Earlier this week, a reader wondered what Harley Bassman, inventor of Wall Street’s most-cited benchmark for rates volatility, thinks of Modern Monetary Theory (MMT).
The question came in response to a reprint of a piece Bassman penned in 2016 on the (remote) possibility that the Fed could eventually embark on a gold purchase program.
In fact, Bassman has weighed in on MMT on a number of occasions over the past two years. For example, last year he wrote that while he doesn’t “believe MMT is viable over the long-term, it is unlikely my personal horizon will overlap its eventual denouement”.
That line gave MMT patron saint (and now New York Times bestselling author) Stephanie Kelton a chuckle. Bassman’s point, obviously, was that while he’s still in the camp which believes the theory is too good to be true (that’s an oversimplification of his position, but in the interest of brevity, I’ll use it here to describe MMT critics as a group), it could be quite a while before the proverbial chickens come home to roost once overt deficit monetization is undertaken.
Bassman was writing well before anyone knew about something called “COVID-19”. The pandemic will likely have a lasting economic legacy, and part of that legacy will be that, at least temporarily, it prompted policymakers to cast aside concerns about debt, deficits, and the monetization of debt and deficits by central banks.
This is a subject that deserves every bit of the press it’s received — and then some.
In a good note out earlier this week, TD’s head of global macro strategy, James Rossiter, takes up the discussion.
The point here isn’t to beat any dead horses (so to speak), but rather to highlight some additional color in order to help connect the dots for readers, who have shown quite a bit of interest in this ongoing debate.
“What has been remarkable about the stimulus measures announced to date is the relatively similar sizes of deficit expansions and central bank QE programs”, Rossiter notes.
“For Canada and the UK, our expectations are that QE will almost perfectly match deficits in 2020 [while] in the euro area, QE is likely to outstrip government deficits”, he adds, before noting that “while in the US, the deficit looks set to out-strip QE significantly, the Fed has been more creative in its crisis response than many of its G10 peers, so QE perhaps does not proxy as well the central bank’s crisis response”.
Obviously, this is not a coincidence. Monetary policy as it manifests in asset purchases is being calibrated based on the size of the debt governments issue to finance virus relief spending.
The buying of government bonds by a nation’s central bank is by definition debt monetization. To say otherwise is to traffic in nonsense. The fact that there’s an intermediary (primary dealers) doesn’t change this. That setup is simply a Wizard of Oz-ish “pay no attention to the man behind the curtain” dynamic.
In their note (dated Wednesday), TD inadvertently underscores how maddeningly self-referential this has all become. Consider this passage, for instance:
As debt levels pile up and QE becomes increasingly pervasive, questions are being asked about the independence of monetary and fiscal policy. As the chart above shows, it looks like the two are acting in a coordinated fashion, with QE closely matching new debt. The re-activation of the UK government’s “ways and means” account at the Bank of England has led to accusations of debt monetization there (we disagree: it is an accounting operation until the government issues more debt later this fiscal year; it has yet to be used in 2020). Pure monetary transfers to households from the governments in Japan and elsewhere have been dubbed “helicopter money” (we disagree: the government is borrowing to finance them). As central banks are swift to say, they are merely pushing back on the same negative shock as the government, so it only appears that fiscal and monetary policies are being explicitly coordinated.
Note the second “we disagree” point. I don’t want to come across as derisive towards TD (because their note is actually quite good) but, I’d be remiss if I didn’t ask readers to consider whether it makes sense to say that monetary transfers to households are not “helicopter money” because “the government is borrowing to finance them”, when it is the central bank that is buying the newly-issued debt. That is question-begging at its best (or worst, depending on how you want to look at it).
TD goes to ask what might well be described as one of the most important questions of our time, economically speaking. To wit:
What’s so wrong about central banks permanently raising the monetary base as the government borrows, especially if it’s debt monetization rather than helicopter money?
That passage generally hews to the accepted notion that we aren’t splitting hairs when we distinguish between debt monetization and helicopter money.
I would argue that it is splitting hairs — especially at this juncture, when governments and central banks are so clearly coordinating.
Here is TD laying out the supposed difference:
Creeping into the Unconventional: “Simple” debt monetization is when a central bank buys new government debt in the primary market and holds it permanently. If the central bank precisely buys all new debt issued, then debt held by the private sector would not change, despite higher government deficits and debt. This would limit the response of yields on government debt to the new supply of debt, and could boost economic activity by eliminating Ricardian equivalence. There are also more subtle ways to conduct debt monetization, as some G10 economies have done in the 20th century.
Uncharted Territory: Despite the term being thrown around loosely, helicopter money is something that’s been off-limits for most central banks since inflation targeting regimes became the norm. There have been more subtle, indirect ways of central banks financing government debt, but on the whole, the concept of a central bank buying government debt and writing it off (or simply printing money for the government) has seen little practice in developed economies in the last few decades.
The Ricardian equivalence point is notable, but before I touch on that, I’d kindly ask readers to consider whether there really is a distinction between the two purportedly different concepts described in those simple paragraphs.
While it’s true that the market would take note (and that’s putting it mildly) of a decree by a government to cancel all of its own debt held by the central bank, one wonders if that would really have much of an effect on consumer inflation expectations in developed economies.
