Bank of Jinzhou
Identifying the country of origin of the world’s largest banks at any given point in time seems to be a good way to predict financial crises.
In 1989, the top 10 largest banks in the world ranked by assets were all Japanese. In 2007, the world’s largest banks were RBS, Deutsche Bank and BNP Paribas – all European. Now, it’s China’s turn to enjoy the limelight:
Chinese banks occupy four of the top five spots.
Why are we seeing this pattern? A theory is that high USD banking assets of a given firm reflects excessive credit creation, overvalued currencies and overvalued assets on bank balance sheets. It is surprising to see four Chinese banks among the top five in the world, given that China’s share of global GDP is just 15% – despite a record high trade-weighted exchange rate.
Shorting Japanese banks in 1989 or Western banks in 2007 would have been good decisions. Can the same be said for shorting Chinese banks at the present point in time? I believe the answer might be yes.
In this blog post, I will delve deeper into Bank of Jinzhou, a provincial commercial bank that IPO’d in 2015. Despite humble origins, it has ramped up its risk exposure rapidly through short-term borrowing, investments in opaque financial instruments and very aggressive accounting.
Jinzhou… wait, where?
Not many people have heard of Jinzhou – and much less visited the city. It is right in between Beijing and Shenyang in China’s northeast Liaoning Province.
It is not a major city by Chinese standards, given its 1.1 million population. The economy is dependent on the petrochemical industry, textiles, pharmaceuticals and building materials. It is not a particularly wealthy city either: a GDP per capita of CNY 43,000, almost half of Beijing’s CNY 82,000 and CNY 50,000 for China as a whole. It is a sleepy northern Chinese city reliant on heavy industry and state-owned enterprises.
It is surprising that a major bank can spring up from such a backwater city. Bank of Jinzhou’s market capitalisation is now close to HK$50 billion – over US$6 billion. Total assets as per June 2016 were CNY 422 billion – about US$61 billion. For those who remember, Icelandic bank Kaupthing’s total assets reached €58 billion just before the GFC. In retrospect, the loan exposure of Icelandic banks turned out to the out of whack with Iceland’s ability to fund their international expansion. The same can probably be said for some of the Chinese regional banks.
Bank of Jinzhou has essentially grown from nothing to a major powerhouse in less than 20 years. The truth is that the underlying business is very sleepy – hardly growing at all. What the bank has been doing over the past five years of breakneck growth is ratcheting up its risk profile substantially, hiding this risk through aggressive accounting and thereby boosting bottom-line EPS numbers.
Adding value to shareholders?
The way commercial banks add value to its shareholders is by amassing customer deposits and then using these a low-cost base to fund loans whose interest rates reflect underlying credit risks. The combination of scarce banking licenses and deposit insurance allows banks to pocket rentier profit in excess of capital costs – because most customers care about safety of principal above everything else. As long as they know that deposits are worth 100 cents on the dollar, they are willing to accept lower deposit rates than they would have received in a fully competitive environment.
Deposit franchises have been particularly valuable in China thanks to the regulatory environment where PBoC sets lending and deposit rates that virtually ensure a fat spread to any bank sitting on customer deposits.
In an economy with excess financial savings, a pure lending franchise should not be as profitable. Because as long as borrowers have no problem getting a loan, why would third parties lend money to a bank and then let it relend the money to the ultimate borrower? Unless the bank has a superior ability of assessing the credit quality of borrowers, that third party can simply lend to the borrower directly, thereby pocketing the spread himself.
So from this perspective – that building a deposit franchise is how you add value to shareholders – it is very difficult to see that Bank of Jinzhou has added any value at all over the past few years:
Its demand deposits have stayed flat more or less, growing at just 8% per year – well below China’s overall M2 growth of 14%. From my own research, Bank of Jinzhou’s retail customer satisfaction appears to be pretty poor compared to the larger state-owned banks. The way that the bank has grown, has been to lure customers to the bank by offering attractively-priced short-term time deposits and constantly increasing its interbank borrowing. I suspect that most of these time deposits are not the savings account type sold through the bank’s own branches, but rather a type of interbank borrowing via certificates of deposit. The average cost of deposits has gone up from 2.2% to 2.8% over the past two years, reflecting the greater use of CDs. Interbank borrowing costs have been steady at about 5.5%, not too far from the overall prime rate in China. The bank now has a major reliance on wholesale funding that has to be rolled over constantly, exposing the bank to interest rate and refinancing risk.
