IMF Blog IMF Blog

Ten Take Aways from the "Rethinking Macro Policy: Progress or Confusion?"

On April 15-16, the IMF organized the third conference on "Rethinking Macro Policy."

Here are my personal take aways.

1. What will be the "new normal"?  

I had asked the panelists to concentrate not on current policy challenges, but on challenges in the "new normal." I had implicitly assumed that this new normal would be very much like the old normal, one of decent growth and positive equilibrium interest rates. The assumption was challenged at the conference.

On the one hand, Ken Rogoff argued that what we were in the adjustment phase of the “debt supercycle.” Such financial cycles, he argued, end up with debt overhang, which in turn slows down the recovery and requires low interest rates for some time to maintain sufficient demand.  Under that view, while it may take a while for the overhang to go away, more so in the Euro zone than in the United States, we should eventually return to something like the old normal.

On the other hand, Larry Summers argued that, while debt overhang was clearly relevant, more was going on.  He expanded on his secular stagnation hypothesis, arguing that, in the context of a chronic excess of saving over investment, keeping the economy at potential may well require very low or even negative real interest rates.  He pointed out that real interest rates had started declining long before the crisis, and pointed also to the further decline in long rates since he first stated this hypothesis at an IMF conference last year. If he is right, the new normal will not be like the old normal.

My own sense, based on the work we have done at the IMF  and work by others, is that debt overhang is indeed playing a role—although one which varies across countries and type of debtor—but that the decrease in real rates, which was visible before the crisis, is likely to persist in the future.   So, I am closer to Summers (although I am not sure that we will need negative rates in the future) than to Rogoff.

2. What the new normal will be matters a lot for policy design 

If Summers is right, this has major policy implications. If equilibrium real interest rates really are going to be negative, the zero lower bound (or, as we have learned, the “nearly zero” lower bound) may make this impossible to achieve.  Even if they are positive but close to zero, and if low interest rates lead to excessive risk taking, this may create very tough trade-offs for monetary policy. Instead, fiscal policy may be a more appropriate tool for dealing with the new normal.

On the latter, perhaps the most provocative conclusion of the conference was offered by Brad DeLong:  If the rate at which the government can borrow (r) is less than the growth rate (g), then, he argued, governments should increase, not decrease, current debt levels. If people value safety so much (and thus the safe rate is so low), then it makes sense for the state to issue safe debt, and possibly use it for productive investment.  And if the interest rate is less than the growth rate, debt is safe: the debt- to-GDP ratio will decrease, even if the government never repays the debt.

One senses that the argument has strong limits, from the likelihood that r remains less than g (the two letters appear to have become part of the general vocabulary), to the issue of what determines the demand for safe assets, to whether r less than g is an indication of dynamic inefficiency or some distortion, to whether, even if this world, high levels of debt increase the probability of multiple equilibria, rollover crises and sudden stops.

3. Can we hope to limit systemic financial risk? 

The main underlying theme in the discussion of financial regulation was whether it had successfully reduced financial risk.  There was agreement on "reduced," not so on "successfully." There is little question that new regulations, from Basel III to Dodd Frank, have reduced risk in banking.  But some panelists, in particular Anat Admati, argued that this was far from enough, while others argued that the risk had been largely shifted to the shadow banking system.

The discussion became granular. The construction of real time systemic risk measures for major banks by the Vlab at NYU, presented by Viral Acharya, shows how much progress has been made in defining and measuring systemic risk, at least in the banking system. Robert Rubin and Philipp Hildenbrand emphasized the heterogeneity of what is meant by shadow banking, and suggested an approach to regulation of the shadow banking system, focused on functions or activities rather than entities. In a nutshell, if it involves maturity transformation, liquidity mismatch and leverage, then it is banking and needs to be regulated.

Yet another theme was the unintended effects of regulation.  Acharya argued that incorrect regulatory risk weights had led to systemic risk remaining too high in the European banking system. Rubin and Jaime Caruana discussed the effects of regulation on the reduced role of traditional market makers, and the resulting decrease in market liquidity.

4. Should monetary policy go back to its old ways?   

Before the crisis, most advanced country central banks had converged to a largely common framework.  Some had only one mandate, to keep inflation low and stable; others had a dual mandate, to also keep the economy operating at potential.  The main instrument was a policy rate, anchoring the term structure of interest rates.

