I want to be happy about this (less beholden to china because the hold a large chunck if T-Bills) but This Cannot be a Good sign .......
It's really not that big of a deal. We're not talking about how much of our debt they are holding and/or acquiring, we're talking about the term composition of it. It's more like talking about fluctuations in the amount in a business' petty cash account than talking about fluctuations in the total amount in all of the business' accounts. Holdings of T-Bills are, in general, naturally going to fluctuate more than holdings of longer term securities. Moreover, there are very good reasons for the specific 2-year 'divestment' trend being referred to here.
When the federal government's operating deficits starting exploding in the summer of 2008, due at first to the economic stall and then later to the TARP program, the Treasury started flooding the market with T-Bills. It wasn't issuing much in the way of new (i.e. on-net) Notes or Bonds. Between the end of June 2008 and the end of November 2008, the amount of T-Bills outstanding almost doubled from $1.06 Trillion to $2.00 Trillion, while the amount of Notes and Bonds outstanding only increased by about $142 Billion (to nearly $3.27 Trillion). In part because T-Bills are what the Treasury was issuing, that's generally what our larger creditors were buying if they were adding to their overall U.S. credit position. China soaked up a lot of that debt for us - and they did so in the form of T-Bills. At that time liquidity was king, and there isn't much in the world more liquid than T-Bills. (Not to mention, since there was much greater new supply of T-Bills than Notes or Bonds, that's where the the U.S. debt demand needed to go, lest the price - for the buyers - of Notes or Bonds be driven up even more and their yields be driven down even more.)
The Treasury then started issuing significant amounts of new Notes and, for the most part, stopped issuing significant amounts of new T-Bills. Since the summer of 2009, the amount of T-Bills outstanding has been declining - it's about $400 Billion less than it was at the end of August 2009. The amount that China holds has also, and predictably, decreased dramatically. That just represents their term composition going back closer to where it was. Before the Treasury flooded the market with T-Bills in mid to late 2008 and made them the credit instrument of choice, China held relatively little in the way of T-Bills. Most of their U.S. debt holdings had been longer term instruments. That's what they've been returning to.
As to what that article says about Chinese holdings of longer term securities - its assertion that China is no longer just converting the T-Bills it is 'divesting' from into Notes and/or Bonds, but is actually reducing its total U.S sovereign debt holdings - that assertion seems to be based on monthly TIC data (i.e. estimates). I'm guessing the author of the piece doesn't know a lot about TIC data/estimates and how some of it is arrived at. It may or may not be true that China is divesting from U.S. debt in general and not just from T-Bills (though it certainly isn't doing so by anything near 97%), but we shouldn't rely on the estimates in question as definitive indication of that. Because of the way those estimates are updated on a monthly basis, they aren't reliable in this regard (and are acknowledgedly not so). (I can go a little into why if you are interested, but I'll leave that for now.) Annual updates to that data are generated a different way and are, though not perfect, more reliably accurate. The last time an annual update was released, estimates of Chinese holdings jumped by almost $270 Billion from what had been estimated in the monthly updates (i.e. for the same month). The temporary estimates had made it look like China was reducing its holdings when, based on better information, it now looks like it had been continuing to increase them.
The point being, this particular bit of information is far from being of the sky-is-falling variety. Sometimes a little bit of information ...