THE GENIUS OF NOURIEL ROUBINI: WHY CHINA CAN NOT OVERCOME IT'S POPULATION DILEMMA...A RICH COUNTRY FULL OF POOR PEOPLE...


Chronicles of a Crash Foretold: Visualizing the Sequence of a Chinese Hard Landing

By Adam Wolfe

Aug 22, 2011 12:30:00 PM | Last Updated

China’s surge in investment from 2008 coincided with a decline in its marginal product of capital and a massive increase in its debt-to-GDP ratio. 

Its marginal product of capital fell from an average of 1.0 in the first seven years of the decade to 0.75 from 2008 through 2010. 


After accounting for off-balance sheet and shadow bank lending, we estimate that China’s outstanding domestic debt increased from 164% of GDP in 2008 to 207% in 2010.

We estimate about RMB9 trillion (US$1.4 trillion) of the RMB43 trillion increase in China capital stock from 2008 through 2010 was wasted on unproductive investments that will result in bad debts. 

These bad debts will choke off investment in the coming years, and there is no alternative source of growth that can provide an offset. The situation is coming to a head sooner rather than later, as local governments are on the hook for RMB900 billion (US$140 billion) in interest payments this year alone, and their off-balance sheet revenue streams are under pressure.

We sketch three scenarios of how this is likely to play out: A “muddle through” scenario in which the government offers a debt exchange and pushes through financial sector reforms; a “slow grind” scenario in which the government offers a partial bailout for existing debts, but all stakeholders share the burden of adjustment; and a “crash and burn” scenario in which high inflation limits the government’s policy response and it cannot prevent or delay a financial crisis.



Over the past decade, 60% of China’s growth has come from investment and net exports, as a rapidly expanding physical capital stock has helped to improve its export competitiveness. 


This was never going to be sustainable in the long run—shifting demographics would have started to reverse the repression of household incomes, which would have bid up the artificially cheap cost of capital making investment more expensive and ensuring that the RMB would drift upward toward fair value—but it was the mania of the 2008-11 investment boom that pushed the Chinese growth model into the abyss.

As Nouriel Roubini recently argued, this investment boom is leading to massive overcapacities that will generate colossal nonperforming loans (NPLs), risking a hard landing sometime after 2013. 


Here, we quantify the problem, sum up the likely costs and provide three distinct scenarios for the endgame.

GETTING A HANDLE ON THE SITUATION

The Stock Problem

China is now the second-largest economy in the world and, based on our calculations, it is the third-richest country in terms of physical capital, or the accumulated value of fixed assets. 


We base our estimates on an update of Nehru and Dhareshwar (1993), which calculated the stocks of 92 emerging and developed markets from 1950 through 1990.

 Using their estimates of each country’s initial capital stock in the early 1950s, we employ the perpetual inventory method to estimate capital stocks in 2010. In this calculation, the flow of new investment each year is added to the existing capital stock, which depreciates over time. 


We hold their assumption that existing capital stocks depreciate at a uniform 4% rate each year across every country, even though this is certainly a small error that compounds over time.

While China ranks near the top of our league table for total capital stock, it comes close to bottom on a per-capita basis. This suggests that it has years of capital accumulation ahead of it, and that China’s infrastructure is not over-built in the aggregate. In the long run, this is certainly true. 


However, in the near term, the infrastructure that China has built is not adding to output fast enough to pay for itself, and, with an investment rate of nearly 50% of GDP, this is no small problem.



Figure 1: China Is a Rich Country, Full of Poor People

Source: Nehru and Dhareshwar (1993), IMF, World Bank, National Statistics, RGE




The Flow Problem

While there are some problems with China’s stock of fixed assets, the real problem has to do with its flow of investment. 


Any country that builds a new highway is going to see a jump in GDP (gross capital formation) and an increase in productivity (more efficient movement of goods and labor). 


Simultaneously, building a railway of equal value parallel to that highway would double the boost to GDP, but the productivity gain from each project would necessarily be less, which would weigh on growth in the longer run. 


