Red Capitalism: The Fragile Financial Foundation of China's Extraordinary Rise
Page 8
The approach would have cleaned up the banks completely and the liability would have been indisputably with that department with taxing authority. To have done so, however, the MOF would have had to include the required debt issues in its national budget and received the approval of the National People’s Congress. An unfunded IOU, in contrast, is entirely “off the balance sheet” (biaowai ) and would only have been approved as a part of the overall bank-restructuring plan approved by the State Council. Indeed, it is possible that the use of IOUs did not even require State Council approval, as these instruments are purely unfunded contingent liabilities. Contingent liabilities are not included, at least publicly, in the national budget, or anywhere else for that matter.
Then, of course, repayment of its IOUs does not rely on the national budget: it turns out that the banks themselves would be the sole source of cash for funding these payments. Footnotes in the ICBC’s audited financial statements and the ABC IPO prospectus indicate that IOU repayment would come from recoveries on problem loans, bank dividends, bank tax receipts and the sale of bank shares. In other words, the banks would be indirectly paying themselves back over “the following years” since it is entirely unlikely that the MOF would sell (or be allowed to sell) any of its holdings in the banks. Since such funding sources represent future payment streams, it appears that the co-managed funds simply hold the two banks’ NPLs on a consignment basis; they are a convenient parking lot. Given the experience of the AMCs in problem-loan recovery (see below), there is little likelihood that either bank could do much better. The establishment of the Beijing Financial Asset Exchange in early 2010 is highly suggestive of how the banks will dispose of the bad loans in those co-managed accounts. The shareholders of this new exchange, which is located in the heart of the Beijing financial district, include Cinda Investment, Everbright Bank and the Beijing Equity Exchange. Its stated mission is to dispose of non-performing loans by means of an auction process. Perhaps this exchange will lead the disposal process for the two banks. But what entities have the financial capacity to acquire large NPL portfolios and who will take the inevitable write-down? In the end, the MOF will have to issue a bond to cover the net remainder of both of its IOUs or else extend their maturities. Other than avoiding a discussion with the National People’s Congress, it is entirely unclear exactly what is gained by taking this approach.
All of this simply serves to focus the light on the one practical source of repayment: bank dividends. This takes the story right back to bank dividend policies noted in Chapter 2. As will be discussed in regard to CIC in the next chapter, the Ministry of Finance’s arrangement has significant disadvantages, even compared with the far-from-perfect PBOC model.
“Bad bank” performance and its implications
By the end of 2006, BOC, CCB and ICBC had all completed their IPOs and the AMCs shortly thereafter had finished their workouts of their NPL portfolios. Given the weight of the AMCs on each bank’s balance sheet, the question must be asked: how well did these bad banks perform their task? As of 2005, even after the second round of spin-offs, the Big 4 and the second-tier banks still had more than RMB1.3 trillion (US$158 billion) of bad loans on their books. The total of the first two rounds at full face value, together with those remaining as of FY2005, amounted to RMB4.3 trillion. The AMCs were funded by obligations totaling RMB2.7 trillion (US$330 billion), as shown in Table 3.4. These liabilities were all designed, even if poorly, to be repaid by cash generated from loan recoveries. Obviously, as the first portfolio of RMB1.4 billion and parts of the second group of portfolios were acquired at face value, repayment was an impossibility from the start. From their first day of operation, the AMCs were technically bankrupt, and practically little different from the “co-managed accounts” now used by the MOF.
At the end of 2006, when more than 80 percent of the first batch of problem loans had been worked out, recovery rates were reportedly around 20 percent—hardly enough to pay back the interest on the various bonds and loans. While recoveries from the second, largely “commercial,” batch suggest a higher rate, industry sources suggest that actual recoveries lagged the prices paid. As 2009 approached and passed, the Party was faced with the problem of how to write off losses that may have amounted to 80 percent of AMC asset portfolios, or about RMB1.5 trillion. But losses could easily have been even greater than that and even long-term industry participants are unsure just what this figure might be.