Should this framework be maintained? Ben Bernanke’s answer was largely yes. Once economies were out of the zero lower bound, most of the programs introduced during the crisis should be put back on the shelf, ready to be used only if there was another sufficiently adverse shock. The main instrument of monetary policy should again be the federal funds, or, if the central bank wanted to maintain a larger balance sheet and induce banks to hold excess reserves, a combination of the federal funds rate, a repo rate and the rate on excess reserves.

Bernanke however raised the issue of composition and size of the balance sheet of the central bank in the new normal, and the issue was taken on by others, in particular Ricardo Caballero.  With respect to composition:  If, as some argue, there is a "safe asset shortage," does it make sense for central banks to hold, as they did, a large amount of those safe assets?  Wouldn’t it be better for the central bank to hold other assets, and leave those safe assets in private sector hands?  With respect to size:

If the central banks are in a unique position to be able to supply safe assets shouldn’t they do it? I see this discussion as having just started, raising deep issues about the private demand for safe assets, and the potential role of central bank in this context, both in advanced economies and emerging markets.

One question in this context is what is behind the demand for safe assets. Caballero remarked that, if the demand for safe assets is a demand for hedges against macroeconomic fluctuations,  long bonds may actually be safer than short bonds or cash: they are more likely to go up in price when times are bad (and the central bank decreases interest rates), thus providing a better hedge against macroeconomic fluctuations.  But if the demand is instead for assets with known collateral value, investors will want short bonds.  In the first case, the central bank should hold short bonds and leave the long bonds in private hands (the old normal); in the second, the conclusion is the reverse.

5. Instrument rules 

The session led to an exchange between Bernanke and John Taylor on the role of “instrument rules.”  Taylor argued that a deviation by the Fed from a rules-based policy (too loose a monetary policy in the years preceding the crisis), along with a regulatory process which broke rules for safety and soundness, was a key factor in the financial crisis.  He argued for a quick return to a rules-based policy for the instruments, what he called "renormalizing monetary policy."  Bernanke disagreed and answered that, in the complex world in which we live, the right rule would be very complex, and keeping rigidly to a simple rule would be counterproductive. Judging the central bank on how it fulfills its mandate, rather than requiring it to follow a simple rule, he argued, is the way to proceed.  I agree with him.

6. Macroprudential tools or financial regulation  

The semantic issue of what is meant by "macro prudential tools" came up, and Paul Tucker suggested the following definition:  "the choice of dynamically adjusting regulatory pararameters so as to maintain systemic resilience." The definition is heavy, but helpful.   I indeed see the main difference between "macro prudential tools" and “financial regulation” as the fact that the first are "dynamically adjusted" and the second is not.

The issue of when one should go for "dynamically adjusted" tools, or simply for tougher financial regulation (say, go for variable capital ratios or, instead, for higher but constant capital ratios), which I see as central, was not discussed in this form.  Tucker however argued for the use of macro prudential tools to deal with "exuberance," not necessarily for fine tuning in more normal times.  Rubin remarked that we were not very good at telling when times were exuberant rather than normal.

Also discussed was the issue of the relative roles of monetary policy and macro prudential tools.  From a macro viewpoint, Shin argued, one could think of both as affecting the demand and the supply of credit.  From a financial viewpoint, one could think of monetary policy as a general instrument, and macro prudential as specific instruments.  During the discussion, Lars Svensson argued that a simple cost-benefit analysis suggests that monetary policy is a very poor instrument to deal with financial risk.  In a cost-benefit analysis of the Swedish case that uses Riksbank estimates, the unemployment cost of increasing the interest rate far exceeds the benefits of a potentially better future macro outcome from a potential reduction in the probability and severity of a financial crisis. I found his arguments convincing.

7. Should central banks keep their independence?   

Throughout the conference, e.g., in Gill Marcus’ talk, and actually throughout the various meetings which took place during the IMF meetings in the following days, policy makers remarked and complained about the heavy burden placed on monetary policy in this crisis, and the danger of a political backlash against central banks. Even as the crisis recedes, it is clear that central banks will end up with substantially more responsibilities—whether they are given in full or shared—for financial regulation, financial supervision, and the use of macro prudential tools. While even the use of the policy rate has distributional implications, these implications are much more salient in the case of regulation or macro prudential tools, such as the loan-to-value ratio. The general consensus was that these distributional implications could not be ignored, and that while central banks should retain full independence with respect to traditional monetary policy, this cannot be the case for regulation or macro prudential tools.

