While most traders over the past month have been obsessing over developments in Washington, the real action – most of it under the radar – has played out in China, where as discussed over the past few weeks, domestic liquidity has tightened notably, culminating with an unexpected bailout by the PBOC of various smaller banks who defaulted on their interbank loans as interested rates particularly on Certificates of Deposit (CD) – which have become a preferred funding conduit for many Chinese banks – soared. Ironically, these mini PBOC bailouts took place only after the PBOC itself decided to tighten conditions sufficient to choke off much of the shadow debt funding China’s traditional banks.

As a result, the interbank CD rate rallied strongly, leaving a narrower or negative spread for some smaller banks, whose legacy carry trades (see below for details) suddenly became unprofitable. Also, as reported last Tuesday, several small banks failed to meet overnight repo obligations. This liquidity tightness has been mainly due to escalating financial deleveraging, as the PBOC has lifted market rates and rolled out stricter macroprudential policy rules.

But all those events in isolation seem as merely noise against what otherwise appears to be a relatively benign, even boring, backdrop: after all, neither China’s stock, nor bond markets, has seen even remote volatility in recent months, and certainly nothing compared to what was experienced one year ago, when the Chinese turmoil nearly led to a bear market across developed markets. Then again, maybe the markets are simply once again behind the curve due to all the inherent complexity of China’s unprecedented, financialized and extremely complex pre-Minsky moment ponzi scheme.

Last last week, Deutsche Bank analysts led by Hans Fan released what is the definitive research report summarizing all the latest troubling trends facing China, which judging by capital markets, nobody is paying any attention to. They should, because as Deutsche Bank puts it, if taken too far, they threaten an “uncontrollable liquidity event“, i.e., the financial cataclysm that Kyle Bass and other perma-china-bears have been waiting for.

And, as usual, it all started with rising interest rates, which in turn is leading to increasing funding pressure, which if left unchecked, could lead to dire consequences for China’s underfunded banking system.

Here is a fantastic explanation of everything that has happened in China in recent weeks, and more importantly, what may happen next, courtesy of Deutsche Bank. We urge readers to familiarize themselves with the content as we will refer back to this article in future posts.

* * *

Only in early stage of financial deleveraging

China’s monetary policy has been shifting gradually towards a tightening stance since 2H16. Targeting the liabilities side of the banking sector, the PBOC hiked rates of monetary tools, such as MLF, SLF and OMO (Figure 1), and withdrew liquidity on a net basis after the Chinese New Year (Figure 2). At the same time, it targeted the asset side of the banking sector when it rolled out stricter MPA rules by including off-BS WMP credit in broader credit assessment and imposing stricter-than-expected penalties on banks that fail to comply.

As a result, the key indicators in the money market, including repo and CD rates, all suggest stretched domestic liquidity. For example, the 7-day repo rate, which is the most representative liquidity indicator, has exceeded the interest rate corridor ceiling of 3.45% several times this year (Figure 3). Moreover, the interbank CD rate spiked to 4.6% on 20 Mar 2017, up c.180bps from last year’s low (Figure 4).

We summarize in the below diagram recent financial deleveraging efforts by regulators.

Why push forward financial deleveraging?

We believe the PBOC aims mainly to contain the fast-growing leverage in China’s financial sector. In our view, the country’s financial leverage basically relates to speculators borrowing excessive wholesale funding to grow assets and chase yield, rather than relying on vanilla deposits. To measure this, we believe one of the good indicators of financial leverage is the credit-to-deposit ratio, calculated as total banking credit as a percentage of total deposits. The higher the ratio, the more fragile the financial sector, and the more likely the banking system will run into difficulties to finance unexpected funding requirements. Traditionally the loan-to-deposit ratio was widely used to measure system liquidity risk, but has become increasingly irrelevant in China, as banks are growing their bond investments and shadow banking books to extend credit.

As shown in Figure 6, the credit-to-deposit ratio in China’s banking system has risen sharply by 27ppts since 2011 to reach 116% as of February 2017. We see the rising credit-to-deposit ratio basically is a function of increasing reliance on wholesale funding to support strong credit growth. As of end 2016, borrowing from banks and NBFIs accounted for 17% of total liabilities, against 8% 10 years ago (Figure 7).

