After a decade of unprecedented liquidity injections by central banks to preserve the western financial system, global QE has peaked.
First, the aggregate balance sheet of major central banks started to shrink earlier in the year, a reversal that took investors many months to notice but judging by recent market volatility, it is finally being fully appreciated.
Second, beginning this month the Fed's bond portfolio run-offs as part of its QT are roughly offsetting the combined tapered net QE purchases by the ECB and BoJ. Worse, QT is now set to dominate.
Some facts: between mid-2008 and early 2018, the "Big-6" central banks expanded their balance sheets by nearly $15tn, most of it due to explicit targeted purchases of domestic assets (QE) in addition to other forms of liquidity injections (collateralised lending such as the ECB's TLTROs or FX interventions equivalent to foreign-asset QE).
According to Deutsche Bank estimates, the four major central banks involved in QE (Fed, ECB, BoJ and BoE) are now collectively holding $11.3tn of securities accumulated through their asset purchase programs.
Why is the above important? Because as Deutsche strategist Michal Jezek, now that liquidity is contracting makes for a timely moment for looking at the proportion of relevant asset classes owned by central banks and putting the ECB's corporate bond holdings into a wider context.
To begin, as Jezek confirms what we have been saying since the start of 2009, "clearly, QE matters." As central banks reduced the free float of some securities and QE has worked its magic on confidence and growth, asset valuations reached unprecedented levels while volatility became suppressed. A couple of years ago, a quarter of the global bond market was trading with a negative yield. With global QE fading, this proportion has now fallen by half but remains significant.
And the combined valuation of global bonds and stocks has more than doubled since the crisis, led by stocks.
So in light of the recent jump in volatility, and given potentially fragile valuations, QE tapering has been gradual and well telegraphed with the aim to minimize disruptions. Certainly in the case of the ECB CSPP, the market has focused on the taper for quite some time and while many believe it has been largely priced in, Italian bondholders are becoming increasingly uncertain if that is indeed the case.
That said, as DB cautiously admits, "the end of the overall ECB QE and start of global QT may not have been fully internalised by the market yet and the absence of a large, price-insensitive buyer of last resort should keep volatility higher."
To be sure, acquiring a portion (or most) of a securities market over time does not necessarily mean shrinking it. Consider that while the ECB took nearly €175BN of Eurozone corporate bonds out of circulation but net issuance over that period has nearly offset it and broader net issuance from global IG corporates in EUR has well exceeded that volume, as we show in our most recent Issuance and Fund Flows report.
However, QE does shrink the market compared to the counterfactual of no QE because the extra supply response (more issuance because the central bank is buying) should normally be a lot smaller than the net amount bought.
With that in mind, Deutsche Bank presents its estimates of the share of relevant asset classes owned by central banks. The most extreme case is Japan where the BoJ owns almost half of the JGB market, followed by the Eurozone where the ECB holds nearly a third of the relevant covered bond market. Other than that the central banks own 20-25% of their government bond markets and 15-20% of the relevant part of the ABS/MBS market.
As Jezek next notes, compared to all the above numbers, it might seem that there is less scarcity in the Eurozone corporate bond space with the ECB owning "only" about 8.5% of the relevant bond universe (and 20% of the CSPP-eligible subset worth €850bn +). While there is a much greater liquidity constraint than in government bond markets, it is probably true that if healthy net issuance continues, the ECB could carry on conducting meaningful net CSPP purchases well beyond 2018.
Why would they?
According to Deutsche, one reason is that they might decide to put more weight on corporate and less weight on government bonds in their upcoming reinvestment strategy, with zero net QE. Against that, there seems to be limited willingness of the ECB to take unsecured corporate exposures given their recent experience with the troubled Steinhoff and Atlantia credits in the CSPP portfolio. In other words, it now seems more likely that the ownership share will slowly begin to decline as net corporate purchases cease and the IG market keeps growing.
Unless of course the European bond market suffers a cardiac arrest, in which case the ECB will go all out again.
And while the BoE owns about 3% of the UK GBP IG market given its CBPS programme was limited in scale from inception, in bank-based Japan (with a less developed and small corporate bond market) the BoJ has accumulated some 15%, by our estimates.
Looking ahead, Deutsche Bank suggests that as global QT sets in, the turn in the credit cycle still seems relatively distant, with default rates low and still falling, while declining issuance in US high yield has kept spreads supernaturally tight.
The bank forecasts lower but still decent growth at least through 2020 with rate curves bear steepening in the next couple of quarters not only in Europe (consensus) but also in the US (non-consensus) before they start flattening into a slowdown.