China: 1999 Asset Management Corporations

As part of this nationwide campaign to liberalize the command economy, Chinese authorities established four operationally similar asset management corporations (AMCs) to remove non-performing loans (NPLs) from state-owned commercial banks (SOCBs) (García-Herrero 2006; Fung and Ma 2002).

Officials launched the AMCs following a series of capital injections to the SOCBs, which began in 1998 with an RMB injection equivalent to USD 33 billion. This was followed in December 2003 by a targeted injection of USD 22.5 billion in the two “best-performing” SOCBs, China Construction Bank and the Bank of China (García-Herrero 2006). The third and final restructuring wave began in April 2005 with the injection of USD 15 billion into the Industrial Commercial Bank. All told, the Chinese government injected an estimated 20-24% of 2004 GDP into the banking system (García-Herrero 2006, 313).

The executive order was issued a year after the actual setup of the four AMCs and did not include any explicit operational terms or details (Fung et al. 2011).

Although the AMCs were mandated to conduct other operations aside from NPL management and disposal, it does not appear from sources consulted that they were given special powers for debt collection or recovery from debtors.

Cinda Asset Management Corporation, by far the most successful of the four AMCs in terms of cash recovery, was mandated to collect what it could from the borrowers; otherwise, it would repackage the loans for discounted sale on the secondary market (Fung and Ma 2002; Bonin and Huang 2000).

In addition to NPL purchase and resolution, the AMCs were responsible for managing debt-to-equity swap conversion agreements to support approximately 500 “large and heavily-indebted” state-owned enterprises (SOEs) (Fung et al. 2011; Fung and Ma 2002). According to Fung et al. (2011), the government involved the AMCs in this restructuring effort in order to achieve SOE profitability within three years.

Apart from its mandated responsibilities, the AMCs were allowed to involve themselves in certain commercial activities. Such operations included the brokering of securities in order to facilitate the sales of equity holdings through stock listings. The AMCs thus intervened in the markets on behalf of the government, competing directly with the private sector and potentially raising a conflict of interest between policy mandates and commercial incentives (Fung and Ma 2002).

The four AMCs were officially structured as SOEs, “state-owned non-bank financial institutions set up by the State Council,” until the Ministry of Finance was able to “propose solutions to final AMC loss” (Fung and Ma 2002).

The three institutions interacted with the AMCs as follows:

• The Ministry of Finance (MoF) controlled the chairmanship for each of the AMCs’ boards of directors.
• The People’s Bank of China (PBoC) issued licenses defining the AMCs’ business scope and supervised the corresponding state-owned commercial banks.
• The China Securities Regulatory Commission (CSRC) supplied securities-related business permits and regulated the AMCs’ business operations (Fung and Ma 2002).

In order to carry out its debt-for-equity swap operations and resolve legal or logistical disputes, the AMCs would occasionally have to interface with other public institutions, such as the State Economic and Trade Commission, the Ministry of Foreign Trade and Economic Cooperation, and state and local governments (Fung and Ma 2002).

While the four largest state-owned commercial banks had no “formal” or “direct” equity stake in their partner AMCs, the informal relationship was not clear-cut (Fung and Ma 2002). The AMCs were staffed by former bank employees, while the president of each bank also served as the designated party chief of the corresponding AMC (Fung and Ma 2002).

Beginning in April 2003, the newly established China Bank Regulatory Commission (CBRC) became the main supervisory and regulatory authority responsible for overseeing the four AMCs. However, the AMCs were still required to report on a monthly basis to the MoF and PBoC and undergo onsite inspections by central bank officials (Fung et al. 2011).

Ultimately, NPL transfers to the AMCs totaled approximately 13.5% of GDP by March 2005 (García-Herrero 2006).

The State Council allowed the AMCs to be financed by four sources: MoF equity, borrowed funds from the PBoC, AMC-specific bonds, and commercial borrowing from other financial institutions (Fung and Ma 2002).^FThere was no evidence that the AMCs borrowed from private-sector sources as of 2001. Assuming there was no outside funding, Fung and Ma estimate that the partner banks swapped RMB 1.4 trillion in NPLs for a total of RMB 1.168 trillion in new AMC bonds and RMB 232 billion in cash (from the MoF and PBoC) (2002, 4). For a more detailed (but still speculative) breakdown of funding sources, please refer to Fung and Ma 2002. There are no official documents describing the sources in detail. Fung and Ma (2002) provide estimates based on limited public information.

The MoF initially provided an equity capital injection of RMB 10 billion to each AMC; because the transactions do not appear in state budget documents, it remains unclear how this funding was raised. While some portion of the investment probably comprised in-kind contributions (such as operational equipment), Fung and Ma (2002) treat the equity as cash in their analysis. The authors estimate that the RMB 40 billion accounts for about 3% of total funding to the AMCs.