The changes in the bank’s funding structure has been coupled with massive investments in what Bank of Jinzhou calls “debt securities classified as receivables”. Bank of Jinzhou’s traditional loans grew at a 16% CAGR, more or less in line with the overall M2 growth in China. But its debt securities classified as receivables grew at a 200% CAGR, from a very low base in 2012:
So what are these instruments? According to the annual report, they comprise “beneficial interest transfer plans” issued by trust companies, security companies, insurance companies and asset management companies – also known as trust beneficiary rights (TBRs). In essence, they offer banks a way to bypass the regulator’s interest rate and loan ratio controls. Their fair value is said to approximate their carrying value, i.e. the principal amount subject to provisioning for credit losses. But so far, the provisioning for credit losses has been incredibly modest so far: just 0.3% in 2015 compared to 1.9% for Bank of Jinzhou’s own loan portfolio.
Are these provisioning expenses for their beneficial transfer interest plans realistic? The average yield of 8.12% is now higher than for its normal loans. How can the yield for beneficial interest transfer plans be higher yet enjoy much lower credit losses? It’s a mystery, especially considering that overall market yields have come down in recent years – not up. In addition, the yield that Bank of Jinzhou is receiving on its portfolio of beneficial transfer interest plans has been rising successively, indicating ever-higher credit risk. It looks like the bank is going into riskier and riskier type of lending. Meanwhile, they are covering up the poor loan quality through these TBRs, just before the bank’s IPO and a follow-up private placement. If it is such a great business to borrow in the wholesale market at 5.5% and then reinvest these funds into TBRs, why aren’t Bank of Jinzhou’s lenders doing so directly?
What happens in the use of trust beneficial interest plans – also known as trust beneficiary rights – is that loans are moved off-balance sheet to a third party “trust company” in exchange for receivables.
In this arrangement, the bank still acts as a guarantor and retains all the credit risk. But it does not need to recognise credit losses. Since the loan is booked as a bank-to-bank loan, it enables the bank to circumvent regulatory requirements such as loan-to-deposit ratios, at least up until 2015 when China still had LDR requirements.
Bank of Jinzhou is not the only bank that has engaged in such financial engineering, but the scale of its operations is greater than almost any other Chinese bank. At year-end 2015, its shadow banking book reached 160% of its traditional loan book. This ratio will have grown further now that we are well into 2016.
We are now seeing signs that the regulator is stepping up its oversight. In May 2016, CBRC issued “Document 82” that according to UBS will “close a loophole that has allowed banks to build up trillions of RMB of loans outside their loan books as investments”. Analysts believe that CBRC will crack down on the use of these assets in understating NPL ratios. Will there be full NPL recognition for these assets? UBS believes the answer is yes: according to them “the regulation explicitly states that assets that bear loan-like credit risk should be provisioned accordingly”. Most likely, CBRC will give banks a grace period that will allow them to raise capital before the provisions are put into effect.
What’s more though, last month Caixin reported that government announced plans to set up a national centre to register trust products, and perhaps also allow trading in these securities. So far, the trading in trust beneficiary rights has been thin due to a lack of information of what the underlying loans are. Now, active trading in these assets may become reality: “China Trust Registration Co” was set up in the Shanghai Free Trade Zone in 2014 with a registered capital of CNY 3 billion. It is in my view a step closer to market-based pricing of products, similar to what Markit’s ABX Index achieved prior to the GFC. CBRC has not explicitly stated that we will see any kind of mark-to-market accounting for TBRs, but wouldn’t interbank lenders be somewhat concerned if and when they see them trading at cents on the dollar?
Given that China is implementing a form of Basel III, a proper weighting of trust beneficiary right loan exposure would make banks such as Bank of Jinzhou borderline undercapitalised, depending on how far the regulations will go. A one-off retroactive enforcement of the provisions in Document 82 would be devastating. The bank’s new loan-to-deposit ratio of 227% (including all TBRs) would technically not breach any regulatory requirement as China, but such a high ratio might still be seen as problematic from the regulator’s point of view.
