Sunday, April 04, 2010
A Reply to my Critics on Local Debt
Victor Shih
Since the publication of my editorial in the Asian Wall Street Journal on local debt, there has been a wave of interest on this issue. Several investment banks have issued reports on local debt, and some of them have disputed my main finding that current local government investment vehicle debt stands at around 11.4 trillion RMB. The World Bank likewise addressed this issue and came up with a much lower estimate on local investment company (LIC) debt. In the discussion below, I outline some reasons why I still adhere to my estimate that existing local investment vehicle debt stands at around 11 trillion RMB. Furthermore, I once again reiterate that local debt is a serious problem which will require decisive actions from the Chinese government.
Some points people have raised about my estimate of local debt:
1. The Chinese government claims that there is only 6 trillion RMB in local investment vehicle debt.
My response: A. This widely cited figure was produced by a 6/2009 CBRC survey of the situation. The exact methodology is unclear, but informants state that the CBRC extrapolated this amount on the basis of a partial study of a few provinces.
B. Other government agencies have provided conflicting and higher amounts. For example, a MOF research team uncovered "well over 4 trillion" in late 2008 (excellent Credit Swiss research even states that the 4 trillion was a YE 2007 figure).
C. The CBRC finding concerns only bank loans, but total debt should also include bond issuance and accounts payable, which constitute triangular debt.
D. if we sum the gross debt of just the top 50 or so LICs, we quickly arrive at gross debt of over 2 trillion (try adding the gross debt of Guangdong Highway, Guangdong Transportation Group, Chongqing Highway, Beijing Basic Construction, Shanghai Urban Construction and Development Company, Shanghai Pudong Development Co., Tianjin Urban Basic Infrastructure, Binhai Development...etc.), so the remaining 8000 or so entities only owe 4 trillion (on average 500 mln RMB each)?
2. The 11.4 trillion is too high when compared with total bank loans in various categories.
My response: A. First of all, total loans outstanding at the end of 2009 was well over 40 trillion RMB, and I think it is completely reasonable to believe that nearly 1/4 of it was loans to LICs. In fact, I wouldn't be surprised that a higher share of bank loans ended up in LICs.
B. Some analysts have trouble believing that such a high share of medium and long-term loans ended up in LICs. When we consider how many LICs there are and the vital role they play in the local economic strategy, it is not surprising that likely as much as 3/4 of new medium and long term loans in 2009 ended up in LICs.
C. Beyond medium and long term loans, many LICs are holding companies with subsidiaries engaged in a wide range of businesses. For example, the LICs run thousands of hotels across China, and loans to these hotels would be classified as loans to the service industry. Thus, in addition to medium and long term loans and loans to infrastructure, it is perfectly reasonable for a sizable share of working capital loans, trust loans, and loans in the "other" category to end up in LICs. Again, gross debt of these entities would also include bond issuance and debt owed to each other.
3. LIC debt can be calculated by subtracting government spending on basic infrastructure from the total infrastructure spending figure. In that light, LIC debt only increased by 2.8 trillion RMB in 2009.
My response:
A. First, as pointed out, LIC are diversified holding companies which do not only engage in infrastructure construction. For example, thousands of subsidiaries of local investment companies engage in real estate development and absorb some share of the real estate loans. The figure generated using the method above, however, may be meaningful one-day when the government decides how much of the existing LIC debt it will seek to take over as part of a bail out.
B. The calculation above assumes that much of the extrabudgetary revenue from local governments derived from land sales went to infrastructure construction. According to excellent research done by Standard Chartered and UBS on land sales, much of the land sales revenue is spent on compensating original residents, leaving only a minority share for actual investment. Thus, a realistic application of this methodology would lead to something like 3.5 trillion RMB in new loans to LICs, not just 2.8 trillion.
4. My estimate of 12.7 trillion in future LIC debt is baseless and is way too high for YE 2011.
My response:
A. To be sure, I now think most of this debt will not realize by YE 2011 also. However, it would not be far-fetched to think that most of this debt will be realize by YE 2012. This estimate is not "baseless" as it comes from the hundreds of lines of credit that banks have granted to local governments. As long as banks more or less adhere to these lines of credit, they will lend this amount to local governments at some point in the future.
