Kartik B Athreya

On behalf of Tom, the organizers, and the team here in Richmond, I’d like to welcome each of you to today’s “Symposium on Monetary Policy and Financial Stability.” I’ve been given a role by the organizers — so I thank them for that and hope that I won’t make them regret it.

As a research director, I’ll report on how I’ve seen research in monetary policy and financial stability economics as having moved and how those respective research programs might help the policymakers here attack the questions posed in today’s gathering.

The two questions around which the panels are organized are of course, tightly related. I’ll start roughly with the second, and you’ll see that it’ll bleed hopelessly into the first as I move on.

My formal education in monetary policy, now 20 years old, emphasized money as an institution to alleviate trading frictions arising from informational asymmetries and decentralized trade execution and settlement. I’m thinking of the work of Randy Wright, Steve Williamson, Nobu Kiyotaki¹and others.  The models that then quantified the implications of those frictions, though usually via some shortcuts for tractability, were clear: Monetary policy was largely a distraction. That is, maybe it mattered, but certainly not so much (Though, Lagos and Wright did make the point that when you model search frictions, the costs of inflation are much higher than when you do the typical Lucas-style calculation).²

At any rate, what I learned pointed to “low-for-long” nominal rates as not a scary, but rather a sensible thing. Namely, those models were almost universally ruled by the black-hole-like gravity of the Friedman rule — zero nominal rates all the time. Indeed, they accepted the deflation that would typically come with it (!). I’ll note that no less a figure than Narayana Kocherlakota once wrote a paper with Hal Cole titled “Zero Nominal Interest Rates: Why They’re Good and How to Get Them.”³ And I didn’t see any connections drawn to today’s topic of how monetary policy and financial instability were linked. And as far as I could see, neither did anyone else.

So imagine my surprise when I first saw central banking practice from the inside and realized it bore no resemblance to how I’d learned to frame things. The nominal short rate in late 2000 was about 650 basis points (bps), so just about … 650bps higher than where I’d learned it ought to be. And, then, by January 2002, it was about 175bps. The shock was plausibly a clearly financial one as the tech boom went bust. And policymakers seemed to think this 475bps reduction mattered not just to keep inflation on track in the face of financial instability — here a large decline in equity prices — but to preserve employment and credit conditions too! And policymakers definitely didn’t seem pleased with the triumph over shoe-leather distortions they had presumably just delivered. So clearly, I was more than a bit out to lunch.

Of course, many — and they would say “sensibly!” — never got on board with models where monetary policy was either tangential to managing the real economy or was something to be set monomaniacally to achieve the elimination of transactions distortions. And I started seeing how policymakers instead thought hard about how nominal rate policy ought to be set to balance employment (or real activity) and price stability goals.

Luckily, the New Keynesian project — led by, among others, Rotemberg and Woodford and Richmond Fed Hall of Famers Goodfriend and King, and my colleague Alex Wolman — exploded right around 2000 and allowed a way for many economists of my vintage to understand why you didn’t want just zero nominal rates all the time.⁴ But quantitatively, if it gaveth a way to avoid zero nominal rates all the time, it tooketh away too: A raft of papers made a very clear case that monetary policy ought not to do much besides doggedly pursue price stability, modulo trending relative prices perhaps.⁵ Accommodation — at least beyond the lowering warranted by a falling r* — was not desirable.

But while these models of monetary policy also had a clear message, and while they provided clarity on when “accommodation” might be appropriate, they still gave us nothing about how financial stability should affect, or be affected by, monetary policy.⁶

Then, in 2008, the global financial crisis (GFC) hit us with gale force. And just as before, central banking practice was perhaps once again way ahead of theory. Central bankers were not so sanguine about pursuing price stability alone and ignoring either the real economy or financial stability. As the zero lower bound began to dominate, policymakers explored — and repeatedly found — creative ways to matter for long-term rates and ways to shore up intermediation beyond the narrowest kinds of lender of last resort activity.

My personal rearview mirror suggests that the response to the GFC and recession certainly did not offer an immediate repudiation of the massive and previously unimaginable actions of world central banks.  Policymakers acted, real disaster was plausibly avoided and key to today’s events: Nothing terrible in financial stability occurred because of it (at least not yet) … all while something near price stability was maintained throughout.

And low-for-long is a place where we’ve stayed, aside from a brief normalization, for a full decade now: And our balance sheet has remained huge. And has private sector stuff in it. And we have made an even stronger near commitment to low future rates. And our implementation operations are forever altered. And the fiscal environment has changed the supply of interest-bearing safe assets at a time when we too have one to offer the marketplace! But, throughout all this, we seem to have done well at avoiding anything like serious financial stability.

In this sense, maybe the central banking community can be happy with just how much it can accommodate without creating financial instability — banks were even a bright spot in 2020, pointing to the value of being well capitalized.⁷ We’ve been forced into a weird experiment, and financial stability seems not to have been compromised (famous last words, perhaps). Future work will hopefully tell us much more about this natural experiment, including the effects of the innovative and massive monetary policy response. I realize that many think we are creating the conditions for future financial instability by holding rates low now and purchasing assets and other like-interventions, but I am not as worried, because I take signal from the way I see the past two decades.

