Market Fears on Chinese Banks Are Misplaced; Opportunity Abounds

By Edward Friedman, CFA, MBA

China and its financial position evoke fierce debates in the investment community as some are skeptical that its GDP numbers are reflective of the country’s economic realities and some believe that the banking system is about to collapse due to enormous amounts of non-performing loans (NPLs) and wealth management products (WMPs). Various organizations are trying to estimate the “true” level of NPLs in the Chinese banking system, with estimates running between 8% and 15.5%, as recently estimated by the International Monetary Fund (IMF).

It is important to note that not all Chinese banks are created equally. There are policy banks that are wholly owned by the Chinese government and which direct loans to areas the government would like to grow. There are smaller regional banks that employ a hybrid model of government-directed loans and merit-based loans, and then there are the largest banks that are run under international standards.

There is no doubt that debt levels in China have increased since 2009 and that WMP balances have increased dramatically in the past few years. In this article we ask whether these create systemic problems in China’s financial system that may lead to a financial crisis similar to the one the US led the world to in 2008 or if it is a problem concentrated in a few 100% government-owned regional banks. We argue that it is the latter and as a result not all banks should be painted with the same brush, leading the major banks to trade at bottom valuations.

Non-Performing Loans (NPLs)

Let’s start with non-performing loans. It is difficult to really assess the level of NPLs in the overall Chinese banking system. In a recent report, Citibank estimates that 8.6% of corporate loans are non-performing. We believe this number is highly overstated as the analyst reached this number by classifying all the loans of companies that cannot fully cover their interest costs from earnings as non-performing. In the next step the analyst added the total allowance (total loan loss cushion) in the banking system with the pre-provision operating profit and concluded that the whole system can absorb an immediate, as opposed to a gradual, increase in NPLs to 6.7% without actually hurting book value, meaning the system overall will not need to raise capital, nor will its capital ratios be negatively impacted even if NPLs rise to 6.7% from today’s reported level of 1.6%.

We have been conducting a similar exercise on the major banks that we hold, China Construction Bank (CCB) and Industrial and Commercial Bank of China (ICBC), each year since we bought them almost six years ago.

We also tried to see what NPLs the market is implying for these two banks. We increased provisions as high as possible and assumed a coverage ratio (loan loss cushion divided by NPLs) of 100% all the way until we lowered the banks’ book value to the market value.

The following table depicts how high NPLs can rise immediately without hurting book value. We assumed for the maximum figures, like Citigroup, a defaulted loan recovery rate of 20% and we also calculated implied figures with no recovery rate at all:

  Current NPLs Maximum NPLs Market Implied (20% Recovery) Market Implied (0% Recovery)
China Construction Bank 1.63% 6.1% 8.4% 6.6%
Industrial and Commercial Bank of China 1.55% 5.9% 9.4% 7.5%

Royal Bank of Canada, for example, can absorb an increase in NPLs to only 3.3% without hurting its book value, while its market implied NPLs are less than 0.3%.It is important to note that assuming no recovery at all is a very onerous assumption as only about 30% of these banks’ loans are unsecured, while the rest of the loans are either secured by real estate, deposits or some other form of collateral.

In their latest reports, the banks exhibited one interesting trend, an increase in exposure to personal loans, mostly mortgages, at the expense of corporate loans. The following table depicts the changes in the share of corporate and personal loans compared to total loans since 2009:

Loan Type H2/09 H1/16 H2/09 H1/16
Corporate 69.1% 64.2% 69.5% 52.4%
Personal 21.1% 30.5% 22.6% 34.9%
Mortgages (included in personal) 15.3% 22.6% 17.7% 28.6%

Source: Company reports

It is easy to see why this trend persists. Though overall reported NPLs of these banks are in a narrow range between 1.5% and 1.6%, non-performing corporate loans are around 2.6% at CCB and 1.8% at ICBC, while non-performing personal loans are 0.6% and 1.3% respectively. Personal NPLs have increased in the past few quarters, but these are still much lower than those exhibited by corporate loans.

Wealth Management Products (WMPs)

Let’s now turn to the other part that spooks the markets-wealth management products. Several reports that we read lately raised alarm over the overall banking system’s exposure to off-balance sheet WMPs, various investment funds, trust beneficiary rights and others. These reports claim that if the entire balance of these products goes bust, the respective banks’ equity will be wiped out and then some, creating a crisis akin to the global financial crisis.

We do not dispute that many banks, including the two that we have interests in, hold on- and off-balance sheet various products that are designed to enhance yield. The following table depicts the holdings of our banks in such products and their equity as of the end of 2015:

Product ICBC CCB
On Balance Sheet Assets 260 billion RMB 240 billion RMB
Off Balance Sheet Assets 3.32 trillion RMB 1.36 trillion RMB
Breakdown % of Total Value (On and Off Balance Sheet)    
Non-Standard Credits 20% 29%
Banks and Deposits 68% 55%
Equities 12% 16%
Book Value 1.8 trillion RMB 1.42 trillion RMB
On and Off Balance Sheet/Book Value 198% 112%

Source: Company reports

The part that is considered the riskiest in a wealth management product is non-standard credit assets (NSTCA). This part is mostly made of loans the bank will not make under normal circumstances. Some, obviously, may see this as a proof of their contentions that these are irresponsible products. However, we spoke to both banks and they explained to us the process of loan approval under this vehicle. One prime example is a loan against receivables. When adjudicating such a loan, the banks that we own will only lend to a client the bank already has a relationship with and it will check the entities that issued these receivables to ascertain that these will not go under.Source: Company reports     

The latest proliferation of WMPs and the potential risks they may pose to the Chinese banking system if done wrong prompted the government to limit banks’ exposure to them. Under these regulations, called Notice A, a bank cannot hold more than 35% of its total WMP balance in these NSTCAs, or no more than 4% of the banks total assets.