With some 12,000 staff, the AMCs had their own operating expenses, including interest expense on their borrowed funds. An estimate of operating losses exclusive of any loan write-offs is shown in Table 3.5. The table uses loan recoveries as a source of operating revenue, an incorrect accounting treatment. But reports indicate that, indeed, the AMCs did use recoveries to make interest payments on their obligations to the PBOC and the banks. Had they not, the banks would have been forced to make provisions against the AMC bonds on their books or the MOF would have had to make the interest payments. There is no indication that this happened. For the sake of arriving at an estimated recovery, figures used in Table 3.5 are assumed to be 20 percent for loans acquired at full face value and 35 percent for loans acquired through an auction, where the AMCs are assumed to have paid 30 percent. Operating expenses are based on 10 percent of NPL disposals, as stipulated by the MOF.
TABLE 3.5 AMC estimated income statement, 1998–2008
Source: Caijing , May 12, 2008; 77–80 and November 24, 2008; 60–62
Note: US dollar values: RMB 8.28/US$1.00
RMB billion
1st Round: 1999–2003 FY2003 US$ billion
Total acquisitions 1,393.9 168.4
Disposals to 2008 1,156.6 139.7
Recovered, assume 20% 231.3 27.9
less:
Interest expense on PBOC loans/AMC bonds, 1999–2003 190.0 22.9
Operating expense, assume half of MOF target 10% of disposals 57.9 7.0
Total operating expense 247.9 29.9
Pre-tax gain/loss −16.6 −2.0
Registered capital 40.0 4.8
Retained earnings −16.6 −2.0
Accumulated write-offs—Round 1 −925.3 −111.8
2nd Round: 2004–2005 FY2005 US$ billion
Total acquistions—face value 1,639.7 198.1
Total acquistions—auction value, assume 30% 491.1 59.3
NPLs remaining from Round 1 237.9 28.6
Assumed disposals—100% 1,639.7 198.1
Recovered on auction NPLs, assume 35% 171.9 20.8
Recovered, Round 1 remainders, assume 10% 23.7 2.9
Total recoveries 195.6 23.6
less:
Interest expense on PBOC loans/AMC bonds, 2004–2005 95.0 11.5
Operating expense, assume half of MOF target 10% of disposals 82.0 9.9
Total operating expense 177.0 21.4
Pre-tax gain/loss 18.7 2.3
Registered capital 40.0 4.8
Retained earnings −2.1 −0.3
Accumulated write-offs—Round 2 −533.2 −64.4
Write-offs − Round 1 + Round 2 −1,458.5 −176.1
The resulting analysis suggests that the four AMCs lost their RMB40 billion in capital entirely, with estimated write-offs of RMB1.5 trillion (US$176 billion) yet to be taken. This represents a loss rate of around 50 percent. While the profit or loss of the AMCs is only a rough guess, the amount of the write-offs is a more accurate figure and, what is more, they remain on the balance sheets of these four non-public, non-transparent enterprises.
The reason write-downs have not been taken is straightforward. A full or even partial write-off would lead to the outright bankruptcies of the AMCs, confronting the government with a difficult choice: either the banks would suffer significant losses on the AMC bonds or the MOF would have to bear the burden and explain to the NPC. At the outset of the reform process and the creation of the bad banks, their closure and full write-offs, including MOF payment on their bonds, had been part of the plan and explained as such.
&nb
sp; Over the years, however, the plan had been changed and the MOF had assumed responsibility as a result of its bureaucratic victory over the PBOC. Now, in 2009, the banks seemed to be performing like world-beaters and the AMCs were noisily talking up their panoply of financial licenses; everyone had deliberately forgotten the history. Why should the MOF rock the boat when it is far easier to defer any decision until a more convenient time?