In November, CBRC issued another proposal that urged banks to step up their corporate governance, including by assigning a department to oversee off-balance sheet risks and employing external auditors to assess their off-balance sheet businesses. We are getting closer to full recognition of the loan exposures that reside on regional bank balance sheets.
The role of Wealth Management Products
PBoC recently announced that wealth management products (WMPs) that are held off bank balance sheets will be included in its framework for estimating risk to the financial system. Bank of Jinzhou itself has greatly expanded its exposure to WMPs over the past few years. The line is not always that clear, because many customers buying WMPs issued by a major bank assume that their counterparty is the bank, when the bank is simply an intermediary. There have been several cases over the past few years, where banks have had to compensate WMP buyers for losses, even though the WMPs were technically off-balance sheet. In mid-2016, Bank of Jinzhou had a CNY 25 billion balance of principal-guaranteed WMPs and another CNY 25 billion balance of non-principal protected WMPs. In a circular loop, Bank of Jinzhou has also invested in another CNY 15 billion worth of WMPs. In my mind, this circularity can only be a reflection of outright speculation with either shareholder of depositor funds.
Reading through WMP prospectuses makes it clear that there is a connection between TBRs and WMPs. Zhongyuan Bank for example, explicitly stated in one of their WMP prospectuses that 10-90% were invested in publicly-traded securities and 10-90% in TBRs and similar instruments. So we know that much of the assets in WMPs are in fact traditional loans that were originally sold to trust companies to hide credit losses. But as long as the WMP market keeps growing, the mismatch between 90 or 180 day maturity of a WMP and the long maturities of thinly traded TBR assets may not be much of problem. Administrators can simply pay off WMP interest by issuing new WMPs. Only when the market starts shrinking will we know if the underlying assets truly have cash flows that can support the attractive yields that WMP offer.
Warning signs flashing red
In the year 2000, Bank of Jinzhou’s major shareholder “Huaqiao” transferred 10.5 million shares to “State-owned No. 777 factory”. Huaqiao later sued the 777 factory, claiming that the shares were transferred illegally without Huaqiao’s knowledge about the situation. The shares represented 9% of the total share capital at the time. Between 2008 and 2010 the company changed its biggest shareholder over 30 times – before bank was even listed. Why? It is unclear, but a theory is that the beneficial shareholders of 777 factory didn’t want the shares to be tracked back to them. According to the prospectus, the lawsuit had still not been settled by mid-2015, and the 2015 annual report makes no mention of it. If Huaqiao gets its shares back, Bank of Jinzhou will have to issue warrants at CNY 1.00/share equivalent to 7.5% of shares outstanding. So the new shareholders will have to take the burden via dilution, while the owners of 777 factory effectively were given a gift from them. 777 factory may sound like an old state-owned enterprise from the Khruschchev era, but it was set up in 1999 just prior to the share transfer and appears to belong to electronics company Jinzhou Huaguang, whose website coincidentally shares the same 777 moniker: www.hg777.com.
In 2011, Bank of Jinzhou applied for a listing in Shanghai. Shortly thereafter, China Securities Regulatory Commission (CSRC) questioned the legality of some of the company’s loans and whether they had breached requirements on collateral on some of their loans. CSRC also asked whether Bank of Jinzhou had provided loans to certain SOEs as part of deals to unload non-performing loans. This is akin to NPL roll-over, but with a third-party standing in between. And lastly, CSRC also queried on whether any relevant people in the bank had been involved in criminal offences. No smoke without fire? For whatever reason, Bank of Jinzhou scrapped its Shanghai IPO after waiting for 3 years and decided to list in Hong Kong instead. It could be because of the long waiting list for doing an IPO domestically. Or in my view, more likely because international investors just happen to be more gullible.