This is the situation in which China finds itself today—a rising flow of investment, with a lower return on its capital.



Figure 2: The Flow Investment Overwhelms Consumption (% of GDP)

Source: National Bureau of Statistics




When demand for China’s exports collapsed in 2008, policy makers in Beijing needed an alternative source of growth and so they opened the flood gates for local leaders to build nearly anything they wanted, so long as it created jobs and GDP. 


This led to a massive investment in high-speed railways, highways, airports, luxury apartments, wind power and heavy industrial capacity. 


While in the aggregate China still has a deficit of infrastructure, the speed in which these new projects are coming on line negates the usefulness of each of them. There was no time for a real assessment of which new highways were needed, let alone how those roads would be affected by the new high-speed rail line planned along the same route. Instead, anything that was on a local leader’s wish list was given instant approval. 


The reason every leader had such a long wish list was because the National Development and Reform Commission had previously judged most proposals as lacking in any economic justification. However, even if these plans had been sound, production simply cannot be shifted around fast enough to keep up with the pace of China’s current investment rate. 


The end product of poor planning and rapid investment is clear: There are massive traffic jams going into Beijing while the city is ringed with empty luxury apartments and villas. 


Wind-powered generators sit unconnected to the state grid, while new coal-fired plants are built down the road. High-speed railways run nearly empty, while coal shipments overwhelm the existing freight network. The result has been a collapse in the marginal product of China’s capital, or the amount of GDP each dollar of investment generates, a problem that will compound as long as policy makers cannot find alternative sources of demand.



Figure 3: Investing for GDP, Not Profit (RMB, trillions)

Source: Nehru and Dhareshwar (1993), IMF, World Bank, National Bureau of Statistics, RGE



A decline in the productivity of capital is a warning sign of problems to come, as it implies investment growth is not sustainable at its current rate. 


China’s marginal product of capital fell from an average of 1.0 in the first seven years of the decade to 0.75 from 2008 through 2010. Given our growth forecasts for 2011-12, we expect China’s marginal product of capital to remain close to 0.7 this year and next. 


South Korea and Malaysia saw similar declines in their capital productivity in the five years prior to the 1997 Asian financial crisis. 


Brazil’s ratio fell sharply in 1996, two years before its currency crisis. The marginal productivity of capital in Spain declined sharply in 2002 and stayed low until the bottom fell out in 2008. 


In the U.S., the ratio declined steadily from 2004 through the crash in 2008. Although China does not run a current account deficit or have an overvalued currency, it too will run into a wall as long as it continues to pour funds into unproductive investments.



The Debt Problem



The weakness in China’s growth model, as with all of those above, is that the vast majority of these unproductive investments were financed with debt. 


We estimate that China’s domestic debt-to-GDP ratio increased from 164% in 2008 to 207% in 2010, as total debt outstanding increased by 60% to RMB82 trillion (US$12.4 trillion). 


A sharp increase in debt-to-GDP is another warning sign of coming problems. 


In Hyman Minsky’s financial instability theory, an increase in debt without a similar rise in output indicates that debt is drifting from “hedge” borrowing, in which cash flow can cover debt obligations, to “speculative” borrowing, in which cash flow can only cover interest payments or “Ponzi” financing, which requires ever rising asset prices to cover even interest payments. 


That much of the increase in 2010 came from off-balance sheet lending, similar to the special purpose vehicles we saw during the U.S. subprime bubble, compounds the risk. It seems that loans that could not be repaid have been rolled over into debts that regulators cannot easily find. 


If banks find themselves unable to keep rolling over these obligations, NPLs would quickly eat through the state-owned banking sector’s provisions.



Figure 4: China’s Growing Debt Burden (RMB, trillions)

Source: People’s Bank of China, Ministry of Finance, Asian Development Bank, RGE estimates



A decline in China’s marginal product of capital was inevitable in 2008, given the output loss from exports, but by maintaining its reliance on investment for growth China has been throwing good money after bad. 


