China Deleveraging Revisited
Probably the biggest question facing China and international investors is whether China is deleveraging. The Chinese debt and fixed income markets have changed though to the point that people can see almost anything they want. You choose to see China is deleveraging or is still leveraging up, and you can make a case for that. Let us go through the data so we can see how I reach my conclusions, but then how data is being obfuscated, omitted, or not considered.
- Total bank loans in 2017 was up 12.7% with nominal GDP up 11.2% representing a small boost in total leverage. Stock aggregate financing to the real economy was a bit lower than bank loans at 12% but still representing a small additional leveraging to the debt to GDP ratio. In fairness, this aggregate number is solid.
- What is noticeable about the Chinese debt numbers is how the sub-categories diverged so significantly. Put another way, how China arrived at those aggregate numbers is interesting and instructive. In short, some (not as much as some people argue) but some financial deleveraging took place, some corporate deleveraging took place but households and government blew up their debt levels. What is notable is that the “deleveraged” sectors did not see absolute falls in debt, they just saw slower growth rates below nominal growth so they remain decidedly weak. The government and household sectors appear to be rapidly looking to match the indebted corporate sector so that the enormity of the corporate debt load is matched across all sectors of the Chinese economy. This matters at it makes it that much more difficult for the government or household sector to play a larger role if so needed. We are not at those levels yet, but rapidly moving towards them.
- In the corporate and financial sectors, there actually was progress made on deleveraging. Claims on non-financial institutions in 2017 rose a paltry 6.3% and are up only 32% since 2014. The current corporate debt sector problems are due more to the explosion of debt between say 2009-2014 (rough divisions). Debt growth has slowed but given that the slowing is let’s say roughly in line with or moderately beneath nominal GDP growth in any one year, the corporate sector debt level has remained elevated in absolute terms.
- It is important to note, in my opinion, that 2017 was likely an outlier in terms of debt growth and industrial profitability. The story of 2017 in China was entirely and completely the growth in basic input commodity prices like coal and steel but more broadly related inputs, primarily metals like tin and aluminum. To put this in perspective, if we remove the impact of metals processing from manufacturing profits, profit growth drops from 6.6% to 3.9%. Put another way, 42% of the growth in manufacturing profits can be attributed to areas like ferrous metals. Add to this the fact that mining and related sector profit was up 151% and in reality the 2017 boost to profits and corporate deleveraging due primarily to a very narrow group of industries. On average, outside of this narrow group, profit growth was quite low. Liabilities in mining and ferrous metal smelting actually declined in absolute terms but grew slightly faster than revenue in other industrial sectors. I say all this because “deleveraging” and industrial profit growth in 2017 China was driven almost exclusively by financial market driven commodity price gains engineered by Beijing. For so many reasons, expecting triple digit price gains to drive the economy is risky at best. 2017 was not a broad based economic recovery in China. On the downside, these numbers are very unlikely to go up definitely nowhere near previous rates. On the upside, Beijing seems intent on establishing a price floor to keep these firms profitable.
- What is notable is where the debt growth is occurring. Government and households have seen debt grow rapidly. Even more noteworthy is that while this has gained increasing attention, this is not simply a 2017 phenomenon but a multiyear explosion in debt growth by the government and household sectors. Bank claims on government grew 27% in 2017 but grew 98% since 2015 and 207% since 2014. A similar phenomenon is playing out in the household sector. Total official household debt has grown 21%, 50%, and 75% in aggregate over the last three years. What is important to note is that while the non-financial corporate sector remains the largest individual line item for banks, it is no longer as dominant a line item as it once was. The other bank line items like financial lending, investment, households, and government are now larger that straight corporate lending.
- This paints a larger story that China is attempting to reign in corporate debt levels, mildly albeit, but to continue to hit growth targets they are using various fiscal stimulus and household debt to prop up the economy. This also paints a story of continued additions of net leverage which I would expect to continue to rise over the next 1-3 years for reasons which I will get into later. One of the enduring stories in recent Chinese economic history is the never ending assurances that they do not worry about growth targets and can deleverage anytime but they never seem ready to accept the inevitable tradeoff between reducing all debt growth and lowering GDP.
- There has been a decline in financial leverage but here I think this change has been overstated. Much of the financial deleveraging that has taken place has simply moved into other instruments. Consequently, while financial to financial lending is effectively flat, banks have increased their holdings of “portfolio investment” domestically by a healthy 9.3%. Bank to bank WMPs decreased but this was largely offset by higher retail sales of WMPs. The primary point here is that while there has been a slowing of growth in financial leverage, let’s not overstate the change.
