There was a lot of good stuff in the March 2021 TIC data. Given that Q1 2021 was the most solidly reflationary since the recession and GFC2, we find the same across most of its headline data. From UST buying (while bonds were, in general, selling off and yields were rising; more on this in a minute) overall to bank liabilities expanding the most in several years, that had been the overall theme for the three months of TIC up to and including March.

On the flip side, as much as the first quarter had been most evidently looser, it was also the weakest yet when compared to prior periods. Furthermore, in global bond yields, meaning money, it didn’t seem to get beyond Feb 24-25-26 leaving March as somewhat of a question. Here we also find a few interesting TIC components which offer some clues.

The bigger picture is simply this: does reflation still have legs? Upon that, the entire inflation case rests. The headline numbers don’t really tell us much beyond the facts we already know.

Or, are we already seeing signs – very familiar signs – of its premature demise, maybe even rolling over already toward what would then be the next global dollar shortage/surge (Euro$ #4b; or Euro$ #5 if you’re picky). In that circumstance, truly transitory “inflation.”

Taking the former first, contrary to established belief foreigners actually sell LT UST’s specifically and US$ assets in general when those things are in high demand. Dollar shortages across the world compel liquidation of dollar assets, thus the confusion. Tight money raises the demand for UST’s among market participants who aren’t forced into selling them. Foreigners, typically FOI’s (Foreign Official Institutions), liquidate into a market starving for these very assets.

Conversely, when dollars are less hard to come by – reflation – foreigners tend to buy these assets back even as the rest of the marketplace seems to be selling them as growth and inflation expectations rebound.

Up to now, foreign holders aren’t exactly buying more LT UST’s but they clearly have been forced to sell a whole lot less than immediately in March 2020 and the few months afterward. You can see above how the inverse correlation works; around September, less foreign selling indicating less forced UST liquidations consistent with reflationary higher yields since.

In March 2021, foreign holders went nuts in UST’s (the opposite of what they’d done in February, as noted last month). The net increase of all types of recorded US$ assets among all the various foreign holders surged by a monthly record of $208 billion.

Amidst the buying, corporate bonds found favor more thoroughly reversing the decline which had dated back to late 2018. The inflection began around November, consistent with pretty much everything.

That’s not all. Another key reflationary indication was the first quarterly increase in reported/measured bank liabilities. According to what banks told Treasury about their operations during Q1, total estimated US$ foreign liabilities rose by $126 billion. While liabilities had spiked much more, by $544 billion, during Q1 2020, that was due to the absurdity of the Fed’s “overseas” dollar swaps being roundtripped back home.

Apart from it, bank liabilities hadn’t “organically” increased like they had in Q1 2021 since the first quarter of 2017 – not coincidentally, this prior had turned out to be the first sustained period of what would more completely become Reflation #3 and then later “globally synchronized growth.”

The liabilities chart below both confirms that as well as suggests why it wouldn’t ever transform further into actual growth, falling short as never more than temporary reflation (the balance sheet rebound started strong enough, but then faded quickly).

Already, an important inconsistency. In 2017’s case, the Q1 liability increase kicked off that year’s “dollar crash” whereas in 2021’s Q1 the dollar’s exchange value had stopped falling.

Even the monthly pattern in bank liabilities raises those same suspicions; the increase was due to January and February, while the reported balance declined sharply in March matching the behavior in at least yields (including TIPS). In other words, while the combined quarter appears wholly onboard with reflation, it was only true from its first two months.

Then there is Japan.

You might remember the Japanese playing a crucial role in turning Reflation #3 into Euro$ #4 right from the beginning of 2018 (if not, you can go here and read all about it). To sum up: Tokyo banks had been a key source of dollar redistribution funding into Asia and particularly China, but unlike Western Economists and central bankers they paid closer attention to Xi Jinping’s radical redrawing of the Chinese political landscape, thus actually appreciating what that had meant (no global recovery in 2018 and beyond).