We estimate China “wasted” about RMB9 trillion in investments from 2008 through 2010, based on the decline in its marginal productivity of capital, which equals roughly 22% of the increase in credit over the same time frame. 


In an optimistic scenario, not all of this will end in default, but, given that the government has already admitted that only 27% of local government investments can generate sufficient cash flow to cover their debt obligations, it is probably not too far from the mark. 



In a less benign scenario, even good investments could go bust if the banking sector chokes on bad debt and demand falls sharply. 


There is no reason to believe that RMB9 trillion in defaults and NPLs is as bad as it can get. 


China has continued to fund wasteful investment projects this year in order to maintain relatively strong headline growth, and it will continue to do so next year given the lack of external demand. It is unlikely to be able to keep playing this game much beyond that point.



It’s All the Government’s Problem



We previously estimated in the July China Monthly that the contingent liabilities of China’s national government stood at 77% of GDP (RMB30 trillion) at the end of 2010, compared with the 17% that it recognizes on its balance sheet. 


This means that, although the Ministry of Finance’s balance sheet can absorb the bad debts that we expect to start piling up next year, it does not have enough room to do this and force out more government-led investments at the same time. 


The government’s fiscal position is much more constrained than it was before the global financial crisis, when we estimate that its contingent liabilities amounted to roughly 54% of GDP. 


Additionally, in every scenario we present below, growth slows significantly. Without high nominal GDP growth, China’s government at all levels will have to constrain its borrowing to stabilize its debt-to-GDP ratio.

Figure 5: Not Much Room for the Government’s Balance Sheet to Grow (RMB, billions)

Source: Ministry of Finance, Ministry of Railways, People’s Bank of China, National Audit Office, Asian Development Bank, RGE estimates



Beijing almost certainly will bailout the state-owned banking sector and before the end of 2013, but it is not clear how much it is willing to offer.


 About 53% of the bank loans issued to fund the investment boom will come due between 2011 and 2013, assuming the maturity profile on local government debts roughly matches those of developers and state-owned enterprises (SOEs), which have also borrowed heavily to build physical capital. Banks have so far been evergreening the losses on these debts, largely through off-balance sheet conduits, but higher interest rates and regulatory changes will narrow the scope for banks to keep rolling over these bad debts. 



China’s banks are currently sitting on about RMB560 billion in provisions above their existing NPLs. 


This will prove helpful for delaying the crisis, but will not be enough for the banking sector to absorb the coming hit beyond next year without government support.



We think that Beijing is more likely to offer only a partial bailout because, so far, it has been reluctant to recognize all of its contingent liabilities. 


The National Audit Office reported that China’s local governments had debts of RMB10.7 trillion at the end of 2010, about RMB4.3 trillion less than we estimate and RMB3.5 trillion less than Moody’s estimates that local governments borrowed from the banks. 



It seems unlikely that Beijing will assume debts that it does not officially recognize as government obligations, making those loans extremely risky, whether on- or off-balance sheet. 


Likewise, the Ministry of Railways is very nearly bankrupt and will never be able to meet its debt obligations with its own cash flow, yet it continues to issue corporate bonds that lack an explicit guarantee from the Ministry of Finance. 


Additionally, developers with close ties to local governments have shifted a large portion of their debt financing offshore and underground, and Beijing may not be willing to bailout foreign investors or underground financers.



And It Ends With a Growth Problem



Whether Beijing offers a full or partial bailout of the financial sector, Chinese growth is set to slow sharply in the coming years. 


Deleveraging will be inevitable for local governments and state-linked borrowers that have funded unproductive investment projects with substantial amounts of debt. 


The strength of the central government’s balance sheet will allow it to assume some of this burden, and the high savings ratio of China’s households will ensure there is enough liquidity to prevent a run on the state-owned banks. 


However, the central government will not be able to lever up fast enough to offer a sufficient bailout and maintain high investment rates at the same time, and households will not be able to offset the slowdown in investment with higher levels of consumption.