It is likely that a substantial amount of funding for the AMCs (about 14%, or RMB 192 billion) was provided by the PBoC in the form of cash lending; however, only one AMC, Huarong Asset Management Company, disclosed funding from the PBoC on its balance sheet. According to Fung and Ma (2002), the PBoC funding did not “seem to have the expected monetary consequences,” in that China’s overall reserves remained stable and average growth only amounted to 8% per annum at the time.^FFor a more detailed explanation of the monetary implications of PBoC financing, please refer to Fung and Ma, 2002. The authors (2002) speculate that some “offsetting movements,” namely the 65% decline (RMB 340 billion) in PBoC claims on the four SOCBs in 2000, contributed to the decrease in the banks’ liabilities. The SOCBs may have used their free reserves resulting from the NPL transfers to pay off their PBoC liabilities, or the central bank itself could have sterilized its lending to the AMCs by recalling its previous loans to the SOCBs. It is also possible that the PBoC financing might have been presented in the form of “liability transfers,” which would have involved the AMCs paying for their NPL purchases by assuming the SOCBs’ liabilities to the PBoC (Fung and Ma 2002).  

Most of the funding—83% of policy transfers, or RMB 1.168 trillion—was financed by the issuance of non-tradable AMC bonds to the big four state-owned commercial banks. The “general belief” in Chinese financial markets was that the government would not let the AMCs default on their bonds; “the Chinese AMCs [were thus] viewed as government-sponsored agencies” (Fung and Ma 2002). However, the government asserted a policy of “constructive ambiguity,” encouraging speculation about which institution (e.g., the MoF, the PBoC, or state-owned commercial banks) would bear the risk of default (Fung and Ma 2002). Without an implicit state guarantee, the AMC bonds would probably not have achieved a better risk rating than their underlying NPLs (Fung and Ma 2002).

The Chinese government set an interest rate of 2.25% on both AMC bonds and PBoC credit to the AMCs. This rate was based on the official one-year bank deposit rate in 2001. Using Fung and Ma’s (2002) estimates (MoF equity amounting to RMB 40 billion combined with the PBoC policy purchase of RMB 1.4 trillion), the four AMCs collectively borrowed RMB 1.36 trillion and thus owed over RMB 30 billion in annual interest.

The partner AMC-bank pairings were designated as follows:

• Orient Asset Management – Bank of China
• Great Wall Asset Management – Agricultural Bank of China
• Cinda Asset Management – China Construction Bank^FCinda also differed slightly in that it ultimately acquired NPLs from the China Development Bank, one of China’s three policy banks (Fung and Ma 2002, 1).
• Huarong Asset Management – Industrial and Construction Bank of China (Fung and Ma 2002)

According to García-Herrero et al. (2006), separation of NPL recovery operations between the four banks was seen to have had two advantages. First, it would be easier to divide and conquer the overwhelmingly diffuse client base and small loan amounts. Second, it seemed advantageous to assign specific AMCs their partner banks based on specialization (e.g., Great Wall would deal with NPLs from the agriculture sector).

At some unspecified point between the first and second NPL transfer, the AMC/partner-bank model evolved into a more traditional auction model. This allowed AMCs to place bids for the remaining assets (García-Herrero et al. 2006). Cinda Asset Management, which had far outperformed its counterparts in terms of maximizing asset recovery, won the bid from the four SOCBs by promising 30% asset recovery (Fung and Ma 2002; García-Herrero et al. 2006).

Chinese authorities faced a “daunting” NPL problem in the 1990s, mostly due to the significant amounts of policy loans left over from the centrally planned economy (Fung and Ma 2002). In addition, weak SOE performance and unregulated credit-risk controls on state-owned commercial banks contributed to the proliferation of bad loans (Fung and Ma 2002).

These “policy-based” transfers were also explicitly authorized and financed by the government (Fung and Ma 2002). In order to avoid issues of moral hazard, the government took responsibility for bank losses resulting from policy lending that occurred only before 1996. Officials declared an end to policy-driven NPL transfers in 2001 (Fung and Ma 2002).

Between 1999 and 2000, RMB 1.4 trillion in NPLs were transferred at face value from the state-owned commercial banks to four partner AMCs in exchange for government cash, credit, and AMC bonds. These transfers accounted for over 20% of the big four banks’ combined loan portfolios and amounted to about 18% of the country’s GDP in 1998. However, the scope of the NPL transfers may have been much greater than what was revealed at the time (Fung and Ma 2002).^FFor further analysis and commentary regarding the scope of uncertainties surrounding the NPL problem, please refer to Fung and Ma 2002.

The second wave of transfers employed an auction-based model to award NPLs to the highest-bidding AMC. Cinda Asset Management, which promised 30% recovery and had previously outperformed the other AMCs, won the auction (García-Herrero et al. 2006, 315; Fung and Ma 2002, 12).

The AMCs claimed an average NPL resolution rate of 49.4% by March 2005 (García-Herrero 2006).

If the AMCs were unable to dispose of all transferred NPLs by the specified end date of 2006, the MoF would either take over the nominal equity or directly write off the remaining bad debt (Bonin and Huang 2000). As of March 2005, the AMCs had collectively recovered RMB 136.3 billion (USD 16.9 billion) in cash by the disposal of assets (García-Herrero et al. 2006).

Chinese officials announced upon establishment that the AMCs would operate for ten years; however, this provision was never explicitly stated in the State Council executive order (Fung et al. 2011). This timeframe was decided upon such that sunset provisions could be included from the outset (Bonin and Huang 2000).

Later, the government set the end of 2006 as the deadline for the resolution of all NPLs held by the four AMCs (García-Herrero et al. 2006).