Back in May, this website was the first to explain the "mystery" behind Belgium's ravenous Treasury buying which in early 2015 had turned into sudden selling, and which we demonstrated was merely China transacting using offshore Euroclear-based accounts to preserve anonymity. Since then theme of Belgium as a Chinese proxy has become so popular, even CNBC gets it.
Consequently, we were also the first to correctly warn that China had begun liquidating its Treasury holdings (a finding which left none other than Goldman "speechless"), which also helped us predict that China is about to announce its currency devaluation three days before it happened as the conversion of Chinese reserves from inert paper to active dollars hinted at a massive effort to stabilize the currency, and thus unprecedented capital outflows.
As a result, the only data point which mattered in yesterday's Treasury International Capital data release was not China's holdings, which actually "rose" $1.7 billion in the month when China actively devalued its currency and then spent hundreds of billions to prevent the devaluation from becoming an all out FX rout, but the ongoing decline in Belgium holdings. As the chart below shows, Belgium, pardon Euroclear - which is a clearing house not only for China but many other EM nations who park their reserves in Belgium - sold another $45 billion in Treasurys last month, bringing the total to a dangerously low $111 billion, down from $355 billion at the start of the year.
Lumping Belgium and China holdings into one, as we have done since May, shows that as expected, Chinese selling continued in August, and the result was another drop of $43 billion in TSY holdings in the month of August, which incidentally mirrors perfectly the previously announced decline in September Chinese FX reserves, which according to official data declined from $3.557 trillion to $3.514 trillion.
According to the chart above, while to many Quantitative Tightening is a novel concept, the reality is that China (+ Euroclear) have been dumping Treasurys and liquidating reserves since January when total holdings peaked at $1.6 trillion last summer, and have since declined to $1.38 trillion. It means that China has sold a quarter trillion dollars worth of Treasurys in the past year, in the process offsetting what would have been about 25% of the Fed's QE3.
However, the real number is likely far greater.
While China's official (declining) FX reserve data (a real-time mirror of the TIC data from China's perspective) is a useful guide to what is happening with China's reserves (primarily US Treasurys, as well as European sovereigns and various other unclassified assets), it is also manipulated by the politburo which does not want to give an overly pessimistic picture of what is happening in China. As a result, a far more accurate representation of FX flows comes from the data showing FX purchases for the whole banking system (PBOC plus banks), as this number is far more difficult to rig.
As we previously reported, in September FX purchases decreased by US$120 billion in September (vs. a decrease of $115bn in August). This is a troublin discrepancy with both official Chinese reserve data and US TIC data as the scale of decline in this data is significantly larger than that in PBOC’s FX reserves (-$43bn) and its foreign assets (-$42bn), suggesting that banks have resorted to their own spot FX positions to help absorb outflow pressure.
More from Goldman:
Given possible PBOC balance sheet management (e.g., short-term transactions and agreements between with banks, e.g., forward transactions, FX entrusted loan drawdown or repayment), we interpret the FX reserves data with caution, as it might not give a complete picture of the FX flow situation. The large gap between today’s data and the other PBOC data for September suggests that banks might have used their own spot FX positions to help meet some of the outflow demand, although banks’ overall FX positions might still have been squared with the PBOC via forward agreements. This also partly explains why new RMB loans exceeded RMB deposit generation by a large margin (of over RMB 1tn) in September, as shown in yesterday’s money and credit data--apparently corporates and households converted a large amount of their RMB deposits into FX. Overall, today’s data indicates that the outflow situation might have improved only modestly from August. Note that today’s data do not include information on possible forward transactions between the PBOC and onshore or offshore banks.
The problem with China's data - and incrasingly the US - is that it is completely unreliable. So to get a full picture of what really hapepned, we will have to wait for SAFE data on banks’ FX settlements on behalf of their onshore clients (due on October 22nd) to have a more complete view on the FX-RMB conversion trend among onshore non-banks. That report captures banks' FX transactions vis-à-vis non-banks through both spot and forward transactions (for August this data showed an overall FX outflow of $178bn).
But even with the incomplete picture we have, we can draw two conclusions:
- Chinese FX outflows are actually accelerating, not slowing down, despite the PBOC's (and TIC's) best efforts to show that China has sold "only" $250 billion in Treasurys in the past year
- Chinese TSY selling has so far not impacted price of 10Ys adversely because it took place in a time of "great unrotation" from stocks into safety assets, and a surge in global deflation fears provoked by... China. Now that the Politburo appears to have fooled markets that China is "fine" once again, and inflationary fears reemerge, preserving the bid for 10Y paper may not be quite so easy especially as China continues to fight capital outflows by selling reserves.
- The recent data on rising Chinese credit creation had nothing to do with an improvement in the economy, and everything to do with covering the discrepancy between official (declining) capital outflows, and unofficial (increasing) capital outflows.
Bottom line: China's capital outflow is getting worse, not better, and continued to drag not only its economy lower, but accelerate China's disposition of Treasury paper. Anyone hoping for a quick rebound in China's economy will be severely disappointed.