SpongeBob SquareBanks.

It might seem strange to suggest that developed market central banks aren’t buying enough bonds right now, but that is precisely where the debate is headed in the second half of the year.

“Enough” isn’t a normative term in this case. “Enough” simply refers to whether monetary policymakers are prepared to absorb and soak up a sufficient amount of new supply tied to fiscal stimulus to keep a lid on yields, thereby suppressing borrowing costs for their respective sovereigns and averting any kind of damaging “tantrum” in markets.

Asset purchases have, of course, ballooned over the course of the crisis. Somehow, “unprecedented” seems like an inadequate adjective when it comes to describing the sheer scope of the effort.

That visual is from Morgan Stanley, but there’s nothing special about it. It’s just derived from public information and you can construct any number of variations on it if you’re so inclined. The point is always the same — namely to provide some visual context for the recent splurge.

As Morgan writes in the chart header, the ECB has moved beyond the capital key with its Pandemic Purchase Program (PEPP), and that’s apparently causing some internal consternation at the bank.

As documented here, deviations from the capital key were material during the first two months of the facility’s operation, which is a euphemistic way of saying Christine Lagarde went from asserting that the ECB isn’t in the business of tamping down spreads in the periphery to launching a program (with €1.35 trillion in firepower) aimed at doing just that.

PEPP’s “flexibility” is arguably its most powerful feature as it allows the ECB to ensure that vulnerable locales (i.e., Italy) don’t lose market access as they struggle to recover from the pandemic.

But, now that the crisis has abated, the hawks are circling.

“‘Flexible’ doesn’t mean ‘unrestricted’”, Jens Weidmann said recently. “It’s important to me that monetary policy doesn’t set wrong incentives for public finances. In this context, the capital key offers the ECB a sensible guideline”.

This is just another manifestation of the global charade aimed at obscuring the fact that QE is arm’s length debt monetization/deficit financing. A German high court tilted at this windmill in May, largely to no avail (the German parliament backed the ECB this week). The plaintiffs in the case will apparently challenge PEPP on similar grounds to the original case, which challenged PSPP (i.e., “legacy” QE, if you will).

In any event, the point is that as governments borrow and spend to help their economies recover from the crisis, central banks will need to absorb the supply. The reason I include the bits about the ECB and PEPP is that things are a bit more complicated for nations like Italy which do not issue their own currencies.

Walking back out of the tall grass and onto the fairway (as it were), the broader issue is how much of the increased duration supply central banks will absorb in the back half of the year following the initial bonanza.

This will differ by locale and also depends on whether buying is front-loaded or back-loaded. Consider this from JPMorgan:

Over the past few weeks we have seen a number of central bank announcements on balance sheet policy. These have effectively reinforced the contrast between a front-loaded response from the Fed in the form of aggressive balance sheet expansion in late March and April with the building and perhaps more persistent response from the ECB as it extended its PEPP program not just in size but also in terms of time horizon to mid-2021, which we see likely extended again in 4Q20. At the same time, the likely more back-loaded fiscal response, particularly in the US where initially the deficit has been financed heavily by bills with bond issuance stepping up more from May onwards, has raised questions over what it means for overall duration absorption by central banks.

The Fed put a floor under Treasury (and MBS) buying at the June meeting, so we know what the minimum is in terms of monthly purchases. Meanwhile, the ECB expanded “regular” QE in March, and then topped up PEPP in June, while the BOE recently upped its own purchase program, albeit while simultaneously hinting at a wind down. As for the BoJ, they are clearly inclined to accommodate Abe’s massive stimulus measures, but YCC means purchases are to a certain extent dictated by the market, assuming you want to call a situation where the central bank owns nearly 50% a “market”.

Rolling all of this up, JPMorgan notes that the implication is “a downshift in the aggregate G4 QE pace from first to the second half of the year”. Quite a bit of this is due to the Fed backing off the previously unimaginable pace of purchases from March and April.

That doesn’t answer the question about how much duration supply the benefactors with the printing presses plan to absorb, though.

In order to address that, JPMorgan converts central bank holdings and outstanding government bonds to 10-year equivalents. That’s a (relatively) straightforward exercise, but extending it into the future requires assumptions. Here’s JPMorgan:

For the Fed, we assume it will distribute its purchases with weights across the curve that are roughly proportional to the maturity distribution in the market. For the ECB, we assume that the average maturity of its holdings remains broadly stable. For the BoE, we assume the average duration of its holdings continues to grow, but at a slower pace than in 2Q as its purchase pace slows. And for the BoJ, we assume the average maturity of its holdings will continue to gradually decline as the increases in its purchase schedule suggests the increases in its purchases in 2Q have been focused in the sub-10y maturity bucket.

The takeaway is that only the ECB appears poised to keep soaking up duration at a materially faster rate.

This matters to the extent stimulus is funded at the long-end. “The question is who will step up to buy in the second half when we expect coupon supply to be significantly higher than Fed purchases”, BofA’s Ralph Axel and Olivia Lima wrote this week.

The somewhat cruel reality is that in order to attract buyers, yields need to rise, but rising yields aren’t exactly conducive to funding massive stimulus. BofA notes that pension funds and insurance companies are “the main buyers in the long end”, but they’re also “the most price sensitive”. For Axel and Lima, it would take “a 50bp increase in rates… to move the needle” when it comes to stoking demand from the LDI crowd.

Absent that demand, and assuming a continual pickup in economic activity, long-end yields could rise, steepening the curve.

“Given the likely deterioration in bond demand from private investors relative to 2019”, a decrease in the amount of duration absorbed by G4 central banks in the back half of the year “poses some upside risk to long-end yields, in particular in the second half”, JPMorgan says.

Of course, if the wheels fall off for the economy, it could stoke a safe-haven bid for Treasurys. At this point, maybe we should stop calling it a “safe-haven bid” and just call it demand tied to the assumption that if the macro outlook goes south again, one can squeeze out more gains in bonds by riding the next leg of the rally, which would invariably push US yields to zero and below.

In the same vein, another turn for the worse economically would likely force the Fed into ramping up buying beyond the new floors established last month.

For Jerome Powell, there are quite a few overlapping, self-referential dynamics at work. One way to “solve” that moving Venn diagram is yield curve control.

---