What makes ‘whatever-it-takes’?

Stay informed with free updates

Simply sign up to the Global Economy myFT Digest — delivered directly to your inbox.

When the macro faeces hits, what should central banks do: announce massive monetary stimulus with a big fat headline number, or take a leaf out of the Draghi playbook and say more vaguely “whatever it takes”?

(As an aside, it makes us sad that the top results for “whatever it takes” on Google, Bing and DuckDuckGo were all an Imagine Dragons song, not the Super-Mario version. (Editor’s note: not all of us are sad about this.))

You might think a more explicit promise to buy a bajillion dollars of government debt is the best way to quell any financial turmoil, but if you were around for the 2012 or 2020 versions you’ll probably guess that open-ended commitments are more powerful — even if they are inherently nebulous.

Still, it’s nice when economists actually examine situations like 2020, prove our intuition, and quantify it. From an new NBER paper by Kathryn Dominguez and Andrea Foschi of the University of Michigan:

. . . We find that on average a central bank’s first whatever-it-takes announcement lowers 10-year bond yields by an additional 25 basis points relative to size-limited announcements, suggesting that communication of potential policy scale matters. Our results for yields hold for both advanced and emerging economies, while exchange rates go in opposing directions, muting their response when we group all countries together.

What is so interesting about the Covid-19 hit of 2020 is that it was a uniform shock that affected virtually every country in the world, and central banks reacted in stages, with different policies at different times.

Between March 2020 and December 2021 there were 166 bond-buying announcements from the 22 central banks studied, of which 120 had a clear limit and 46 were open-ended. Of the open-ended announcements, 14 in some way made an explicit “whatever it takes” reference in the statement or subsequent press conference. This is what allowed Dominguez and Foschi to test the differing impact.

As you’d also expect, the impact was the greatest on the first day of announcement, with negligible effects from subsequent reiterations or new but similar announcements (zoomable version).

Of course, there ARE some complications that make definitive conclusions a bit tricky.

For example, some days saw multiple announcements from different central banks, which makes it trickier to isolate the impact of each individual central bank (eg, whatever the Fed does is likely to have a greater impact on most countries than whatever its own central bank does).

And sometimes there might be a “whatever it takes” reference in the press release, but the messaging is clearly that there are some kind of limits. In other cases there might be a hard number mentioned in the press release.

For example, on March 18 the ECB revealed the €750bn Pandemic Emergency Purchase Programme, but said in a statement that it “will do everything necessary within its mandate”. Similarly, on March 15 the Fed said it would buy “at least” $500bn of Treasuries and $200bn of MBS, but chair Jay Powell later stressed that there was absolutely zero cap on what it was planning to do.

The Michigan economists also stress in their conclusion that even though the impact is more powerful it “does not follow that central banks can rely on whatever-it-takes policy in future crises”. Because:

It is important to note that whatever-it-takes statements embody constructive ambiguity: they are inherently less transparent than announcements with explicit size and duration information. This form of purposeful policy vagueness allows for the possibility that no policy interventions will be taken if the announcement itself is all that it takes. It is also noteworthy that central banks rarely describe the criteria they will use to determine when their whatever-it-takes policy interventions will have accomplished their objective.

Along with the reduced transparency of open-ended operations, there are other downsides to whatever-it-takes policymaking. After a whatever-it-takes announcement is made, it may be harder to impress the market again. Our estimates indicate that subsequent open ended announcements have less impact on asset prices. Whatever-it-takes announcements set a high bar, potentially leading to ever escalating market expectations for large-scale intervention. These types of announcements will also be counter-productive if they inadvertently heighten investors’ fears that economic circumstances are even worse than was thought, or that more standard (size limited) policies are not up to the task. Markets may also worry that if central banks go ‘too big,’ they will have limited their options to address the next shock.

Finally, whatever-it-takes policies are likely to increase moral hazard. Large-scale asset purchases will inevitably increase incentives for risk taking by financial institutions that hold a high share of eligible securities.

Lastly — and a point Dominguez and Foschi don’t explore — is that “whatever it takes” announcements often also require a certain amount of willing gullibility on behalf of investors.

For example, it is often forgotten that Mario Draghi’s “whatever it takes” statement was preceded by “within our mandate” — and he only pledged to preserve the euro, not ensure that every member of the eurozone would forever remain one.

But investors wanted to believe. That meant that it didn’t matter that the “Outright Monetary Transactions” scheme that the ECB hurriedly unveiled after Draghi’s speech was a damp squib. And that kind of central bank magic is harder to bottle and replicate.