B. Although the State Council has called for more caution in lending to local investment vehicles, we still see local governments aggressively trying to raise money from the banks. Hubei, for example, has an investment plan worth 12 trillion RMB, and plans on investing 6 trillion RMB between now and 2012 (please see http://nf.nfdaily.cn/epaper/21cn/content/20100324/ArticelJ07002FM.htm). Of the 6 trillion, at least 3 trillion will come from bank loans and other forms of debt. If Hubei is able to realize its ambition, we are already 1/4 of the way toward my 12.7 trillion estimate. Thus, unless the central government harshly restricts overall credit, I think local governments at the provincial and municipal levels will have no trouble borrowing an additional 12.7 trillion by YE 2011 or 2012.
Beyond critizing my estimate, some investment bank reports also argue that whatever the debt amount, the Chinese government is fully capable of addressing this issue and in heading off a financial crisis. On this point, I mainly agree with my colleagues, but I still don't think the problem is trivial, especially in light that local governments seem determined to take on trillions in additional debt in the coming two years to finance ambitious investment plans. My main worry is that unless Beijing decisively restricts local investment projects, local investment companies will continue to borrow in large quantities in the coming two years.
Even relatively bullish investment bank report suggests that new non-performing loans in the banks can increase by 2-3 trillion RMB in the next couple of years. To be sure, this is well within the government's ability to handle and likely will not lead to any kind of financial crisis. However, this remains a daunting problem for the government and for current shareholders of China's banking stocks. This will require the China Investment Corporation to inject tens of billions of dollars into banks through Huijin. Additional asset management companies will have to be formed to take over the NPLs. This is a lengthy and difficult process involving numerous ministries and interests, which is expected to generate a great deal of uncertainty. If the expectation indeed is a couple of trillions in NPLs, it deserves careful watching rather than dismissal.
Finally, some investment bank reports suggest that the enormous sum of state assets must be considered along side of the debt. If debt ever becomes a problem, the Chinese government can always sell state assets to repay the debt. Here, I am in complete agreement with my colleagues. It will be a great day when the Chinese government decides to privatize trillions in state assets to raise money to repay local debt. The record of the Chinese government in privatization, however, is spotty at best. Even in the late 1990s, when the fiscal shape of the central government was at its weakest, only small SOEs were privatized, often through murky processes to insiders. Since then, both the central and local governments have done their utmost to maintain the dominance of large state-owned corporations through protectionism and subsidies from both the budget and the financial system. Instead of privatizing these firms and allowing them to compete on equal footings with private and foreign firms, they are given every advantage so that they can dominate the domestic and even the global markets. The financial system in particular channels the bulk of its resources to the state sector. Unfortunately, it does not seem privatization is anywhere near on the horizon. Instead, we can expect trillions more being poured into state entities, including local investment companies, in the foreseeable future.
Victor Shih
Since the publication of my editorial in the Asian Wall Street Journal on local debt, there has been a wave of interest on this issue. Several investment banks have issued reports on local debt, and some of them have disputed my main finding that current local government investment vehicle debt stands at around 11.4 trillion RMB. The World Bank likewise addressed this issue and came up with a much lower estimate on local investment company (LIC) debt. In the discussion below, I outline some reasons why I still adhere to my estimate that existing local investment vehicle debt stands at around 11 trillion RMB. Furthermore, I once again reiterate that local debt is a serious problem which will require decisive actions from the Chinese government.
Some points people have raised about my estimate of local debt:
1. The Chinese government claims that there is only 6 trillion RMB in local investment vehicle debt.
My response: A. This widely cited figure was produced by a 6/2009 CBRC survey of the situation. The exact methodology is unclear, but informants state that the CBRC extrapolated this amount on the basis of a partial study of a few provinces.
B. Other government agencies have provided conflicting and higher amounts. For example, a MOF research team uncovered "well over 4 trillion" in late 2008 (excellent Credit Swiss research even states that the 4 trillion was a YE 2007 figure).
C. The CBRC finding concerns only bank loans, but total debt should also include bond issuance and accounts payable, which constitute triangular debt.