Mechanically, with private consumption having fallen to 34% of GDP, it would take a massive surge in consumption to offset the slowdown in investment growth. 


Consumption growth would have to increase 1.5 percentage points for every percentage point slowdown in investment if China is to maintain the same growth rate, holding everything else equal. 


Chinese households save a bit more than 30% of their income, roughly on par with their Confucian neighbors, but private consumption has fallen to 34% of GDP because household incomes have failed to keep up with the growth of national income. 



Instead, income has flowed to state-owned enterprises in the form of retained earnings, which they have plowed back into fixed investments. 


The new Five-Year Plan calls for SOEs to pay higher dividends, which could be used to fund fiscal transfers or social services for households, but the increase is too small to make much a difference. Given the need for SOEs to pay down some of their debts, they will need to maintain a rather high savings rate in the coming years, and forcing them to pay out more than this may prove counterproductive. 



Wages are beginning to outstrip productivity, which will start to rebalance China’s growth model, but this is also moving too slow to make much of a difference in the short term. This year’s wage gains look to be roughly on par with last year’s, when private consumption grew a paltry 6.6%, according to our estimates. 


Wages may increase 12-15% this year, but even at that pace they will fail to keep pace with per-capita nominal GDP once again, further pushing down labor’s share of national income. Additionally, the government will not be able to force through the financial sector reforms that would be necessary to boost the return to savers (households) and raise the costs of borrowers (SOEs), while the banking sector is in crisis.



 For these reasons, and those RGE Chairman Nouriel Roubini recently outlined, Chinese private consumption is unlikely to suddenly surge while the country’s growth model collapses.



Looking beyond China’s unique domestic factors, it is clear that economic history is not on Beijing’s side. 



In a sample of 24 countries that have seen a structural peak to their investment share of GDP, growth slowed from an average 5.5% in the three years preceding the peak to 0.9% in the three years following.


 There were no exceptions—growth slowed in all 24 countries, though some countries performed much better than average.



Figure 6: Nobody Escapes a Growth Slowdown When Investment Peaks

Source: IMF, RGE calculations



While a growth slowdown is inevitable, a hard landing is not. 



Other Asian economies (South Korea in 1991, Vietnam in 2007) have managed the transition with only a modest slowdown in growth, at least in the short term, and there are reasons why China, too, might escape the worst of the coming crunch. 



First, liquidity is unlikely to be a problem. With the required reserve ratio (RRR) at a record-high 21.5% for the largest commercial banks, there is about RMB16 trillion worth of liquidity sitting on the balance sheet of the People’s Bank of China (PBoC) that could be used in a crisis (likely requiring the sale of China’s foreign assets). 


Additionally, financial repression and a closed capital account should prevent a run on the banking sector. 



Second, the central government’s clean balance sheet will allow it to easily assume its contingent liabilities without too much strain, while the strong asset position of local governments could potentially allow them to repay their bad debts with good assets. 


Perhaps most importantly, it is not too late for Beijing to get ahead of the problem.



THE ENDGAME: THREE WAYS TO KILL A GROWTH MODEL



In all of our scenarios, Chinese growth slows sharply and the government is forced to deal with its contingent liabilities. The situation is coming to a head sooner rather than later, as local governments are on the hook for RMB900 billion (US$140 billion) in interest payments this year alone, assuming a 6% rate on RMB15 billion in debts. 


Since on-budget revenues have been allocated elsewhere, the shortfall on debts for non-income generating investments will have to come primarily from land-lease sales. 


We estimate that land sales are likely to generate RMB1 trillion in revenue this year—barely enough to cover their existing off-balance sheet obligations and much of this revenue has already been pledged to cover other commitments, like the government’s social housing construction efforts. 


Next year, land sale revenues are likely to decline, while local governments’ debt service costs will rise, assuming a large portion of their debts due this year have already been rolled over. Some revenue and expenditure juggling, combined with pressure on local banks, might get local governments through 2012 without major defaults. Much beyond this, however, and the situation will become untenable.