Proof of silly lending: According to this article, Bank of Jinzhou has had a very close relationship to high-flying solar power company Hanergy. One of Bank of Jinzhou’s major shareholders “Yinchuan Baota” has in the past supported Hanergy’s share price through open market purchases. Hanergy later issued 500 million shares in a private placement to a Yinchuan Baota subsidiary at a heavily discounted price. Bank of Jinzhou itself lent Hanergy CNY 8 billion, 70% collateralised against Hanergy’s own shares. Bank of Jinzhou sold on part of the exposure to Hanergy via Wealth Management Products to its customers, but still kept CNY 3.4 billion on its own balance sheet. Most of its exposure was reportedly in the form of BTRs, ending up as “debt securities classified as receivables” on Bank of Jinzhou’s balance sheet. Yet, even though Hanergy’s fell almost 50% in a day and was subsequently suspended, Bank of Jinzhou’s provisioning expenses for debt securities classified as receivables stayed low. The exposure was not insignificant – 70% of CNY 3.4 billion equals over CNY 2.4 billion, about 37% of Bank of Jinzhou’s 2015 profit before tax. The bank got out of the mess by a CNY 2 billion loan from two “independent” financial institutions and a CNY 2.6 billion repayment by now possibly-bankrupt Hanergy. It is a mystery where the money came from – does the money truly exist or did repayment simply reflect a circular loan arrangement where the ultimate loan has little to show for in terms of asset backing?
What will happen to Bank of Jinzhou?
China has been in a wonderful position of having had zero currency risk and zero credit risk over the past decade. It is unclear how long this situation can be sustained. Many investors and analysts believe that maturity mismatches and lengthening credit chains are nothing to be worried about, since the government is the ultimate backstop – the government would never let anyone fail. What they are missing is that the government bailing out every financial institution, and propping up every asset market in the country would require resources of such a scale that it would likely lead to currency depreciation. Investors are slowly waking up to this fact.
In practice, we will likely see a mix of both. The currency will continue to depreciate and credit risks will be passed on to outside parties, whenever possible. In July this year, Caixin reported that China Development Bank is calling on all financial institutions to stop lending to Liaoning Province government entities and Liaoning Province SOEs. This was in response to 7 defaults by a Liaoning steel maker. To clarify – Liaoning is the exact province where Bank of Jinzhou is located. This observation goes against the common view that China will simply follow Japan’s path towards a zombification of the economy.
At the same time, we should not expect a Lehman moment anytime soon. Financial repression for households in China have created the world’s highest savings rate, leading to an oversupply of financial savings in the banking system. Banks are not in lack of funding, and so they are unlikely to go bankrupt.
What is more likely to happen, in my opinion, is a gradual recognition of non-performing loans, after foreign capital markets have been tapped to exhaustion. Special mention loan ratios keep on increasing at a steady pace, and so the fundamentals are clearly getting worse.
Thought experiment #1: Imagine that the bank had not engaged in short-term borrowing via additional time deposits and interbank borrowing, and had not invested shifted assets to TBRs at all in the first place. Let’s further assume that the tax rate stays flat at 24%, in line with the historical effective rate. In that scenario, Bank of Jinzhou’s net profit would have been around CNY 1.9 billion in 2015, giving the bank a P/E ratio of 23.3x and a P/B of 1.9x.
Would you invest at a Chinese regional bank at a peak of a credit cycle at a 23x P/E ratio? No-one in their right mind would. Yet, the situation is likely worse than this. Because not only is the bank trading at a high multiple to its long-term earning capacity, it is also at risk of eventual recognition of credit losses from the new loan exposure it has taken on. It is also possible that the impairment charges the bank has taken on traditional loans is understated, given the potential for shuffling bad loans into its TBR portfolio.
Thought experiment #2: Let’s say that the regulator succeed in its effort to enforce Document 82, forcing all banks to recognise impairment losses in their receivables portfolios. I assume here that Bank of Jinzhou will have to make provisioning charges of 1.9% on their receivables book, in line with their traditional loan book. A fairly conservative assumption. Net profit would fall by about 50%, the Bank’s P/E ratio would rise to 16.7x and the P/B ratio would be 2.0x.
The big question is how the capital markets will respond. Will interbank rates go up? Will trust between financial institutions dissipate? Will depositors be scared when they find out that some of the WMPs were supported by nothing but hot air? After the news came in November out that bank WMP exposure would have to move on-balance sheet, interbank rates started rising to a 15-month high.
