Remarks by Kristalina Georgieva, IMF Managing Director at the Bank for International Settlement
Dear Fabio, dear Pablo, dear governors—thank you for inviting me here to share a few thoughts on the challenges faced by central banks as our stewards of monetary policy.
Over the past 20 years, we have seen the Great Moderation give way to the global financial crisis, followed by a long spell at the effective lower bound, followed by a succession of large shocks, from the pandemic to the war in the Middle East.
And we have seen the unstoppable march of technology, not at an even tempo but, as ever, in discrete jumps—from robotics to stablecoins to AI bots trading 24/7.
Each recent shock has had profound implications for fiscal and monetary policy, making policymaking harder, touching all three core central banking tasks: price stability, financial stability, and payment systems.
Today I will focus on what in many jurisdictions is top of mind for businesses and the general public—price stability—addressing three points: first, demand management in the face of the energy shock; second, what AI means for monetary policy; and third, the pressure on long-term interest rates arising from market concerns about the mix of stubborn inflation and high public debt.
Let me start with the Strait of Hormuz, noting upfront that—rather unusually for me—I am less optimistic than consensus on the size and duration of the supply shock its effective closure has caused.
It is true that six months into the war the global economy has not fallen off a cliff. Oil did not reach $150 per barrel and, despite inflated fuel prices and hardships in several Gulf countries, we remain on track for world growth of about 3 percent this year.
To explain this resilience, let me recognize a few of the countries represented in this forum, without trying to be comprehensive: Saudi Arabia and the UAE for rerouting oil around Hormuz; the U.S. and Norway for ramping up oil and gas exports; China for curtailing its oil imports; Nigeria and India for their refining; and all International Energy Agency members for their coordinated reserve releases.
These steps and others, including gas-to-coal switching in some countries, diversification away from hydrocarbons in many, and demand reduction in all, have helped us get through this shock—so far.
But, to quote the Game of Thrones, winter is coming. Oil, gas, and petroleum product prices remain high—just last Friday, U.S. diesel prices broke a new record. Shipping through Hormuz is at about one-tenth of its pre-war level. Reserve drawdowns will hit their limits. Depleted reserves will need restocking. And now we have AI as the thousand-pound gorilla in the room, with its voracious appetite for energy.
As central bankers, you are in the business of demand management. One key question you confront is how restrictive should policy be if the energy outlook were to worsen? The answer depends on your assessment of the risk that a price-level shift could trigger second-round effects and a broader inflation process.
To make that assessment, you look to the data: output and capacity utilization; job market indicators; financial conditions; underlying inflation and inflation expectations; and much more. And, despite all the data, we know that reading the tea leaves remains as much art as science.
At the IMF, our reading is not only that the price-level shift is less extreme than in the gas supply shock experienced by Europe in 2022, but also that there may have been more slack going in, which would limit the risk of the shock propagating.
But again, let me share my worry that this global ordeal may be far from over.
With that in mind, it is comforting to have watched many emerging markets come of age in recent years in terms of central bank independence and the sophistication of their policy frameworks. Here, in my first slide of two, we see this progress.
Next, let me turn to AI and what it means for monetary policy—as a positive demand shock today and a positive supply shock tomorrow, rolled into one.
So, here is my hypothesis: that AI is likely to add to inflation pressures in the short run while its longer-run effects are unclear—views are split on whether it may be deflationary or not.
With large upfront investment in AI infrastructure—reversing a decades-long progression toward lower capital and energy intensity—and with rising equity prices boosting household wealth, a positive demand effect likely dominates in the short run—although this could go into reverse if there were to be a major market correction.
But there is also a positive supply effect, with productivity expected to increase over time. As our new chief economist Silvana Tenreyro argues in a recent Bank of England paper titled “Productivity and Inflation Dynamics,” higher productivity boosts both supply and demand, the latter through its effects on real income.
Whether the outcome is inflationary or deflationary depends on the magnitudes of the two effects and their timing. With AI expected to deliver a lasting uplift to productivity growth, Silvana’s logic suggests it will also raise expected permanent real income, which could lift demand in advance of expanding supply, pushing up the neutral interest rate, r*, and inflation—the leads and lags are crucial.
Ken Rogoff goes a step further, injecting into the argument the demand effects of a changing income distribution. His logic is that if AI hollows out the middle of the job market—which our own research also flags as a risk—and if AI retains its winner-takes-all dynamic, the boost to demand could be muted, possibly to a point where the positive supply effect dominates.
For now, we can safely conclude that the question of whether the recent bout of high inflation would be followed by a return to ultra-low inflation and interest rates has been answered. In the short run, no.
And as for the longer run, yes, there is a possibility that—through the expected uplift in productivity—AI could reduce costs and be deflationary. But in how long is this long run? And how confident are we that we will get the right combination of productivity gains and demand responses?