Sure, such a move would push up market-based measures of inflation expectations dramatically (indeed, for the first day or two after a hypothetical cancelation of central bank-held government securities, there would be all manner of volatility), but does the public in the US or Japan or France or the UK really know enough about QE to react in a way that would trigger a self-feeding hyper-inflationary spiral? The average American doesn’t even know what the Fed is, let alone what the Fed does, let alone what QE is. And we’re supposed to believe that if the Treasurys on the Fed’s balance sheet were simply canceled, that Americans would suddenly behave in a way that would drive up prices for things like bread? The vast majority of Americans aren’t even financially literate.
But let’s leave that extreme example aside, because it’s controversial and, in the event the public did figure it out, it’s possible things could go decisively wrong.
TD assumes simple debt monetization, where the charade is maintained, and then elaborates on why this may be desirable under the current circumstances. To wit, from the same note:
There are some fundamental economic arguments in favor of debt monetization, especially in the current circumstances. First, expanding the monetary base permanently should in theory be somewhat inflationary, and in a world where debt levels are surging, even if to still-sustainable levels, a little extra inflation could be a welcome development to deflate debt. Second, to the extent that fiscal spending is financed by the central bank, then Ricardian equivalence is less likely to hold, as individuals and firms won’t expect to have to pay higher taxes in the future so that the government can pay off its debt. They are therefore more likely to spend now, when the economy most needs it. While it’s not been used recently, there are a number of examples of developed economies successfully pursuing forms of debt monetization outside of wartime during the 20th century. In many (if not most) cases, the policies were complex, opaque, and importantly, not inflationary. Japan, under guidance of Finance Minister Takahashi conducted monetary financing in the 1930s to try to boost economic activity, and Canada did so in various forms from the 1930s until the mid-1970s, in part via the Bank of Canada’s Industrial Development Bank.
Those are all crucial points. As Bassman wrote in 2016 (and as plenty of other commentators have reminded the world at a time when debt continues to pile up), “there are only two avenues out of a debt crisis – default or inflate, and inflation is simply a slow-motion default”.
That kind of “slow-motion default” wouldn’t necessarily be the worst thing in the world for lower- and middle-income earners. If they can secure enough in the way of wage growth to keep pace with inflation, the people who will be disproportionately impacted by a “slow-motion” default would be creditors of all sorts.
On the Ricardian equivalence bit, note that I’ve discussed that at length in these pages, most recently in “‘It’s The Private Sector, Stupid’ – Trump May Have A Ricardian Equivalence Problem“. If the central bank holds the debt, then it may change the psychological calculus that, if you buy the theory, says the private sector will preemptively tighten its belt during periods of belt-loosening by the public sector.
But perhaps the most crucial point of all comes when TD notes that we crossed the Rubicon on this years ago, once it became clear that the total unwinding of central bank balance sheets was unlikely at best, and probably impossible. Of course, if you don’t ever allow the assets to runoff, then it’s debt monetization. Here’s one last excerpt from the note:
Central banks that have been conducting QE for over a decade and fully reinvesting maturing assets, like the BoE and ECB, are to some extent already conducting stealth debt monetization, and it will certainly ex-post look like debt monetization if balance sheets are not run down any time soon. But inflation over this period has remained tame, and in most cases below central banks’ targets (and certainly never far above them). The distinction that needs to be made here is that true debt monetization involves an explicit, ex-ante commitment by the central bank to permanently reinvest maturing debt forever, whereas the BoE and ECB currently see their reinvestments as temporary aspects of their policy regime, and still intend to one day let debt assets mature (even if that may be a long, long way in the future). The Fed avoided this label to some degree between 2017 to 2019, when it allowed its balance sheet to slowly run off.
The Fed may have “avoided” that label temporarily, but it was always explicit that the post-financial crisis balance sheet would be larger than pre-crisis, and that was before COVID came calling.
Now, the situation for the G4 looks like this:
Do not kid yourself: That is never (ever, ever) going to be unwound — not completely. This will never be “normalized”. In fact, it makes no sense to use that term anymore, because we’re a dozen years in — what you see in the chart is normal.
And that, folks, is the point.
“How long until a central bank decides that it needs to maintain the current size of its balance sheet to successfully implement monetary policy, and announces that it intends to fully reinvest its government debt indefinitely?”, TD asks.
We do not have to wait until central banks say that explicitly. The fact that the Fed was forced to start buying bills last autumn after the consequences of balance sheet rundown became clear in funding markets is evidence that even tentative steps down the road to unwinding balance sheets will cause technical tremors in the market’s plumbing. Giant strides down that road would lead to tantrums and worse.
But the funniest thing about all of the above is that the entire discussion rests on an observably false narrative about the sequencing of government spending.
In other words: Everything said above only makes sense if you assume something that isn’t even true in the first place.
On Saturday, in a tweet, Stephanie Kelton wrote that “MMT shows that it makes no sense to ask, ‘How should the government finance its spending?'”. She added the following, which I’ll present without further comment.
There is only one way to pay. All spending is already “money-financed.” So many people fail to understand this and end up saying that MMT advocates monetary-financing. That is wrong.
I explain this in the book, using what I call the S(TAB) model. Taxes And Borrowing are secondary to Spending. They do not finance the spending.
By the time the government sells bonds, the spending has already been financed.