If there has been a lot different about the past few months, it was reflected in the TIC figures and then some. What is usually pretty easy to decipher, there were instead all sorts of shifts across the most important categories. For one, the foreign official sector was busy in December buying up UST’s and dollar assets just as the foreign private sector was dumping them. We had already heard about Japan’s activity with regard to UST dispositions, but in reality overall it wasn’t enormous, just different.

The Treasury Department estimates the private side selling a net of $40 billion in UST’s, and just $32 billion overall (+$12 billion agencies, +$5 billion corporates, -$9 billion equities). The official sector that had been furiously selling UST’s up until October purchased a small amount of US$ assets in November ($1 billion) and a larger one in December ($18 billion). It was the largest monthly accumulation for foreign central banks since June 2014.

That may not have been a surprising shift, however, given that there has been some correlation between less “selling UST’s” and even net buying in the past during months when the private sector is selling them instead. It makes sense if we think about it from the perspective of what central banks are attempting to do by “selling UST’s”, delivering dollar balances to those banks. Therefore, if there is heavy selling in the foreign private sector which includes foreign banks, they have already raised their own balances such that their local central bank may not be needed.

The most prominent example, but not the only one, of this private/official displacement was in the middle of 2013 during the so-called taper tantrum. In May 2013, the official sector purchased an enormous quantity of UST’s, as reserves, while the private sector was selling them. The situation continued the very next month, June and the full rout in MBS, where private selling was enormous but for the second straight month official activity despite what was a massive dollar warning pushed toward net accumulation.

That raises the question as to how we might classify and attribute this particular bout of selling in the private sector. Was it an organic part of the bond market “reflation” trade, meaning selloff? Or was it a possibly a margin call from central banks calling back past “dollar” operations like maturing forward cover? It is, of course, impossible to determine from raw figures. However, we should note the drastically different behavior in reported bank liabilities in December and therefore Q4 overall.

The established quarterly pattern had already been violated with weaker initial monthly retracements in both Q3 (July) as well as Q4 (October). In Q3, at least, the quarter end drain proceeded largely as it had in those prior quarters dating back to the middle of 2013, not coincidentally. In December 2016, by contrast, there was hardly any drain at all.

It was, in fact, the least negative final month of any going back to, once again, the bond market “reflation” rout in the middle of 2013. I have to think both of these quarterly results were attributable to the private “selling of UST’s” which suggests to me a margin call raising up cash balances in the aggregate. It is once more unclear as to the motivation for it.

That meant overall for the month of December, net activity, netting both private and official, was to the sell side, -$14 billion combined and the third combined negative of the past four months. It suggests redistribution of dollar activity (margin call to central banks) than perhaps a change for it. The specific target was apparently Japan, again no surprise as the figures from the Japanese side were release over the weekend, as well as the direction and intensity of JPY through at least the middle of December (it shifted back to appreciation again on December 15; suggesting that the selling of UST’s, if these patterns hold, might also have abruptly changed starting with January 2017’s figures).

With regard to China, the TIC estimates for that country somewhat disagree with the same related figures from the Chinese side (SAFE). There was less “selling” officially in January from China’s view but not December, perhaps indicating a difference in timing or booking settlements of transactions, as the behavior of CNY strongly indicated in December other kinds of “dollar” influence and transactions that are scheduled (“ticking clock”) to expire (by inference) either later this month or the middle of next month.

It is difficult to draw conclusions from one month (or quarter) of data, so further confirmation next month and those thereafter is needed. If there is indeed a change in pattern from the banking side, that might require testing other assumptions about what different behavior might indicate – as well as what response from the official sector there may be in relation. It would not be out of the question for that to happen, as something similar occurred in the TIC data in the “reflation” trading of late 2013 and into 2014. After all, 2013’s massive dollar warning in terms of bank liabilities was also to the plus side.