Which banks are more leveraged? Joint-stock banks and city/rural banks

As we have long argued, the risks are not evenly distributed in China’s banking system; there are notable differences in the balance sheet structures of different types of banks. As shown in Figure 8, medium-sized banks, which mainly include joint-stock banks, recorded the highest credit-to-deposit ratios and hence are most reliant on wholesale funding. At the same time, small banks, which mainly include city/rural commercial banks, also delivered notable increases in credit-to-deposit ratios, despite a lower absolute level. The credit-to-deposit ratio for small banks has increased by 30ppts since 2010, vs. 14ppts for the big-four banks in the same period.

On the liabilities side, medium-sized and small banks mainly rely on wholesale funding, i.e. borrowing from banks and NBFIs. As of 1H16, wholesale funding made up 31% and 23% for medium-sized and small banks, respectively, against only 13% for big-four banks, as shown in Figure 9.

A closer look into interbank CDs – funding pressure ahead

Wholesale funding for smaller banks has been obtained mainly by issuing CDs in the interbank market. Interbank CDs have supported 20% of smaller banks’ assets expansion over the past 12 months. Since the introduction of interbank CDs in 2014, CD issuance recorded strong growth and the balance jumped 89% yoy to Rmb7.3tr in Feb 2017 (Figure 10), or 3.4% of total banking liabilities.

Joint-stock and city/rural banks account for 99% of issuance (Figure 12). In the coming months these banks have ambitious CD pipelines. More than 400 banks announced plans to issue CDs worth Rmb14.6tr in 2017. This represents 60% yoy growth from the issuance plan in 2016. Investor-wise, WMPs, various asset management plans and commercial banks themselves are the major buyers, which combined make up 79% of the total balance (Figure 13).

However, we view banks that are more reliant on CDs as more vulnerable to rising rates and tighter regulations.

Reflecting tighter liquidity, the interbank CD rate has rallied strongly, with the 6-month CD pricing at 4.6% on average. Some CDs issued by smaller rural commercial banks have been priced close to 5% recently. This would have pushed up the funding cost and notably for smaller banks. If banks invest in low-risk assets such as mortgages, discounted bills and treasury bonds, this would lead to a negative spread. Alternatively, banks can lengthen asset duration, increase the risk appetite, add leverage or slow down asset growth. Among these alternatives, we believe a slowdown in asset growth is the most likely.

Caixin previously reported CDs are likely to be reclassified as interbank liabilities, capped at 33% of total liabilities. This potential regulation could add funding pressure for banks with a heavy reliance on interbank liabilities. With Rmb4tr interbank CDs to mature during Mar- Jun 2017 (Figure 16) and interbank liabilities exposure approaching the limit (Figure 17), joint-stock and city/rural banks are subject to notable funding pressure.

We show the listed banks’ issuances in the chart below. INDB, SPDB and PAB are among the most exposed to interbank CDs.

* * *

What are the implications?

Are we close to a “tipping point”?

For now, probably not, especially in a year of leadership transition. In our view, the risk of an uncontrollable liquidity event is low, as the PBOC will do whatever it takes to inject liquidity if needed. In the domestic liquidity market, the PBOC exerts strong influence in both the volume and pricing of liquidity. With 90%+ of financial institutions directly or indirectly controlled by the government, PBOC will likely continue to give liquidity support. In 2H15, the central bank established an interest rate corridor to contain interbank rates within a narrow range and pledged to inject unlimited liquidity to support banks with funding needs.

However, continuing liquidity injections do not come without a cost. A bigger asset bubble, persistent capital outflow pressure and a lower yield curve over the longer term are side effects that China will have to bear. At the same time, the execution risk of PBOC itself is rising.

Implications on system credit growth

We expect system credit growth to moderate from 16.4% yoy in 2016 (16.1% in Feb’17) to approximately 14-15% yoy in 2017 (Figure 23). As a result, the credit impulse is likely to trend lower from the current high level (Figure 24). The slower credit growth is mainly attributable to several factors: 1) a tighter liquidity stance to push up the funding cost of smaller banks and to force them to slow down asset growth; 2) further curbs on shadow banking; 3) a higher  bond yield to defer bond issuance; and 4) slower mortgage loan growth.

Appendix A – Liquidity flows in China’s interbank market

New deposits supported 55% of asset growth in China’s banking system in 2016. The remaining 45% of new assets were mainly funded by borrowing from PBOC (19%) and borrowing from each other (19%, including bond issuance). While borrowing from NBFIs remained flat for the entire system, it was the main funding source for medium-sized and small banks. We summarize the liquidity flows in China’s interbank market in Appendix A.