- Within this there are a number of big picture moves playing out that I do not think are getting the necessary attention in how all of this fits together. First, because 2017 is such an outlier due to the price boost of commodities (which is a statistical issue worthy of exploration by itself) we would be mistaken to just carry that forward in how we view the range of variables we look at. In reality, as you will see, I think it much more accurate to view 2017 as the year when China did not address the problems facing it in the good times preparing the expected slow down. Second, PPI finished 2017 at 6.3% and it is quite likely to finish at least 2% lower. That means that nominal GDP should finish distinctly lower meaning leverage will likely increase more in 2018. China got a nice, likely one time, inflation bump in 2017, but the prices right not do not indicate a similar bump is likely to come in 2018. Third, the real concern is that corporate debt levels remain very worrisome even if they have not worsened and now the debt levels of other sectors is rising offsetting the ability to address any shocks. They word I have chosen to use it increases the “fragility” of the system. Household and government sector are rapidly consuming their slack to assist the corporate sector should we need a recapitalization.
- Then I think there numerous specific mechanisms that have not gotten enough attention. First, financial deleveraging and base commodity prices that drove the nominal GDP rebound are intimately tied up. If China is going to attack financial leverage, this is going to put significant pressure on commodity prices. Though many believe that prices have gone up due to capacity cuts, this simply is not true. Total capacity in virtually all metals has continued to increase in absolute terms. It seems highly unlikely China will be able to keep commodity prices buoyant andhave significant financial deleveraging. These two and too closely linked.
- The problem with 2017 is not what happened but what did not happen. For instance, in a good economic year, NPLs did not decline. We know they will go up if the economy slows and yet no progress was made in 2017. Actual debt to equity swaps executed were a small fraction of the announced number and bankruptcies by any standard remained quite low. Given that nominal GDP is likely to fall by at least 2% in 2018, it seems like a missed opportunity to accomplish little in the way of actual clean up.
- This matters because I believe a mountain of evidence indicates the underlying foundation of the Chinese financial system continues to erode. For instance, in 2017 banking new loan to new deposit ratio topped 100% and since 2014 this number has totaled 90.2%. This implies that the lending binge from 2009-2014 is not returning the necessary cash to pay back the debts and increase the deposit base to make new loans because there are large ongoing capital write offs. If we add in the additional investments banks have made, this number far exceeds 100%. This is not the sign of a healthy banking system.
- This is showing up in the unsexy assets a bank holds. While total banking system assets in China have grown about 45% in the past three years, this has come at the expense of cash, central bank deposits, and reserves. Over the past three years, as bank assets grew 45%, bank cash, deposits with the PBOC, and reserves grew a total of 9.7%. This pushed down CDR (cash, deposits, and reserves) from 27.1% of assets in December 2014 to 20.5% in December 2017. The last time this number was this bad was in 2006 at the tail end of the bank restructuring. A drop of almost 7% in three years is significant and must slow before stopping. That number will likely continue to decline weakening the ability of the Chinese banking system to manage shocks.
- However, it is not just banks working to prop up the Chinese economy but the PBOC as well but not necessarily how you think. In 2017, PBOC depository corporation and “other” assets grew 31% and 187% from December 2015. Excluding “foreign assets” such as exchange reserve, these two line items comprise nearly the remaining one third of PBOC assets so this is not starting from a small base. To put the absolute size in perspective, this amounts to roughly 15% of all lending in China. In 2017, total banking system assets increased 19.3 trillion RMB and PBOC assets grew 2.8 trillion RMB. There are couple of important implications. First, financial deleveraging was replaced by PBOC asset purchases not systemic reductions. Second, what is notable is that the PBOC is not classifying these as government claims which seems puzzling given their previous declarations they would accept local government bonds on repo. However, for the moment let us assume these are at most LGFV debt. Third, it clearly belies Beijing claims about desiring deleveraging as this was a clear policy channel that is being used to boost credit and meet growth targets.
- I don’t think we have strong evidence that Beijing is producing fraudulent systemic bank data the same they do for other areas of the economy and financial markets though they are clearly working to hit headline numbers while letting the underlying numbers continue to erode. I think you have to be quiet suspicious of financial data in China so I am not about to defend its accuracy, but what is more readily apparent is that China is clearly focused on achieving the top line objective for targets like to debt to GDP even as loan to deposit ratio goes on rising and CDR to assets continues to fall.
- There are two overarching economic problems I see. First, Chinese assets are wildly overvalued. This stems in my opinion from a money supply that is vastly in excess of the size of the economy. The differences are enormous. In nominal exchange rate terms, the US economy is nearly 50% larger than the Chinese economy. However, Chinese RMB M2 is effectively twice as much as the USD for what is effectively the global currency. When put against the size of the economy, the numbers are even more striking. There is more than twice as much Chinese RMB as there is economy. Conversely, there are actually 30% few USD than there is against the size of the economy. This of course translates into enormously higher asset prices. This translates into lending against overvalued assets which is why the PBOC is so worried about keeping asset prices high whether it is commodities or real estate. Beijing cannot seriously delever without accepting significant pain from lower asset prices.
- The concept that remains with me is this idea of improved headline numbers but continued hollowing out and increased fragility. Corporate debt relative to assets and revenue has stabilized at very high levels but now debt is growing rapidly in other sectors. This makes the future increasingly fragile and vulnerable to shocks. I think headline numbers have not necessarily improved but they are worsening at a slower rate. However, many of the underlying numbers continue to worsen significantly increasing the fragility of the system and ability to respond.