In TIC, this became obvious in what US banks reported to the US Treasury about their own claims on Japan’s banks and non-banks; right in January 2018, a whole lot less. This had indicated an abrupt change of heart on Tokyo dollar redistribution, implying the Chinese connection via higher risks.

Since November 2020, same sort of thing. All the way back during 2019’s “recession scare”, the Japanese apparently jumped back into the eurodollar business and then kept up at it through most of last year seemingly expecting a big payoff in Beijing.

While the West once again spouts off on China’s purported strength and strengthening recovery, that’s not at all what Xi nor Li Keqiang have been saying especially since late in 2020 (no sharp turns). Eerily repeating the very same 2017-18 setup, China’s economy has only disappointed while Communist officials have said to expect it to continue (not the “disappointment” per se, but the low level rebound that is very disappointing to the rest of the world’s needs and expectations).

Underlying this Japanese anti-reflationary withdrawal, the same is taking place across the Caribbean (we’ll have to ask Emil Kalinowski to stick his head out of his window and confirm). Like Japan, Caribbean “banks” (there are, actually, few banks on these islands; bank offices, sure) had re-entered the eurodollar business (in a very specific, curious manner) only to likewise rethink the idea around the end of last year.

Remember, the dollar stopped declining in early January.

This may be why demand for instruments like US Treasury bills have remained high despite reflation being indicated across many markets and data (including those I’ve pointed out here). TIC also shows us that borrowing in repo – on the collateral side – has continued be unusually high since last March.

The process is complicated and even counterintuitive, but I’ve discussed some of it here. For our purposes, this is the major connection:

But there are variations on the theme which complicates matters even further. Not every repurchase agreement is settled by cash. Some are essentially swaps: one security pledged for another with the agreement that both will be re-exchanged, repurchased if you like, at maturity.

A better way to say this form of repurchase agreement is a collateral swap, typically as one leg in a multi-dimensional transformation process.

To put it in very simple terms, cash and collateral gets borrowed, lent, and relent back and forth between American banks and overseas counterparts, especially nowadays, as it turns out, overseas non-banks. You’d think that if the world was experiencing truly strong and lasting reflationary impulse there’d end up being less need for such extremes.

There has been an understandably tight relationship between LT UST yields and the “demand” for cross-border repo funding (in both directions). While the overall level of resale agreements (shown inverted above) with US banks has come back down since last March, it hasn’t very much especially in comparison to prior reflationary periods. There’s still quite a lot going on (more than $1.1 trillion).

This implied heavy demand for collateral might explain one reason why reflationary yields haven’t been nearly as reflationary as otherwise expected. And this idea is further backed up by the flipside, how much in particularly T-bills US banks continue to borrow from foreign non-banks.

Remember, these aren’t T-bills US banks (including domestic subs of foreign banks) are buying from overseas non-banks, these bills are being borrowed (liabilities to foreigners) presumably because they aren’t as cheaply and easily available via other means even though the US government had issued trillions of them (and the Federal Reserve curiously stopped buying them at exactly the same time).

The amount of bill borrowing peaked last August, the same month that LT bond yields bottomed out. But the decline in borrowed T-bills hasn’t amounted to all that much, repeating the same pattern (bottleneck, then only partly unwound) established during and after GFC1. At least then, T-bill borrowing subsided somewhat more substantially up until Euro$ #2 showed up in early 2011.

This time GFC2 and its aftermath, however, US banks aren’t borrowing bills from FOI’s as much as, again, non-banks. And they’re still borrowing at close to peak levels, hinting strongly at another monetary headwind inconsistent with sturdy reflationary momentum (as well as bill yields moving in on zero across-the-board).

There are actually a few more items in TIC which could be potentially revealing. For now, the headline data confirms the overall impression anyone would have had during Q1: reflationary. Underlying, there are several serious reasons consistent with past historical experience which might go further to indicate why: reflation hasn’t actually been all that strong despite its mainstream characterization otherwise; also, several reasons to more closely watch out for it rolling over.