D. if we sum the gross debt of just the top 50 or so LICs, we quickly arrive at gross debt of over 2 trillion (try adding the gross debt of Guangdong Highway, Guangdong Transportation Group, Chongqing Highway, Beijing Basic Construction, Shanghai Urban Construction and Development Company, Shanghai Pudong Development Co., Tianjin Urban Basic Infrastructure, Binhai Development...etc.), so the remaining 8000 or so entities only owe 4 trillion (on average 500 mln RMB each)?
2. The 11.4 trillion is too high when compared with total bank loans in various categories.
My response: A. First of all, total loans outstanding at the end of 2009 was well over 40 trillion RMB, and I think it is completely reasonable to believe that nearly 1/4 of it was loans to LICs. In fact, I wouldn't be surprised that a higher share of bank loans ended up in LICs.
B. Some analysts have trouble believing that such a high share of medium and long-term loans ended up in LICs. When we consider how many LICs there are and the vital role they play in the local economic strategy, it is not surprising that likely as much as 3/4 of new medium and long term loans in 2009 ended up in LICs.
C. Beyond medium and long term loans, many LICs are holding companies with subsidiaries engaged in a wide range of businesses. For example, the LICs run thousands of hotels across China, and loans to these hotels would be classified as loans to the service industry. Thus, in addition to medium and long term loans and loans to infrastructure, it is perfectly reasonable for a sizable share of working capital loans, trust loans, and loans in the "other" category to end up in LICs. Again, gross debt of these entities would also include bond issuance and debt owed to each other.
3. LIC debt can be calculated by subtracting government spending on basic infrastructure from the total infrastructure spending figure. In that light, LIC debt only increased by 2.8 trillion RMB in 2009.
My response:
A. First, as pointed out, LIC are diversified holding companies which do not only engage in infrastructure construction. For example, thousands of subsidiaries of local investment companies engage in real estate development and absorb some share of the real estate loans. The figure generated using the method above, however, may be meaningful one-day when the government decides how much of the existing LIC debt it will seek to take over as part of a bail out.
B. The calculation above assumes that much of the extrabudgetary revenue from local governments derived from land sales went to infrastructure construction. According to excellent research done by Standard Chartered and UBS on land sales, much of the land sales revenue is spent on compensating original residents, leaving only a minority share for actual investment. Thus, a realistic application of this methodology would lead to something like 3.5 trillion RMB in new loans to LICs, not just 2.8 trillion.
4. My estimate of 12.7 trillion in future LIC debt is baseless and is way too high for YE 2011.
My response:
A. To be sure, I now think most of this debt will not realize by YE 2011 also. However, it would not be far-fetched to think that most of this debt will be realize by YE 2012. This estimate is not "baseless" as it comes from the hundreds of lines of credit that banks have granted to local governments. As long as banks more or less adhere to these lines of credit, they will lend this amount to local governments at some point in the future.
B. Although the State Council has called for more caution in lending to local investment vehicles, we still see local governments aggressively trying to raise money from the banks. Hubei, for example, has an investment plan worth 12 trillion RMB, and plans on investing 6 trillion RMB between now and 2012 (please see http://nf.nfdaily.cn/epaper/21cn/content/20100324/ArticelJ07002FM.htm). Of the 6 trillion, at least 3 trillion will come from bank loans and other forms of debt. If Hubei is able to realize its ambition, we are already 1/4 of the way toward my 12.7 trillion estimate. Thus, unless the central government harshly restricts overall credit, I think local governments at the provincial and municipal levels will have no trouble borrowing an additional 12.7 trillion by YE 2011 or 2012.
Beyond critizing my estimate, some investment bank reports also argue that whatever the debt amount, the Chinese government is fully capable of addressing this issue and in heading off a financial crisis. On this point, I mainly agree with my colleagues, but I still don't think the problem is trivial, especially in light that local governments seem determined to take on trillions in additional debt in the coming two years to finance ambitious investment plans. My main worry is that unless Beijing decisively restricts local investment projects, local investment companies will continue to borrow in large quantities in the coming two years.
Even relatively bullish investment bank report suggests that new non-performing loans in the banks can increase by 2-3 trillion RMB in the next couple of years. To be sure, this is well within the government's ability to handle and likely will not lead to any kind of financial crisis. However, this remains a daunting problem for the government and for current shareholders of China's banking stocks. This will require the China Investment Corporation to inject tens of billions of dollars into banks through Huijin. Additional asset management companies will have to be formed to take over the NPLs. This is a lengthy and difficult process involving numerous ministries and interests, which is expected to generate a great deal of uncertainty. If the expectation indeed is a couple of trillions in NPLs, it deserves careful watching rather than dismissal.
