported NPL ratio has barely moved, the SML ratio has been steadily deteriorating.

Screen Shot 2015-08-20 at 5.03.34 pm


  • Unique to the Chinese banking system is the large prevalence of loans backed by guarantees. Indeed, 20% of all Chinese loans are backed by guarantees – and this proportion increases the smaller the bank becomes.
  • The problem with guaranteed loans, we believe, is that they are much riskier than perceived because the guarantees are often not credible. Indeed, many guarantors are often providing multiple guarantees for many banks’ loans. Furthermore, we believe the popularity of guarantee-backed loans stems from a desire to skirt capital requirements.
  • We believe guarantees are being paid for in order to reduce the capital requirement of the loan, while the bank continues to bear the underlying credit risk, practically speaking.
  • We also suspect, therefore, that guarantees are being sought on those loans that need them most – making them the most risky type of loan.
  • UBS also notes the difficulty in making recoveries on guaranteed loans and the associated discount that is applies to impaired loans by China’s asset management companies (AMCs):
    • “… the only solution to recovery on a defaulted guaranteed loan is often to take the guarantor(s) to court and try to compel a forced repayment or asset seizure. Not only does this tend to be more time-consuming, expensive and laborious than with an asset-backed loan, but the bank is more likely to be left with nothing—and this has reportedly frequently been the outcome in various locales in Jiangsu and Zhejiang provinces.
    • For the same reasons, these loans are unattractive to the AMCs, who often pay rates that are a fraction of what they would offer for an asset-backed loan as they do not want to end up in the courts any more than the banks do. This often means that in the event of defaults on these loans, banks are left in the awkward position of choosing between lengthy, costly legal battles versus selling those loans at prices well below what they normally sell NPLs for.”
  • By way of example of our concerns, consider China Minsheng Bank (HKEx:1988). In 2014, Minsheng had 33% of its loan book backed by guarantees.
  • The growth in Minsheng’s guaranteed loan book was +7%YoY, while the growth in reported non-performing guaranteed loans was +70% over the same period!
    • For context, against book equity of RMB 236B, the guaranteed loan portfolio is RMB 605B; and the impaired guaranteed loans are RMB 10B.
  • Furthermore, Minsheng appears to be readily providing guarantees to loans being held by other banks. In 2014, off-balance-sheet guarantees were RMB 205B, up +94%YoY!


  • We are concerned about the Chinese shadow banking system, which relates to the large extent to which short-duration trust products are sourced to fund long-duration assets – particularly in real-estate.
  • In recent years, we have observed a substantial run-up in trust company assets, as illustrated below. These have appeared both on balance sheet, and off balance sheet, of most Chinese banks.

Screen Shot 2015-08-20 at 5.04.08 pm

  • The benefits of such securitized loans are at least three-fold, we believe:
    • Regulatory arbitrage – risk-weightings on such securities are 25% versus 100% for loans. This reduces the capital requirements of banks and serves to inflate reported ROE, while also increasing the risk of said equity.
    • Lender fragmentation – by securitizing loans, the lending can essentially be fragmented across a very large investor base. This also enables newly issued securities to repay maturing existing securities (yes, this aspect is “Ponzi-like”).
    • Credit quality arbitrage – as a corollary of the above, the continual roll-over of securities into new securities makes it much more justifiable to report low NPL provisions. After all, if there is always a newly issued security that can repay a maturing security, then no security will ever be non-performing (again, the Ponzi framework). This is evident in bank financials that illustrate absurdly low provisions for these securities held on the balance sheet.
  • UBS compiled a useful analysis that illustrates what a revised 2014 CET1 ratio might look like for a number of the Chinese banks, if risk-weightings and loan provisions on owned wealth management products were normalized to the levels of typical loans. Shown below, for example, Minsheng Bank’s CET1 ratio would reduce from 8.58% to 7.83%; and Ping An Bank’s from 8.64% to 6.94%.

Screen Shot 2015-08-20 at 5.04.34 pm

  • Drilling into China Minsheng further, we make the following observations:
    • On balance sheet WMPs jumped by more than 5x in 2014 and now equate to 1.0x book equity.
    • WMPs and other investment securities held in unconsolidated off-balance-sheet trusts more than doubled in 2014 and now equate to nearly 4x book equity.
      • While owners of such off-balance-sheet securities may not have any legal recourse to China Minsheng Bank, in practice, there is a genuine risk that the government forces the banks to guarantee such securities owned by many retail investors.

Screen Shot 2015-08-20 at 5.05.07 pm

Screen Shot 2015-11-11 at 12.08.48 pmAndrew Macken is a Portfolio Manager with Montgomery Global Investment Management. To learn more about Montaka, please call +612 7202 0100.

Leave a Comment

Your email address will not be published. Required fields are marked *