Still, in the longer run, one would hope that economics would do what Reagan asked of it: Help us understand why this level of activity didn’t create financial stability disaster.

From a research director’s perspective, a first step is to acquire a decent handle on where financial instability comes from. And here, we have made progress — between Diamond-Dybvig and the massive literature that followed — that was more specific to the institutions in question. Indeed, the research program has pushed much further into the subtleties of runs — surely one of the things at the very core of financial instability — left unclarified by the early models and the original McCoy of Diamond and Dybvig.⁸ As a result, we know much more about landscapes that feature “fragility,” and when and how. But it must be acknowledged that we researchers still cannot answer the panel’s concern head-on: We don’t know how systematic monetary policy should be calibrated against a backdrop of such fragility. Nor, really, can we say how the monetary policy posture matters for the arrangements that emerge.

These tracks of course got a jump-start in the post GFC period, when work being done on fragile intermediation started to become central to those previously thinking mainly about monetary policy with no concern for intermediation. After all, the incompleteness of most monetary policy models lay in the completeness (in the Arrow-Debreu sense) of the financial markets they implicitly imagined, and from there, you got quickly to the idea that the remit of central bank fiddling probably should be restricted, and you have nothing much to say about how it matters for fragility in financial arrangements or at the household or corporate level.

But once we have models where these markets are characterized by the warts of the real world: limited commitment, unregulated maturity transformation and a vaguely defined safety net, there seems no reason to think central bankers can look only at price stability and get what they want on the real side of the economy.⁹

So research needs to keep investing in work that identifies, ideally from primitives, contractual arrangements that might be privately optimal while leaving the system open to collapse and then connecting that to the level and (state-contingent) trajectory of the nominal rate. This will take a while.

But you policymakers can’t wait until we in research know to act — and luckily, maybe we don’t have to. After all, I’ve argued that you guys keep finding ways to be useful while guys like me can’t figure out why it all seems to work (or at least not blow up).

So as you forge ahead, I want to point to one thing about the post-GFC period that I think has been key to allowing monetary to accommodate while not leading to turbulence on the financial stability side. And this is the robust evolution of macroprudential regulation. I think of people like Nellie Liang and others who led us so forcefully to macroprudential regulation that includes stress testing and other diagnostics and their successors, who now keep that machinery rolling, hold the line on dividend distributions and conceive evermore extreme scenarios to evaluate capital adequacy. I can’t help but think that the shift in the regulatory landscape has been just vital in allowing the kind of fruitful aggression, and even experimentation, on the monetary policy side that we now see.

Now, as a staffer in this specific Federal Reserve Bank, I’m housebroken enough to not show too much excitement for a path to nirvana that runs through an ever-expanding regulatory remit and safety net. The perimeter of regulation matters, and the stuff that happens beyond it is scary and endogenous to policy. And while banks are vastly better capitalized now, fragile intermediation still shows up from time to time (think: prime money funds last year) as a preferred arrangement for the private sector. Indeed, the government-sponsored enterprises are no closer to being reformed. As I noted, how marketplace incentives evolve with the systematic component of monetary policy is not well-enough understood, though leaders like Eric Rosengren continue to urge us to not lose the plot.

And with a nod to Jeremy Stein, I’ll also admit to worrying that it is because monetary policy gets in all the cracks that it’s a bit unnerving. So two cheers, at least, for macro (and micro) prudential supervision, for its role in disconnecting monetary policy from financial stability.

I’ll close with a short thought on the specific question of panel 1: How are monetary policy and financial stability linked? In a sense, I’ve spent my time so far telling you that a lot of work needs to be done to answer this. But I can be a bit more specific. One strand of recent research progress looks to me pretty promising: This is the program that tackles heterogeneity in macroeconomics head-on, with both theory and empirics.

On theory, the splashiest and perhaps most germane component of that work is seen in the arrival of HANK and Her Sisters. But it’s less just HANK, per se, and much more the admission into our monetary models … of balance sheets. The lynchpin of these models is to admit heterogeneity and balance sheets at the household, real-firm and intermediary levels that are rich. This allows us to connect real interest rates, inflation and the financial fragility that households and firms, in collaboration with intermediaries who must themselves issue liabilities and hold nontrivial — and potentially mismatched — gross asset and liability positions.

On empirics, to my taste, Mian and Sufi’s seminal¹⁰ — and forensic — empirical analysis, along with pioneering counterparts thinking about the micro-empirics of firm-side intermediation,¹¹ have shown the way to help the models we now have find quantitative relevance and allow for the kind of clean counterfactuals needed to really gain insight — from a research point of view — on today’s topic. Indeed, the absolute explosion of ongoing empirical work that ties rates to balance sheets, consumption and investment decisions using high-quality micro data at a time when we have models with which to interpret them quantitatively, has to make one feel hopeful.¹²

So in summary: Received economics seems to me anyway not yet in a place where households’, firms’, and intermediaries’ balance sheets susceptibility to collapse can be tightly tied — in even a probabilistic sense — to monetary policy choices. But, it’s hard not to be optimistic in light of the activity levels in the profession in this area. And while that understanding is developing, I take cheer from the creativity of monetary policymakers and the diligence of regulators who together have helped us navigate so well an extremely challenging decade and a half.

So on that happy note, let’s now hear from our distinguished panelists.

Thank you.