In the post-crisis era, one of the more critical debates revolves around whether it’s the “stock” or the “flow” of central bank asset purchases that matters when it comes to supporting risk assets.
Generally speaking, the “stock” effect refers to the notion that the sequestration of assets on G4 balance sheets supports markets by ensuring the supply/demand picture in government and investment grade bonds is skewed towards scarcity. That effectively preserves the “hunt for yield” dynamic that keeps a bid under risk assets like high yield and emerging market credit. It is, the argument goes, the size of central bank balance sheets that’s important.
The “flow” argument revolves around the notion that while “size matters” (so to speak), it’s the marginal bid inherent in monthly asset purchases that is critical when it comes to ensuring that asset prices remain buoyant and that financial conditions remain accommodative. Here’s Goldman, explaining the difference (from a note out last October, following the ECB taper announcement):
A ‘stock effect’, by which the cumulative amount of long-dated bonds owned, and reinvested, by the central bank shrinks the stock available to private investors and acts to depress the term premium. This effect is amplified in countries, like Germany, that have comparatively lower debt, and little net supply. A ‘flow effect’, as the more regular and predictable is the amount of secondary market purchases conducted by the national central banks, the lower the amount of duration risk dealers are asked to absorb when they underwrite government securities in the primary market. This effect is particularly important in countries where sovereign credit risk is higher. And a ‘forward guidance booster effect’: as long as net central bank purchases are running, investors do not expect the central bank to hike policy rates.
Although you’d be hard-pressed to find anyone who falls neatly into one camp or the other (i.e., someone who would argue that one of those two effects is completely irrelevant), some analysts lean demonstrably in favor of the notion that the “flow” matters much more than the “stock”.
Citi’s Matt King, for instance, has been pounding the table on the preeminence of the flow effect for years and back in December, Bloomberg highlighted a Bank of England staff blog reiterating the idea that it is the ongoing support provided by incremental purchases that perpetuates the status quo.
This debate is getting more important literally by the month. The Fed’s balance sheet unwind is taking place in the U.S. against a backdrop of increased Treasury supply (to fund expansionary fiscal policy), a juxtaposition that risks undermining the rest of the dollar bond market as private investors are forced to absorb more U.S. debt. Critics like the RBI’s Urjit Patel have argued that the Fed should calibrate its balance sheet normalization effort to account for the increase in Treasury supply in order to avoid creating a liquidity crisis for emerging markets.
Meanwhile, the ECB will taper purchases to €15 billion/month starting in September and will wind down APP entirely in December, meaning that outside of reinvestments (which aren’t trivial), the flow effect will dissipate.
That’s the backdrop for comments from Macquarie’s Viktor Shvets, who spoke to Bloomberg on Wednesday. First of all, Shvets wants to make sure you appreciate what it is the Fed is effectively doing:
The problem is the Federal Reserve right now is destroying money. They will be forced at some point in time over the next three to six months to stop reducing the size of their balance sheet.
He’s hardly alone in the contention that the Fed will be forced to halt the normalization effort sooner than the committee, and the market, expects. Morgan Stanley was out last week with a 77-page note making the case that normalization will end in 2019 and that the balance sheet will actually begin to grow again in 2020.
But the important point comes at the ~1:45 mark in the clip below, when Shvets takes up the argument outlined above:
Got that? Here it is again:
They seem to believe that the size of the balance sheet is more important than the flow. It is clear from the market that the flow is far more significant than the size of overall accommodation.
He’s probably right. After all, the flow effect represents an ongoing bullish technical; a price insensitive bid that provides continual support for risk assets and acts as a perpetual volatility suppressant.
Of course really, if what you’re leaning on to support a bullish thesis in risk assets is QE, the distinction between the stock and flow effects is only material at the margin, because they have both likely peaked.
Finally, do note that a continual rise in the dollar coincident with tapering (and ultimately, the cessation) of ECB asset purchases works to tighten liquidity on two fronts, as explained by BofAML in a note out last month, excerpts from which I’ll present without further comment.
Note how fast the YoY growth in global QE has declined of late. What’s behind such a quick drop? The ECB has been purchasing fewer bonds in ‘18 and China FX reserve growth has cooled (on the other hand, BoJ QE has been relatively stable, and the Fed continues with modest balance sheet run-off). However, the sharp drop in the global QE pace lately has been exacerbated by the US Dollar rally which began in mid-April. Chart 2 shows the change in monthly QE volumes, just to emphasize the above point. As can be clearly seen, the volume of monthly global QE buying has declined significantly over the last two months, commensurate with the USD appreciation.
As the world’s premiere reserve currency, movements in the US Dollar impart a tightening/easing bias on global financial conditions, beyond just impacting the US economy (chart 3). True to form, there has been a reasonable relationship over time between the monthly rate of change of our global QE numbers (in USD terms) and inflows into European IG credit funds. Put another way, slowing global QE means less crowding into risky assets more broadly… which points to less credit inflows (chart 4).