Our starting assumption is our baseline global scenario, which sees growth in developed markets (DMs) tipping into a stall speed while emerging market (EM) growth is only slightly below potential in H2 2011. 


Our global scenario analysis assigns a 40% probability to DMs falling into a recession in 2012, though the baseline scenario (60%) is that growth remains volatile but DMs narrowly avoid a recession and EMs continue to grow slightly below potential. For our China scenario analysis, we assume the baseline global forecast for 2012. 


If DMs were to tip into a recession in 2012, it would likely lengthen our timeline, but increase the risk of our crash scenario.


 A DM recession could push back the reckoning of China’s investment imbalance if it allowed the PBoC to loosen monetary conditions in 2012, which would lower the costs for floating rate borrowers and allow for easier roll-overs. 


However, this would increase the size of China’s debts and ensure that its marginal product of capital fell further as output growth would slip even with another fiscal stimulus (which would necessarily have to be smaller than in 2008, given the deterioration of local government and bank balance sheets). 


The eventual cost would be higher, even if it took slightly longer to get there.



The two major turning points in our scenarios are the near-term inflation outlook and the size and shape of the expected government bailout that will come in 2012 or 2013. 


We assign a high probability to Chinese inflation easing to about 4.2% in 2012 as commodity inflation eases and monetary policy stays on hold. Lower inflation will take some pressure off banks and borrowers, as well as providing more fiscal space for the government to bail them out. 


However, as the Fed prepares to launch another round of quantitative easing, and with few signs of a significant deceleration in Chinese growth, there is a meaningful risk that inflation will overshoot our forecast. This would pull forward the coming crisis, possibly to as early as mid-2012, during which China’s political transition will be at its peak and policy paralysis will be severe. 


Higher inflation is a necessary but insufficient condition for our “crash and burn” scenario.



Likewise, the size and shape of a government bailout will have a significant impact on Chinese growth in the medium term. As argued above, we expect the government to offer only a partial bailout of local governments, banks, SOEs and developers in 2012 or early 2013, though it will be nearly impossible to avoid taking over the Ministry of Railways’ debts in full. 



A partial bailout would be somewhere in the realm of RMB4 trillion-6 trillion, and would include a debt exchange for local governments, an injection of capital into the state-owned banks and the complete absorption of the railways balance sheet. 



A full bailout would be something closer to RMB10 trillion. Its component parts would remain the same, but a larger share of local government obligations would qualify for the debt exchange. 


Both options would see local governments enter the bond market directly by the end of 2013, where they would have to pay something closer to a market rate to borrow and fund new investments. 


A partial bailout would force serious deleveraging on local governments in order to tap the new municipal bond market, exacerbating the slowdown in investment after 2012.

We assume the following delinquency rates on debts: 30% for local governments (RMB15 trillion outstanding at end-2010), 30% for property developers (about RMB4.2 trillion domestically), 50% for the Ministry of Railways (RMB1.9 trillion) and 7.5% for SOEs (roughly RMB30 trillion). The problems of the commercial and policy banks are largely tied to these debts, and so we leave them out to avoid double counting. 



We do not expect severe losses on mortgages or consumer loans, as falling property prices are likely to be absorbed by the owners’ equity due to relatively high down-payment requirements. This may prove overly optimistic, however, if debts from outside of the banking sector were used to fund down-payments, which anecdotally appears to be quite common. 



As such, China’s national balance sheet is set to take about a RMB9 trillion hit sometime between 2012 and 2015. 


If inflation cannot be tamed in the near term, default rates would rise, along with the ultimate cost. 


This would probably result in another RMB2.1 trillion or so of problem loans, mostly from developers and local governments, which are more interest rate sensitive. Below, we present the three most likely scenarios for 2012-16, leaving aside the combination of high near-term inflation and a full bailout, since the contingent probability of that scenario is rather low and the end result would be similar to our “slow grind” scenario.



Figure 7: A Road Map for the End of Roads

Source: RGE





Labels: , , ,