And one last thought on AI, more for the fiscal authorities than for central bankers: if AI is going to hollow out the middle of the job market, we had better start considering the future of personal income taxation.
That last concern takes me from the gorilla to the elephant in the room: public debt.
The recent years have been defined by vast shocks—with the pandemic in particular necessitating unprecedentedly large fiscal support to households and firms, notably in advanced economies, followed by further rounds of fiscal support during the cost-of-living crisis, especially in Europe.
Regrettably, in far too many countries, post-crisis recovery has not been paired with consolidation. The result is what I call the “stairway-not-to-heaven”: big upward jumps in debt when shocks strike, little or no reduction afterward. Rinse and repeat.
Even after the recent inflation surprise helped lower some debt ratios, the crisis legacy remains a global public debt burden that is at its heaviest since World War II—and on track to exceed 100 percent of GDP within the decade—with many of the highest ratios seen in advanced economies.
At the same time, inflation remains stubborn. In U.S. for instance, inflation has been above target for 5½ years—and we at the IMF have pushed back in our projections the date when we expect it to return to target by no less than two years. In the U.S. and elsewhere, markets judge the recent rate-cutting cycles to be over.
Today, most major advanced economies have public debt paths that call for fiscal policy attention. All of us are aware that U.S., French, and Japanese 10‑year sovereign yields, to highlight just three, are currently at their highest levels since 2007, 2008, and 1996, respectively.
And as these benchmark borrowing costs rise, they lift most of the world’s yield curves up with them—in some emerging markets this more than fully offsets hard-won spread compression.
Core yields are climbing despite central bank bondholdings that keep material amounts of debt and duration off market. Today, even after significant shrinkage, the Bank of Japan still owns 34 percent of own-government debt, the Eurosystem 20 percent, and the Fed 11 percent—as shown here, in my second slide.
Many governments now face high and rising debt-service burdens. Heavy reliance on short-term bills increases gross financing needs—to close to 40 percent of GDP in the U.S. Treasury’s case. This amplifies sensitivity to policy rates and raises the cost of falling behind the curve, failing to control inflation, and then having to double down with higher rates.
It is worrisome that fiscal consolidation seems not to have received the attention it deserves. In Japan, the government has ambitious budget plans to lift potential growth and ease the pain of rising prices. In Europe, significant spending is being accommodated by escape clauses in the fiscal rules. In the U.S., the debt path draws insufficient Congressional attention.
With rising bond yields, the difference between real interest rates and real growth, r – g, is now less supportive than before. For growth to painlessly solve the fiscal problem, large and sustained increases would be needed.
These are the reasons why I chose to use my recent remarks at Jackson Hole to warn about the risk of fiscal dominance. When public debt is high, history has taught us that fears of lower-than-optimal policy rates or future monetization may lurk in the background, pushing bond yields and inflation expectations upward. And, with many central banks now making losses from a mix of low yields on legacy assets and high interest expenses on bank reserves, the challenge for central banks is even greater.
With high debt and fears of fiscal dominance conspiring to increase the risk of inflation expectations moving upward, central banks will need to respond forcefully to future shocks to protect independence, preserve credibility, and deliver on their price stability mandate.
At the Fund, we take very seriously our duty to call out the fiscal risks and create traction for consolidation—this is the underlying problem that needs to be solved. Our annual Article IV consultations with the systemic economies have consistently flagged the fiscal issue and will continue to do so.
But until such time as credible medium-term fiscal plans are put in place, it is important to stand firm with a rock-solid commitment to that most critical purpose of central banking: to ensure inflation is low and stable.
Let me close by proposing for discussion five core features for monetary policy success in these unsettled times:
First, vigilance. When shocks interact, our models may lose some predictive power, calling for especially thorough scrutiny of the data.
Second, agility. Conditions change, and what works in one state of the world may not work in another, so policies must stand ready to pivot and adapt.
Third, communication. Markets need basic guidance on how central banks think, but equally central banks must take care to guard their policy optionality.
Fourth, credibility. The framework that underpins credibility—central bank independence and a clear nominal anchor—must be protected, and keeping inflation anchored when public debt is high may dictate a hawkish bias.
Fifth, humility. This, last but by no means least, is a key ingredient given the speed and magnitude of the changes that central banks confront today, requiring the ability to correct course with no remorse.
Having these five points in mind may come in handy, especially in tough times—times like, for instance, when you are called upon to distinguish between market discipline from orderly increases in bond yields and sudden spikes indicative of vigilantes and malfunction—situations that can necessitate decisive liquidity provision.
Thank you.
Compliments of the International Monetary Fund The post IMF | Monetary Policy with Supply Shocks and High Debt first appeared on European American Chamber of Commerce New York [EACCNY] | Your Partner for Transatlantic Business Resources.