A China s Debt Bomb The United States has had some experience with the aggregator bank concept. The Resolution Trust Corporation was created by Congress to sell off illiquid assets in the wake of a real estate bust in the 1980s and the subsequent collapse of the savings and loan industry. But the RTC model has only limited relevance today. For one thing, all the property had been seized by government bank insurers when the savings and loan associations went belly up. For another, the scale of the problem was much smaller. In the seven years from 1989 to 1995, the RTC liquidated assets with a nominal value of $394 billion probably peanuts compared with the illiquid assets now giving the American financial system so much grief. Arguably, a better example to mine for insights about aggregator banks is the experience of China s asset management corporations over the last decade. China is worth a careful look in any case because the Chinese By Tong Li As this piece is being written, the Obama administration is exploring the idea of creating one or more aggregator institutions, bad banks, to take illiquid assets mostly mortgages and mortgage-related securities of uncertain value off the hands of the private banking system. Indeed, by the time you read this, the bad bank(s) may already have been created. experience with toxic bank assets is continuing, and how it evolves is likely to have a significant impact on the ability of the Chinese economy (and thus the world economy) to bounce back from the current crisis. Unlike the United States financial system, in which the securities markets and banks play roughly equal roles in intermediation, China s financial system has until very recently been dominated by the banking sector. Indeed, since the Mao era ended, banks have largely shouldered the task of funneling public and private savings to enterprises in particular, state-owned enterprises. Thus, for better and worse, the banks have fueled China s remarkable growth over the last three decades and have had great influence over which sectors were favored and which weren t. In the 1980s and early 1990s these lending operations were dominated by the Agricultural Bank of China, the Bank of China, the China Construction Bank and the Industrial 48 The Milken Institute Review
david cowles (all) Second Quarter 2009 49
china s bad banks and Commercial Bank of China. Although commonly referred to as commercial banks, the Big Four functioned like government agencies: investment decisions were not based upon assessments of risk and return with an eye toward maximizing profits, but on government policy. Not surprisingly, non-performing loans (NPLs) rapidly piled up on their balance sheets as funds were channeled to projects that met political objectives, rather than financial ones. A couple of specific factors contributed to the rapid accumulation of NPLs. First, China experienced rapid economic growth from 1992 to 1996, a period in which Beijing cruised full-speed-ahead on infrastructure projects. When growth slowed down over the next few years (due in part to governmental intervention to dampen inflation and in part to the Asian currency crisis), many of these projects became insolvent. Second, what little management of risk Tong Li is a senior research analyst at the Milken Institute. had been attempted by these state-owned banks broke down under the weight of politically driven decision-making. In particular, NPLs were considered the price that had to be paid to restructure (or to close) grossly inefficient state-owned enterprises that provided employment and a social safety network to tens of millions of Chinese. Hence, it came as no surprise when the government decided to bail out the nearly insolvent commercial banks in the late 1990s. Nicholas Lardy, now at the Peterson Institute of International Economics, has estimated that the Big Four as a group were technically insolvent in 1995 as bad loans exceeded their nominal capital. Others estimated that NPLs reached 39 percent of the banks loan portfolios in 1999. This translated to 2.5 trillion renminbi, or approximately 31 percent of China s GDP. To cope with the looming insolvency of the banking sector, Beijing created four asset management corporations (AMCs) one bad bank for each of the Big Four. Each AMC was initially capitalized by the government at 10 billion renminbi, in return for which the Ministry of Finance got all the stock. The AMCs task: buy the non-performing bank loans at face value using a combination of their own capital plus huge sums raised by selling bonds, by borrowing from other financial institutions and, most important, by borrowing from the central bank, the People s Bank of China. The asset management corporations beg for comparison with the United States Resolution Trust Corporation: all were created to hold toxic assets as equity with the sole purpose of liquidating these assets. There s one gigantic difference, however. The RTC inherited the assets of insolvent S&Ls, while the Chinese commercial banks sold their NPLs to the asset management corporations for far 50 The Milken Institute Review
more than they could possibly be worth. That did wonders for the commercial banks balance sheets. NPLs worth Maoknows-what were replaced with cash and other high-quality assets central bank bonds or AMC bonds fully backed by the Ministry of Finance. Thereafter, three out of the Big Four went public. (The weakest, the Agricultural Bank of China, will be restructured into a holding company, but will not be publicly listed.) At current capitalization, the ICBC and the Bank of China are among the largest commercial banks in the world. playing end-game The asset management corporations have used a variety of methods to unload nonperforming loans. Swaps, in which the AMCs accepted equity in state-owned enterprises (SOEs) in return for debt forgiveness, were common. This gave the SOEs leeway to restructure, and, liberated from debt, some have since turned a profit. Some NPLs were sold to private investors, domestic and foreign, at steep discounts. This method has been widely criticized, however, and the terms of transactions questioned. Among other issues, it increases the opportunities for corruption. Note that the primary method of financing the AMCs borrowing from a variety of sources at 2.25 percent interest rather than by printing money implies that someone will, in the end, have to forgive a great deal of AMC debt. To date, approximately one half of the NPLs on the AMCs balance sheets have been liquidated (a figure that includes NPLs purchased after 2003 at negotiated prices), with overall recovery rates of roughly 25 percent of face value. Projecting similar recovery rates, liquidating all of the assets of the AMCs over the planned 10-year period will barely cover the interest on AMC liabilities. To put it another way, when the four AMCs close their Total non-performing loans in China s financial system (2007) INSTITUTION Commercial banks: RMB BILLIONS State-owned commercial banks.................... 1,115.0 Joint-stock commercial banks...86.0 City commercial banks.................................... 