Liquidity injection from PBOC. Over the past 12 months, to offset the liquidity drain from falling FX reserves, the PBOC has injected a huge amount of liquidity worth Rmb5.8tr into the banking system, which is equivalent to 400bps of RRR cuts (Figure 29). Of this injection, 30% and 24% have been made to support joint-stock banks and policy banks, respectively (Figure 30). For details, please see our report, PBOC liquidity facilities: Doing whatever it takes, 23 January 2017.

Borrowing from interbank market. Policy banks and big-four banks are net interbank lenders, while joint-stock and city/rural commercial banks are net borrowers. Joint-stock and city/rural banks not only borrow from policy/big banks, but also from each other. This could potentially lead to stronger contagion effects if some of them run into liquidity stress.

Lending/borrowing between banks and NBFIs. There has been a sharp rise in net claims to NBFIs from banks (Figure 33). We believe this is due to rising shadow banking transactions and also arbitrage activities with funds self-circulating within the financial sector. Clearly as shown in Figure 34, small banks are key lenders to NBFIs

Appendix B – What is driving the financial leverage?

From the accounting perspective, we believe the rising credit-to-deposit ratio is mainly due to bank credit circulating back into the banking system as non-deposit liabilities. In normal cases, when a bank makes a $100 corporate loan or purchases a $100 corporate bond, the bank books the credit to a corporate on the asset side while it also books a deposit on the liability side. We show a normal case in Figure 35. However, if a bank’s money circulates back into the banking system, just like in the two cases we illustrate in the diagram below, the $100 deposit is removed but interbank borrowing or borrowing from NBFIs would increase by $100. While there are likely to be many variants of bank credit circulation, we elaborate on two cases in detail.

Case #1: Bank credit circling via NBFIs

It is well known that NBFIs have been serving as SPVs to channel shadow banking credit from banks to corporates in past years. What is  insufficiently addressed though is that NBFIs also have been acting as channels for bank credit circling. Let us show a simple example below:

• First, Bank A invests in an asset management plan packaged by an NBFI. This is booked as a receivable investment on Bank A’s balance sheet.
• Second, the NBFI invests further in a CD issued by Bank B. Bank B books the CD under interbank borrowing. The money circulates back into the banking system and no deposit is generated.
• In some cases, if the yield of the CD does not cover the cost of issuing the asset management plan, the NBFI will leverage up in the bond market by pledging the CD through repo transactions. The leverage could be built up by two transactions: 1) entrusted bond investment (“Daichi” in Chinese); or 2) entrusted investment (“Weiwai” in Chinese), which we discuss in detail in our 2017 outlook report.
• In this case we use the investment in a bank’s CD as an example. In reality it applies to investment in interbank CDs, interbank negotiated deposits and financial bonds issued by banks, which are all circulating money back into the banking system.

The bank credit circling through NBFIs is growing rapidly. This is evidenced by strong growth in banks’ receivable investments, which reached Rmb21tr as of end-2016 to account for 10% of commercial banking assets, as shown in Figure 36. This represents 80% CAGR in balance since 2013. The majority of these investments was made by medium-sized and small banks. Note that not all receivable investments are credit circling, but we believe it should make up a notable portion. We summarize the structure of banks’ receivable investments in Figure 38.

The NBFI here could be any trust company, broker, fund subsidiary or insurance company. We believe brokers and fund subsidiaries should be the key players, as their bond trading leverage in the interbank bond market is much higher than other participants (Figure 37).

Case #2: Bank credit circling via corporates

Corporate loans may circle back into the banking system as well. This is because many corporates use borrowed but idle cash to buy bank WMPs. Below is a simple example:

• Firstly, Bank A makes a loan to a corporate.
• Secondly, the corporate uses the loan proceeds to buy a wealth management product issued by Bank A.
• Thirdly, Bank A invests the WMP fund in a financial bond issued by Bank B. This corporate deposit would circle back to the banking system as a non-core liability.
• To make this process economic, in many cases it would require leverage. The corporate borrowing cost may be at 4%, but the financial bond issued by Bank B may only yield 3.5%. To compensate the yield shortage, Bank A has to entrust the WMP fund to a third party and to leverage up by pledging the bonds through repo transactions. This process is called entrusted investment (“Weiwai” in Chinese, or entrusting to an external party).

This type of transaction is not an individual case. As shown in Figure 39, corporates purchased Rmb7.7tr WMPs in 1H16. This accounted for 7% of total corporate debt in China, or 29% of total WMP AUM in the system. SOEs, large private corporate and listed companies enjoy ample bank lending resources with low interest cost. However, the lack of attractive investment projects in their own business prompts them to invest in the financial market (i.e. bank WMPs).