This is just what happened. In 2009, as their bonds came due, the AMCs were not closed down and their bonds were not repaid. Instead, the State Council approved the extension of bond tenors for a further 10 years. To support their full valuation on bank balance sheets, the MOF provided international auditors written support for the payment of interest and principal. Each bank’s annual financial report contains language such as the following from CCB’s 2008 report: “According to a notice issued by the MOF, starting from January 1, 2005, the MOF will provide financial support if Cinda is unable to repay the interest in full. The MOF will also provide support for the repayment of bond principal, if necessary.” Of course, a “notice” is not quite a guarantee; the MOF would never commit itself in writing to that. It does mean that it will in some way support the repayment of these obligations, unless at some point it is unwilling or unable to do so. Guarantees always come due at inconvenient times, as their extension in 2009 indicates. Until then, CCB, BOC and ICBC continue to carry these bonds at full value. As Table 3.1 shows, a default, or even a write-down of their value, would significantly impair the capital base of these banks and inevitably require yet another recapitalization exercise.
THE “PERPETUAL PUT” OPTION TO THE PBOC
This review of how the asset-management companies were used to resolve the problem-loan crisis in the banks highlights perhaps the most important part of the banking system: the perpetual “put” the PBOC has extended to the AMCs. In fact, this “put” extends beyond the AMCs to the entire financial system and weakens any reform effort that might be undertaken. It is the Party’s shield against financial catastrophes. In the name of “financial stability” the Party has required the PBOC to underwrite all financial cleanups, of which there have been many—from the trust-company fiascos of the 1990s, the securities bankruptcies of 2004–05 to the banks—at a publicly estimated (and probably underestimated) cost of over US$300 billion as of year-end 2005 (see Table 3.6).7 With this option available to them, bank management need care little about loan valuations, credit and risk controls. They can simply outsource lending mistakes to the AMCs, perhaps on a so-called negotiated “commercial” basis, and the AMCs will be almost automatically funded by the PBOC.
TABLE 3.6 Estimated historical cost to the PBOC of “Financial Stability” to FY2005
Time Period Amount (RMB billion) Use
1997–2005 159.9 Re-lending to closed trust cos., urban bank co-ops, and rural agricultural co-ops to repay individual and external debt
1998 604.1 Re-lending to the 4 AMCs for first-round acquisition of bank NPLs
From 2002 30.0 Re-lending to 11 bankrupt securities companies to repay individual debt
2003 and 2005 490.2 Huijin recapitalizes BOC, CCB, and ICBC
2004–2005 1,223.6 Re-lending to 4 AMCs for second-round acquisition of bank NPLs
2005 60.0 Additional lending to bankrupt securities companies to repay individual debt
2005 10.0 Re-lending to Investor Protection Fund
Total 2,577.8
US$ billion 315.5
Source: The Economic Observer , November 14, 2005: 3; PBOC Financial Stability Report 2006: 4; Caijing , July 25, 2005: 67
The new Great Leap Forward Economy
Added to the still unresolved loans of the 1990s, the US$1.4 trillion lending binge of 2009 will inevitably lead to correspondingly large loan losses in the near future (see Figure 3.7). The borrowers and projects are the same as in the previous cycle—infrastructure projects, SOEs and local-government “financing platforms, which will be discussed further in Chapter 5. But this time, their scale of borrowing is much, much larger; the press has even taken to referring to this as “Great Leap Forward Lending,” harking back to Mao Zedong’s ill-considered Great Leap Forward of 1958–1961. In early 2010, the regulators and Party spokesmen have taken the line that such investments will pay off over time. This is being echoed by brigades of analysts the world over, but the implication is well-understood by the Party itself. As one official put it simply: “In the near term, there will be no cash flow.” In other words, a large portion of these loans, over 30 percent of which reportedly went to localities, are already in default. Can the demise of the AMCs as originally called for in 1999 really be expected when their use has proven to be so great? To what extent can the Ministry of Finance continue to issue its IOUs?
