Whether Janet Yellen admits it or not, you can bet that going into the most important Fed meeting ever (until the next one) the supposedly “data dependent” FOMC has taken a good hard look at what’s happening in China in the wake of Beijing’s not-so-smooth transition to a new currency regime. 

We’ve detailed the story exhaustively, so we won’t endeavor to recap it all here, but the short version is that what was billed as a move to give the market a greater role in setting the yuan’s exchange rate actually had the opposite effect - at least in the short run. That is, the PBoC used to manipulate the fix to control the spot and now they simply manipulate the spot to control the fix, but unabated devaluation pressure has forced China to intervene on a massive scale and that intervention recently moved into the offshore market as well, as Beijing scrambled to close the onshore/offshore spread. This is costing China dearly in terms of FX reserves, the liquidation of which was so massive in August as to prompt Deutsche Bank to brand it “Quantitative Tightening”, as the reserve drawdowns are effectively QE in reverse. This is of course the same dynamic that’s been taking place in Saudi Arabia in the wake of the petrodollar’s demise and mirrors the response across EMs which are struggling to support commodity currencies as prices collapse.

Attempts to quantify the scope of China’s reserve burn have become ubiquitous, as the cost of offsetting the outflows from China effectively serves as a proxy for the extent to which the Fed would, were they to hike, be “tightening into a tightening”, as we’ve put it. 

On Wednesday we showed that Beijing liquidated $83 billion in Treasurys in July. That, as we also noted, “is before China announced its devaluation on August 11 and before, as we also first reported, it sold another $100 billion in Treasurys in August.” 

Today, we get a fresh look at the numbers courtesy of SAFE which shows that on net, banks sold $128 billion in FX to Chinese non-banks in August. Nothing too surprising there, given that we already knew positioning for FX purchases for the banking sector as a whole dropped by $115 billion for the month. 

As Goldman notes however, when you include banks’ forward books the picture worsens materially:

An alternative gauge that we believe is a closer reflection of the underlying trend of currency demand shows a significantly larger outflow of $178bn.

Today's data at US$178bn on our preferred gauge of underlying currency demand (i.e., outright spot plus freshly-entered forward contracts) is significantly higher than any of [the previous] releases. 

This means, as Goldman goes on to point out, that outflows are actually far worse than what’s indicated by simply looking at China’s reserve drawdown, as banks look to be shouldering some of the burden themselves:

We think this suggests that banks used their own FX position (i.e., without resorting to PBOC's FX reserves) to absorb part of the outflow pressure.

So between July and August (inclusive of freshly entered forwards), outflows total around $261 billion. 

But that’s not all. 

Nomura is out with an estimate of what’s taken place since the start of September.

Between onshore spot intervention and offshore spot and forward intervention, the bank estimates China has spent some $47 billion month-to-date stabilizing the yuan, $25 billion of which in the offshore market, reinforcing what we said a week ago after CNH soared the most on record: 

Note the rationale here: this is China intervening in the hopes that said intervention will make further intervention unnecessary. That is, rampant speculation that the yuan will continue to depreciate is forcing the PBoC to intervene in the onshore market and at an extremely high cost both in terms of the country's FX reserves and in terms of the deleterious effect the reserve liquidation has on liquidity. Devaluation expectations are at least partly manifesting themselves in the offshore spot market so ultimately, the PBoC figures it will try intervening there in the hopes that narrowing the spread will mean it has to intervene less in the onshore market. Summarizing the above in four words: one more spinning plate.

Here's what $25 billion in CNH intervention buys you:

What all of the above means is that all in, outflows totaled some $308 billion in the space of two and a half months.

Some speculate that between stepped up capital controls and a Fed hike, the situation may stabilize (there are those who think FOMC liftoff will serve to calm the market by removing uncertainty), but consider the following from Commerzbank's Zhou Hao:

Increasing FX intervention could be one of few options left. Some companies and real-money funds could look to purchase dollars by selling offshore yuan, so China may extend defense offshore. Frequent intervention burns foreign reserves and squeezes onshore market liquidity, so further tightening of regulations likely if CNY still under pressure.

Along with this from Maybank's Fiona Lim:

PBOC may intend to narrow CNY-CNH gap to ease depreciation pressure; it’s likely to buy offshore yuan and allowing onshore to move lower via daily fixing.

In other words, even if you're optimistic that the pressure to intervene is abating (as Lim actually seems to be), the above underscores the extent to which this is a very, very delicate balancing act, or, as we've characterized it over and over, an attempt to manage a collection of spinning plates. 

And then there are those who take the straightforward view (i.e. the view that doesn't include the assumption that a Fed hike would somehow stabilize things be decreasing uncertainty), like Citic Bank chief economist Liao Qun who said the following on Thursday (via Bloomberg):

U.S. rate hike should have negative effects on both the Hong Kong property and stock markets and intensify pressure on the yuan and increase the likelihood of further depreciation. 

There you have it. In the end, the takeaway is the number shown above: $308 billion in FX outflows in two and a half months. 

Your move Janet.