It is important to note that these two banks only sponsor and administer these assets for investors, sometimes by third parties, and so do most banks. The banks clearly state that the off-balance sheet exposure is non-principal guaranteed, therefore the bank clients technically have no recourse to the bank's assets in case of losses. About a third of the exposure mentioned in the table is to investment funds. These funds are akin to the mutual funds that companies like Vanguard, Fidelity or CI Investments manage. Though none of these companies wish to lose money for their clients, clients cannot claim their money from these companies in case of losses.

Why do clients technically have no recourse to the bank's assets? Since the principal is not guaranteed, clients cannot demand their money back in case of losses. Yet, as was proven during the financial crisis, clients who bought AAA rated asset-backed securities demanded their money from the issuing institutions and in some cases they did get it. There is no such precedence in China and the banks we talked to adamantly claim that clients have been warned that they will not receive their money back in case of losses. The regulator also does not take their word for it and lately banks are required to record WMP sales conversations in which the client must hear and consent that in case of losses, there will be no recourse to the bank.


Lately, we also read comments that the largest Chinese banks will be forced to recapitalize once the true scope of their NPLs is revealed and WMP products start defaulting, thus negatively impacting their capital. In addition, new rules on minimum required loss absorbing capacity, otherwise known as TLAC, may require the banks to raise significant amounts of capital in the near future.

We strongly disagree with these statements. ICBC and CCB have core capital adequacy ratio (equivalent to tier 1 capital ratio) of 12.6% and most of this capital is made of equity, retained profits and various reserves, not preferred shares that are capital eligible. Currently, the minimum required ratio is 9.5%, rising to 11.5% in 2018, so the banks have sufficient capital buffer.

Regarding the loss absorption capacity, that depends. The banks have to raise that capacity to 16% by 2025 and 18% by 2028. Though the banks do not report their current ratio, we believe it is somewhere around 12%. With ROE of 17% to 18% and payout ratio of 33%, we believe that by 2025 these banks may not even need to raise any capital, let alone any significant amounts, as many fear.

With high capacity of loan loss absorption and debunking a myth that WMPs will bring down ICBC and CCB, why is it that both trade at a price to book (P/B) multiple of 0.85x and price to earnings (P/E) of 5.5x? For comparison, Canadian banks that have much lower NPLs (which are probably artificially low today due to IFRS rules) and ROEs of ~15% trade at P/B of 1.5x-1.9x and P/E of 10x-12x. European banks, with all their problems (extremely high NPLs, weak economies, low but improving ROEs), trade at P/B ratios of 0.4x to 0.8x and P/E ratios of 8x to 11x.

We stress that in order for a bank to collapse, several events have to come together, as defaults of NPLs and WMPs alone will not do it. At the onset of the financial crisis, the main problem US banks, led by Bear Sterns and the infamous Lehman Brothers, suffered from was insufficient liquidity and large balances of non-liquid assets. ICBC and CCB are very plain vanilla banks as most of their investments are in government bonds and some equities. They have hardly any level 3 assets, the assets considered to be the least liquid—a characteristic shared by some of the better-performing European banks.  Italian bankIntesa Sanpaolo, for example, has 9% of NPLs, common equity tier 1 ratio (CET1) of 12.7% and liquidity ratio of over 100% and this bank is considered as one of the best managed banks in Europe.

In the late 90s and early 2000s, bad loans in China spiked due to imprudent lending to state owned companies (SOEs). The state bailed out its four largest banks through direct capital injection and state bond issues. Though not ideal, China will be able to do the same today due to its relatively lower government debt and there are indications that lately China has been injecting liquidity into the smaller banks which did lend imprudently to local governments and SOEs.

Since that crisis, competition entered the market and the four largest banks became more international, and due to that, the state, through the People’s Bank of China (PBoC), imposed regulations that restrict imprudent lending. Since then, management teams have also changed, and we now believe that these banks are better managed today than they were 15 to 20 years ago.

We purchased these banks almost six years ago. At the time, they traded at P/B ratios of 1.5x, yet despite the decline of this ratio all the way to 0.85x, a decline of 45%, our total return has been in excess of 20%, largely due to their high ROE and high dividend yield. Panic, fear and disbelief can rule the markets for a while but eventually economics, as reflected by ample capital, access to ample liquidity from the PBoC, ROEs of 17% to 18%, dividend yield of 5.5% and relatively low NPLs, will win and at that point investors will discover well-managed banks trading at the cheapest multiples in the world.