8. Little progress on the design of fiscal policy  

The traditional objection to using fiscal policy as a macroeconomic policy tool was that recessions did not last long, and by the time discretionary fiscal measures were implemented, it was typically too late.  Martin Feldstein made the point that some recessions, in particular those associated with financial crises, are long enough that discretionary policy can and should be used.  He noted however that fiscal activism need not mean changes in the overall budget deficit or surplus, but may come from changes in the composition of the budget, for example, an increase in the investment tax credit financed by an increase in corporate taxation.

Interestingly, despite questions by the chair of the session, Vitor Gaspar, on potential improvements in the design of automatic stabilizers (based on the very interesting chapter 2 of the April 2015 Fiscal Monitor), the issue did not register, and I am still struck by the lack of action on this front.  The likely reason is that, for the time being, the focus on most governments is on debt reduction, on the right speed of fiscal consolidation.   Even there however, there has been little work, and even less action on the design of fiscal rules.  In that respect, an interesting remark by Marco Buti was that the excessive number and complexity of the European Union rules reflects in part the relative weakness of the European commission in making sure countries actually implement them.

9. The complex effects of capital flows 

What struck me most in the discussion of capital flows was the recognition that they have complex effects, a theme which resonated with the recent Mundell Fleming lecture by Helene Rey. Globalization and the increase in countries’ gross asset and liability positions, imply that shifts in flows can be very large.  As all three speakers in the session, Agustín Carstens, Maury Obstfeld, Luiz Pereira da Silva—as well as Raghuram Rajan in the final panel discussion, —argued, the effects of these shifts go far beyond changes in exchange rates: they affect the domestic financial system, sometimes for the better, sometimes for the worst.

This conclusion has one clear implication, namely that hands-off policies are not the solution. I see this as one area where the rethinking has been striking, say compared to ten years ago. There was wide agreement that macro prudential tools can decrease the adverse effects on the banking system. There was also agreement that FX intervention can be a useful tool, at least to stabilize the exchange rate. Pereira da Silva argued for a pragmatic approach in Brazil centered on these. But there was no agreement on capital controls. Carstens was against, arguing that, at least for Mexico, which is highly integrated with the United States, capital controls have more costs than benefits and should be avoided. To me, capital controls are not fundamentally different from macro prudential tools; the rules for investors have to be clear, both ex ante and ex post.

10. How much can the international monetary system be improved?

Turning to the design of the international monetary system, Ricardo Caballero touched upon the demand for safe assets by emerging market countries, and argued for a better provision of international liquidity by the IMF and by central banks.  Not only would this be desirable on its own, but, relating it to the earlier discussion, it would also lead to an increase in the world safe real rate, and alleviate the worries about secular stagnation. Maury Obstfeld argued that no exchange rate arrangement was perfect, but that managed float was still probably the best arrangement for most countries.  Two other main issues came up:

The first was how to square the national mandate of central banks with the fact that monetary policies can have large spillover effects on other countries, potentially leading to a bad global outcome.   Jaime Caruana argued for "enlightened self interest," based on the notion that originating countries should, in their own interest, take into account the “spillbacks” of their policies.  This self interest argument however only goes so far: spillbacks may be limited, and taking them into account may still not induce behavior consistent with the best global outcome.

The second issue was closely related, and had to do with the exact nature of the spillovers.  Put in a more formal way, is the Nash equilibrium—the equilibrium which arises when each central bank tries to fulfill its national mandate given other central banks’ actions—suboptimal relative to a cooperative equilibrium. Zeti Akhtar Aziz argued that these spillovers were poorly understood.  Better understanding, she argued, could lead to less disagreements and frictions among advanced economies and emerging markets.

The bottom line 

Not all the questions I had raised in my preconference blog were taken up, and few were settled. Still, to go back to the title of the conference, "Rethinking Macro Policy III:  Progress or Confusion?" my answer is definitely “both.”  Progress is undeniable. Confusion is unavoidable, given the complex issues that remain to be settled.  (My own intervention, on this theme, is available here).