Public investors have also been exposed to continued diluted by recurring capital raises. Hong Kong bank stocks performed very well during 2016, leading some to suspect Huijin is pushing up prices in anticipation of equity issues. There is a prohibition of selling stakes in state-owned enterprises below book value, so in some sense Huijin acting in such a way would make a lot of sense. Indeed, in mid-December this year, Bank of Jinzhou issued another 1 billion of H-shares at HK$7.5/share in a private placement which will help it survive the expected transitioning to full NPL recognition.
If Xi Jinping continues to pursue social stability above all other goals, we should expect bank net interest margins to come down gradually. Lending rates will have to come down to support borrowers and avoid debt deflation. Deposit rates will have to go up to prevent capital flight, either through higher regulated deposit rates or by continued growth in shadow banking assets. Banks will be squeezed in between.
In the transition process to lower net interest margins, banks have been able to use financial engineering to avoid accounting for credit losses in higher-yielding loans. And they have been able to use WMPs to bypass deposit rate regulation and shift bad loans off balance sheets. But eventually, logic tells us that the yield of TBRs and loans should probably not exceed the sum of a bank’s interbank borrowing rate and any periodic impairment losses on the loans by a very big margin. As a regional bank “trying to participate in the big leagues”, and given the warning signs from its participation from the Hanergy debacle and CSRC queries, I don’t believe that the bank’s credit analysis is any better than what other banks are able to achieve. We may sooner or later find that Bank of Jinzhou’s ballooning TBR business will have been value-destroying in the end.
The ideal time to invest in a bank with a deposit franchise is just after a crisis when NPLs peak, when interest rates are relatively high. At that point in time, the currency will typically be weak, there will be little competition among lenders, share prices tend to be low and spreads tend to be fantastic.
In Mexico after the Tequila Crisis in 1995 for example, net interest spreads hit 20% for a short while and were still 12.1% when NPLs peaked at 5.8% in the year 2000 – five years into the crisis. The share price of commercial bank Banorte rose about 50x the following 15 years, while the currency stayed almost flat up until very recently – not a bad return.
Another example of a perfect timing for bank investment is Indonesia after the Asian Financial Crisis. In 1998 lending rates hit 32% and stayed at 27% well into 1999, while NPLs peaked at more than 30% in the early 2000s. The stock price of Bank Central Asia rose about 80x from the year 2000 until today, while the currency just fell another 30% or so – a fantastic investment.
China is in the opposite position right now. Due to a high savings rate, deposits are plentiful and so the competition for lending business is fierce. Spreads are in theory reasonable at about 3%, but the growth in shadow banking has pushed effective spreads down. The real effective exchange rate for the renminbi is close to an all-time high, with only minor depreciation over the past year. The NPL ratio is 1.3% and has shown signs of rising, after bottoming at 0.8% in 2011. The corporate EBIT interest coverage ratio is at a similar level to 2003, when NPLs were at 9.5%. In other words, the two most important factors behind bank profitability – the net interest margin and the provisioning rate – are both under pressure.
The massive growth that Bank of Jinzhou – a Northeastern regional bank with a weak deposit franchise – has been able to achieve is astounding. But a “growth bank” is somewhat of an oxymoron: banks that try to grow very quickly often end up taking excessive risks. I believe that Bank of Jinzhou is no exception. Its LDR is out of this world, and its exposure to illiquid, opaque TBR portfolios is simply massive. A clear catalyst has now come in the form of “Document 82”, where CBRC is explicitly stating that provisioning for credit losses in receivable portfolios will have to be accounted for properly. CBRC is also making it clear that shuffling loans into TBRs will not be allowed going forward. Active trading in these instruments will make their values more transparent, and additionally, recognition for underlying WMP exposure starting from early 2017 will also be problematic for banks engaged in this type of business. Who knows – banks may find way around the new regulation, or the government may find NPL recognition a threat to financial stability and therefore unacceptable. But my base case is that the new regulation can and will affect the potential for banks to continue this charade. From this perspective, at more than 20x P/E in long-term sustainable earnings power at a peak of a credit cycle, significant overhang in terms of bad debt exposure and solid near-term catalysts, Bank of Jinzhou makes for a good short in my opinion.