FIGURE 3.7 Incremental bank lending, 1993–2009
Source: PBOC, Financial Stability Reports, various
Against this background, it is not surprising that questions have been raised about the PBOC’s ability to continue to write the check for the Party’s profligate management of the country’s finances. It is interesting that the PBOC made public its own balance sheet for 2007 and that discussion around a recapitalization was rumored at about the same time (see Table 3.7).8 This may well explain, at least in part, the use of IOUs written by the MOF for the Agricultural Bank of China restructuring. The 2007 figures show that the central bank is leveraged at nearly 800 times its own capital.
TABLE 3.7 PBOC balance sheet, 2007
Source: PBOC, Financial Stability Report, 2008
It should not be surprising, therefore, that in August 2005, the PBOC created its own asset-management company designed to take “problems left over from history” off its own balance sheet. Huida Asset Management Company (Huida) was described in its brief appearances in the press as the fifth AMC and its operations since 2005 have remained mysterious since it did not sell its distressed-debt portfolios to outside investors. Huida was meant to operate as the twin to Huijin; Huijin made investments in the financial system that created problem assets while Huida was to collect on unpaid loans associated with such assets when and if they were taken on by the PBOC as part of its operations to maintain financial stability.
Huida, like Huijin, was a creation of the Financial Stability Bureau of the PBOC and all its senior management were staff in the Bureau, just as others were senior staff of Huijin.9 But unlike Huijin’s bank investments, the PBOC wanted to remove problem assets from its own balance sheet. Consequently, the actual equity investor in Huida had to be a third party and, given its close connections with the PBOC, Cinda AMC was the obvious choice (see Figure 3.8).
FIGURE 3.8 The establishment of Huida AMC, 2005
What were included in such problem assets?10 On Huida’s business license, the targeted assets were related to real-estate loans in Hainan and Guangxi and portfolios assumed as part of the GITIC and the Guangdong Enterprise bankruptcies. Interestingly, these figures are not included in Table 3.6, but can be estimated at around RMB100 billion.11 Despite such explicitness, financial circles at the time believed that the PBOC’s real intention was to put Huida in charge of working out the loans, totaling RMB634 billion, the central bank had made to the four asset-management companies in 2000. With a capitalization of just RMB100 million, whether it assumed the old problem assets or any part of the PBOC’s more recent AMC loans, Huida was going to be highly leveraged.
Assuming Huida did take on some or all of the PBOC’s AMC loans, such a transaction is illustrated in Figure 3.9. As previously described, the PBOC made loans to Cinda AMC in 2000 to enable it to purchase, on a dollar-for-dollar basis, problem-loan portfolios from China Construction Bank. These loans became assets on the balance sheet of the central bank that it then sold to Huida. Huida could only pay for such loan assets, however, if the PBOC lent it money in turn, which appears to have been the case. The net result of such a transaction was that Huida owned the loan assets associated with Cinda, while on its own books, the PBOC now held Huida
loan assets.
FIGURE 3.9 The transfer of AMC loan portfolios to Huida
The only problem with this arrangement is that Huida is a 100 percent subsidiary of Cinda AMC. In other words, Cinda’s loan obligations to the PBOC (and ultimately Huida) were being held by itself. If such accounts could be consolidated, then the assets would offset the liabilities and everything would just disappear! None of this makes sense, except from a bureaucratic angle: the PBOC was able to park problem assets off its own balance sheet and Cinda—as a non-listed, and undoubtedly non-audited, entity—had no need to consolidate Huida on its own balance sheet. At best, these loans became a contingent liability: if Huida could not collect, then the PBOC’s loan to Huida would not be repaid. As noted previously, contingent liabilities (biaowai zhaiquan ) are not considered to be real in China’s financial practice; where in the national budget report are such things mentioned? A look at Cinda AMC’s excellent website fails to provide any proof of Huida’s existence as a 100 percent subsidiary. One wonders if there is a sixth or even a seventh asset-management company lurking out there in China’s financial system.