51.2 Rural commercial banks................................... 13.1 Foreign banks...3.2 Development banks: China Development Bank................................ 18.5 Export-Import Bank of China...7.9 Agricultural Development Bank...64.3 Rural credit unions...644.5 People s Bank of China.................................... 1,033.4 Asset management corporations...906.2 Total nonperforming loans...........................3,943.2 source: the author doors for good, both China s commercial banks (which are holding huge quantities of AMC bonds) and the People s Bank of China may have to write off virtually all of the principal from loans to the AMCs unless, of course, the government chooses to swap AMC debt for cash or government bonds. Perhaps the most interesting question here is whether China s NPL situation has significantly improved after a decade s efforts. Beijing says that there have been steady declines in both absolute amount of NPLs outstanding and in the ratio of NPLs to the total assets in financial institutions. But these figures do not include NPLs held by some of the most troubled institutions, notably rural credit unions as well as the AMCs. Adding it all up, I estimate that China s financial institutions held a total of 3.94 trillion renminbi (about $570 billion) of NPLs at the end of 2007. This figure is a rough (and conservative) estimate. For example, some NPLs held by some institutions like urban credit unions and postal banks are not included because their financial reports are not publicly disclosed. Second Quarter 2009 51
china s bad banks There s no doubt, however, that NPLs of this magnitude are more than an irritant to the efficient functioning of the Chinese economy. Still, non-performing loans, which exceeded one-third of GDP in 1999, are down to 16 percent today. And by almost any measure, the Chinese economy is better prepared to weather financial disruptions now than it was a decade ago. For example, foreign exchange reserves are approaching $2 trillion up twelvefold since 1999. Instability is still a concern, however, if the debt holders ever opted to dispose of these NPLs in a relatively short period. One should always bear in mind that when NPLs are transferred from one financial institution to another, the underlying risks posed to the entire system are not reduced. Commercial banks may seem much healthier in the wake of the transfer to the AMCs. Nonetheless, these NPLs (and thus the risks) remain in the system remember, the commercial banks took AMC paper in return for their NPLs and will have to be digested over a long period. It took the four AMCs more than eight years to dispose of roughly one trillion renminbi of NPLs. Therefore it is not unreasonable to conclude that it will take another decade, if not longer, to dispose of the rest. And this assumes that NPLs won t continue to build. To date, reformed Chinese commercial banks especially those publicly listed have done a good job of containing their exposure to NPLs. Outstanding bank-held NPLs dropped sharply between 2002 and 2005 as lenders dumped them on the AMCs or worked them out with borrowers. Since then, the value of outstanding NPLs has hardly budged, while the volume of outstanding credit has risen by more than 40 percent. But much has changed, of course, in the global economy in the last two years. The International Monetary Fund estimates that China s real growth rate was slightly below 10 percent in 2008, the lowest since 2003. And there s every reason to believe that growth will slow sharply in 2009, as China s export markets shrink. China s underdeveloped, wellcontrolled financial markets have not experienced the horrifying collapse of liquidity experienced in the developed countries in the last six months. And since China lacks a mature credit-rating system for businesses and individuals, there is no practical way to assess the impact of recent events on bank balance sheets. But some deterioration in the quality of bank assets is inevitable in the midst of a global recession. In the first six months of 2008, Chinese commercial banks increased their ratio of loan-loss reserves to non-performing loans by 5.6 percentage points to an historic high of 44.8 percent. And as of December 2008, Chinese financial institutions have written down more than $5 billion as a result of the credit crisis. This is good news in the sense that the banks are not entirely ducking market realities. But it is also a reminder that, in this economic climate, good-quality assets can become illiquid very quickly, or even slip into the NPL column. That s where the latest government policy initiatives fit in. Last November, Beijing announced an economic stimulus package of four trillion renminbi (about $580 billion), with most of the money aimed at job-creating infrastructure development. And the ministries in charge got the message: many shovel ready projects that had yet to be evaluated carefully were hastily begun in the last two months of 2008. Throwing money at the problem isn t necessarily a bad thing: after all, the Obama administration s stimulus package will certainly sacrifice some efficiency in the name of rapid 52 The Milken Institute Review
job creation. But of the four trillion renminbi, only 1.1 trillion is slated to come from the central government s budget. Municipal governments, government-owned policy development banks and commercial banks are expected to deliver most in the form of credit. And this has to cause worry that these new, poorly collateralized loans will end up on the banks balance sheets. Thus, using the nascent private banking system as a source of countercyclical financing has opened the door to yet another, potentially very serious, accumulation of NPLs. I m reluctant to second-guess China s efforts to decouple its economy from the global recession. Policymakers in every country have been forced to act hastily, and mistakes are inevitable. But much is at stake here in terms of the continuing development of a modern financial system capable of sustaining China s leap to affluence. And it would be unfortunate if the methods the government uses to offset collapsing demand for Chinese exports today undermined the efficiency of the financial system for decades to come. There are parallel concerns, incidentally, for the U.S. financial system. The threats to the Chinese system are more daunting (a) because the government has yet to figure out how to unwind NPLs left over from the 1980s and 1990s, and (b) because China is depending on commercial banks to finance a good part of the current stimulus. But there is bound to be pressure there already is pressure on U.S. banks to do their part by making loans that may or may not be prudent. This seems different only in degree from the Chinese government s diktat that the banks finance new infrastructure in a hurry. By the same token, there are bound to be enduring long-term consequences to the degree of government intervention in banking, undermining the efficiency of financial intermediation. m Second Quarter 2009 53