The July update to the Treasury International Capital flows (TIC) was mostly the same as we have observed throughout this year. The private segment continues to buy on net while the official sector continues to sell on net. I think it fairly reasonable to conclude that the latter is purposefully designed so that the former can take place. In other words, global central banks (not just China) are attempting to intervene and calm “dollar” (funding) markets so that private entities (local banks and their “dollar short”) aren’t forced into hugely disruptive liquidations, able then to carry out the normal course of trade under the “dollar short.”

Total UST “selling” has picked up in the past few months in the official sector, offset somewhat in July by purchases (net) of agency paper. That means global central banks/governments continue to “sell” US$ assets, mostly UST’s, but at a rate somewhat less than at the start to this year.

abook-sept-2016-tic-official-6m

As a result, private buying of US$ assets (or at least what is recorded and reported via TIC) jumped to a net positive $90.5 billion in July. That was the highest net purchasing since June 2015. Rather than be reassuring, however, the comparison to last June may be more appropriate for the “top” that it represented.

To start with, something is “off” in the data for global banking. Ever since the taper summer in 2013, reported bank liabilities (titled as “change in banks’ own net dollar denominated liabilities”), a rough proxy for global “dollar” money dealing, has fallen into a well-defined quarterly pattern. The reason for the change remains unclear to this day, but for whatever reason banks have been consistently drawing down their dollar-denominated liabilities at the end of each quarter – only to scale them back up to start the next. This is surely related to window dressing, but the degree to which it takes place, meaning the difference between the past quarter drawdown and the initial reflow to start the new one, can tell us something about funding markets in that quarter.

abook-sept-2016-tic-bank-liab-qtrly-pattern abook-sept-2016-tic-bank-liab-6m

As has often been the case, the middle month of each quarter ends up making the difference. One of the more ominous signs (though reported months afterward) at the end of 2015 was the net negative for November. The middle month of each quarter has typically been slightly positive with a few small negatives usually presaging the worst quarterly conditions. November 2015, for instance, showed -$83.4 billion, indicating pretty stark “dollar” difficulties within the global banking sector just prior to the second and (so far) largest global liquidation.

The bank data for July 2016, however, is concerning in terms of the lack of positive flow after quarter end. Banks finished Q2 2016 already at a slightly more negative pace, -$171 billion compared to -$163 billion in March 2016, and -$161 billion September 2015 at the end of that Q3. To start Q2 this year, as has been normal, banks’ dollar liabilities in April rose by +$188 billion more than offsetting the end of Q1 (March), which was also far better than the +$134 billion in October 2015 and thus a good reflection upon global “dollar” conditions at that time. To start Q3, however, only +$47.6 billion was reported for July – a huge discrepancy.

abook-sept-2016-tic-bank-liab-qtr-beg-month abook-sept-2016-tic-bank-liab-qtr-end-month

That is the lowest quarterly starting “flow” going all the way back to the last QE’s almost four years ago. And unlike the end of 2014 and the start of 2015, it is a sharp break in the bank funding/money pattern as compared to the start of Q2 in April (which was unsurprising given the rebound state of the world immediately after February 11).

This is, of course, by no means dispositive; it only adds evidence to what I have suspected for the past few months – that “something” changed in overall “dollar” conditions especially going back to July. Unfortunately, by the nature of monthly data we have no idea in what part of July this “something” might have occurred, especially given the events toward the end of that month relating to Japan and the Bank of Japan. I don’t believe it is coincidence to see this unusually low number in bank liabilities and then find swap spreads at the long end, for example, spend most of the month of August dropping (compressing more negative) yet again.

abook-sept-2016-breaks-swaps-30s-spreads

A more negative swap spread is the same thing as being reported here in TIC with unusually low positive net “flow” in foreign bank “dollars” – less offered or usable balance sheet capacity. It does seem to confirm in its limited, rough capacity what I wrote this morning in regard to China money and “dollars”:

Such drastic changes and blown expectations immediately call our attention to the “dollar.” Why has the PBOC sacrificed its post-August 2015 liquidity regime to stick CNY at around 6.68 to 6.685 – not even 6.70 anymore? The fact that Chinese money markets are now such a huge and obvious mess can only mean that whatever it is about the “dollar” that caused this shift has been judged as the greater of these two “evils.”

As to what specifically caused this renewed disruption I can only guess and speculate (and so I will refrain from doing so). For now, we have to be content with more evidence of that “something” and its effects especially focused on and around China.

abook-sept-2016-tic-chinaabook-sept-2016-chinabor2-hibor-3m abook-sept-2016-chinabor2-shibor