Finally, some investment bank reports suggest that the enormous sum of state assets must be considered along side of the debt. If debt ever becomes a problem, the Chinese government can always sell state assets to repay the debt. Here, I am in complete agreement with my colleagues. It will be a great day when the Chinese government decides to privatize trillions in state assets to raise money to repay local debt. The record of the Chinese government in privatization, however, is spotty at best. Even in the late 1990s, when the fiscal shape of the central government was at its weakest, only small SOEs were privatized, often through murky processes to insiders. Since then, both the central and local governments have done their utmost to maintain the dominance of large state-owned corporations through protectionism and subsidies from both the budget and the financial system. Instead of privatizing these firms and allowing them to compete on equal footings with private and foreign firms, they are given every advantage so that they can dominate the domestic and even the global markets. The financial system in particular channels the bulk of its resources to the state sector. Unfortunately, it does not seem privatization is anywhere near on the horizon. Instead, we can expect trillions more being poured into state entities, including local investment companies, in the foreseeable future.
Comments:
Man alive, China too? Is there a country out there that is NOT in debt? Awesome stats, I know whom I sit under if directed to take classes on Chinese politics!
小史,
Thank you for enumerating the responses to your critics. The sheer volume suggests your research has made quite an impact on popular thinking.
You seem to invoke investment banks and their research quite frequently. While it is gracious of you to give their views air time, I think you can afford to be less conciliatory. Some of us who have published "research" on China at an investment bank (at least since 改革开放 took hold) would not take those reports' conclusions as seriously as you seem to.
As in many other circumstances, the road between the data and their interpretation is filled with unscientific potholes. No employee can afford to be critical of Chinese policy without somehow softening their conclusion and/or using the findings in a normative prescription for how the central government should do the "right" thing. As a recent example, take Citi chief economist Willem Buiter, he of the former Maverecon blog, in his March 24 note "Is China Blowing Bubbles?":
"Even if a large fraction of the loans made during 2009 as part of the government-mandated bank lending boost were to go bad eventually, this does not mean that the policy itself was misconceived from a macroeconomic and social welfare perspective...If even part of the additional productive capacity created through the bank-financed investment stimulus has a positive social rate of return, the policy is superior to paying unemployed workers to dig holes and then to fill them again."
Left unsaid, of course, is the very risk you have highlighted, that this large block of LIC lending has not been directed toward "additional productive capacity". Investment bank research on China has many fine qualities but balance is surely not one of them.
Thank you for enumerating the responses to your critics. The sheer volume suggests your research has made quite an impact on popular thinking.
You seem to invoke investment banks and their research quite frequently. While it is gracious of you to give their views air time, I think you can afford to be less conciliatory. Some of us who have published "research" on China at an investment bank (at least since 改革开放 took hold) would not take those reports' conclusions as seriously as you seem to.
As in many other circumstances, the road between the data and their interpretation is filled with unscientific potholes. No employee can afford to be critical of Chinese policy without somehow softening their conclusion and/or using the findings in a normative prescription for how the central government should do the "right" thing. As a recent example, take Citi chief economist Willem Buiter, he of the former Maverecon blog, in his March 24 note "Is China Blowing Bubbles?":
"Even if a large fraction of the loans made during 2009 as part of the government-mandated bank lending boost were to go bad eventually, this does not mean that the policy itself was misconceived from a macroeconomic and social welfare perspective...If even part of the additional productive capacity created through the bank-financed investment stimulus has a positive social rate of return, the policy is superior to paying unemployed workers to dig holes and then to fill them again."
Left unsaid, of course, is the very risk you have highlighted, that this large block of LIC lending has not been directed toward "additional productive capacity". Investment bank research on China has many fine qualities but balance is surely not one of them.
# posted by 
: 2:16 PM
: 2:16 PM
Dear professor, are the excellent research reports by the investment banks you mentioned in this blog post available online? I am very interested in them, but do not know where to find them for research. Thanks very much.
Post a Comment
# posted by : 8:58 PM
: 8:58 PM
