Christine Lagarde: Monetary policy in an unusual cycle - the risks, the path and the costs

First of all, I would like to welcome you all to this year's ECB Forum.

The theme of the conference is "Monetary policy in an era of transformation", and we have a rich programme ahead of us, exploring the changes that are taking place.

But even if most of us can agree that the economy is undergoing substantial change, I imagine there are more diverging views about where it will end up.

This lack of clarity presents a profound challenge for policymakers, as we must try at once to understand these transformations and to steer the economy through them.

Indeed, much of the policy challenge over the last few years has involved stabilising inflation while facing fundamental uncertainty about the economy.

Nevertheless, we have managed to chart a path through this uncertainty, and we have come a long way in the fight against inflation.

In October 2022, inflation peaked at 10.6%. By September 2023, the last time we raised rates, it had fallen by more than half, to 5.2%. And then after nine months of holding rates steady, we saw inflation halve again to 2.6%, which led us to cut rates for the first time in June.

Our work is not done, and we need to remain vigilant. But this progress allows us to look back and reflect on the path we have taken.

This evening I would like to talk about three specific features that have defined this policy cycle: the risks, the path and the costs.¹

The risks

Let me start with the risks.

In a typical policy cycle, when fluctuations are driven by moderate and short-lived shocks, inflation expectations are usually not at risk. Central banks' price stability mandates and reaction functions ensure confidence in the inflation target.

When faced with typical demand shocks, central banks reach their target by stabilising demand around potential output. And when faced with supply shocks, central banks can in principle "look through" them, as these shocks will usually leave no lasting imprint on inflation.

But this low risk to inflation expectations only applies when shocks are indeed moderate and short-lived. In situations where there is a risk of shocks becoming larger and more persistent, inflation expectations can de-anchor regardless of whether the shocks are demand-led or supply-led.

Central banks must then react forcefully to prevent above-target inflation becoming entrenched.

This was the lesson of the 1970s, when a sequence of supply shocks caused by rising oil prices ultimately morphed into a lasting inflationary shock. And with central banks at the time being seen as ambivalent about bringing down inflation, people revised their expectations about medium-term inflation.

Different studies reach different conclusions about the origin of the current inflation episode. ECB analysis finds that, at the peak, supply shocks were three times more important than demand shocks in explaining the deviation of inflation from its mean.² Other research puts a greater emphasis on demand shocks.³

But this delineation between supply and demand, while relevant, has not been the most important factor in our current cycle.

We needed to base our decisions not only on the source of the shocks, but also on their size and persistence. This was because the shocks were so large and persistent that we faced a genuine risk to inflation expectations.

Two features could have provided fertile ground for people to lose confidence in the monetary anchor.

First, the shocks were large enough to make many households switch their attention to inflation. At the start of 2023, over 60% of respondents in our consumer expectations survey reported that they were paying more attention to inflation than in the past.⁴

Second, the inflationary impact of the shocks risked becoming endogenously persistent, owing mainly to the staggered wage bargaining process in the euro area. Although there is large variation across countries, the average duration of wage contracts is two years, effectively guaranteeing a drawn-out process to "catch up" with past inflation.⁵

We did see some signs that the anchoring of inflation expectations was becoming more vulnerable, especially via a fattening of the "right tail" of the distribution. In October 2022, around four in ten consumers expected medium-term inflation to be at or above 5% and professional forecasters assigned a 30% probability of inflation being at or above 3% two years later.⁶

So, monetary policy had to send a strong signal that permanent overshoots of the inflation target would not be tolerated. As a result, we strongly emphasised our determination to ensure a "timely" return to target. Our aim was to convey our commitment to ensuring that the period of high inflation would be limited and signal a sense of urgency.

The path

But how does monetary policy anchor inflation expectations? It is not only about the policy destination, but also about setting the right trajectory of rates to get there.

This brings me to the second specific feature of this cycle: the rate path.

It was clear from the outset that merely communicating our commitment to reaching our target would not have been enough. ECB analysis shows that, if we had not reacted at all, the risk of de-anchoring would have been above 30% in 2023 and 2024.⁷

It is likely that even moderate policy action would have been insufficient. For example, if rates had stopped at 2%, the risk of de-anchoring would still have been around 24%.

So, when we first started raising rates, we knew that we were far from where we needed to be. The most important factor was therefore to close the gap as quickly as possible. This is why we had a historically steep climb at the start of our rate path, using increments of 75 and 50 basis points for our first six rate increases.

But as policy rates moved towards restrictive territory, the challenge shifted from acting quickly to calibrating the path precisely. In particular, we needed to set a rate path that both delivered a "timely" return to 2% and did so with a high degree of confidence.

This path also required us to take a different approach from the past.

Faced with multiple large shocks, there was significant uncertainty about how to interpret and rank the information we were receiving from the economy.

On the one hand, it would have been risky to rely too much on models trained on historical data, as those data may no longer have been valid. We could not know, for instance, whether shifts in preferences, higher energy prices and geopolitics had changed the structure of the economy.

On the other hand, relying too much on current data might have been equally misleading if they had turned out to have little predictive power for the medium term. As shocks worked their way through the economy, current data could also have reflected lags more than actual inflation trends.

So we constructed a framework to hedge against this uncertainty, blending projections with current data about underlying inflation and monetary transmission. The aim was to combine various pieces of information about the medium-term outlook into a single assessment that could be updated swiftly.

Our forecasts provided a comprehensive assessment of future inflation, assuming the underlying parameters of the economy remained stable. At the same time, looking at current data allowed us to identify the persistent components of inflation and account for structural changes that might have been missing from our forecast models.⁸

In this reaction function, our assessment of the inflation outlook is informed by, but not limited to, our projections. We use various measures to gauge underlying inflation. And when assessing the strength of monetary policy, we consider banks, capital markets and the real economy.

As a result, while the flow of new information constantly adds to and improves our picture of medium-term inflation, we are not pushed around by any specific data point. Data dependence does not mean data point dependence.

This framework helped us navigate the "tightening" and "holding" phases of our policy cycle, and it gave us the confidence to deliver a first rate cut at our last policy meeting.

During these phases, we have seen the "right tail" of the distribution of inflation expectations narrow, consistent with a timely return of inflation to target.

The costs

But while our policy path has helped to tame inflation, it has also dampened economic growth. Interest rates rose steadily and remained high while the economy was stagnating for five straight quarters.

This pattern is unavoidable when central banks face shocks that push inflation and output in opposing directions. But this time, the costs of disinflation have been contained compared with similar episodes in the past.

This brings me to the third specific feature of this cycle.

Given the magnitude of the shock to inflation, a "soft landing" is still not guaranteed. If we look at historical rate cycles since 1970, we can see that when major central banks hiked interest rates while energy prices were high, the costs for the economy were usually quite steep.⁹

Only around 15% of the successful soft landings in this period – defined as avoiding either a recession or a major deterioration of employment – have been achieved following energy price shocks.

But this cycle has so far not followed past patterns.

Inflation peaked at a much higher point than during previous soft landings, but it also decelerated faster. Growth has remained within the range of previous soft landing episodes, albeit near the bottom of that range. And the performance of the labour market has been exceptionally benign.

Employment has grown despite slowing GDP growth, rising by 2.6 million people since the end of 2022. And unemployment is at historical lows for the euro area, and well within the range observed during previous soft landings across major economies.

The resilience of the labour market is itself a reflection of the unusual mix of shocks that have hit the euro area, with labour shortages leading firms to hoard more labour, and higher profits and lower real wages making it easier for them to do so.¹⁰

As a result, the usual propagation from slower growth to heightened unemployment risks and lower demand did not happen to the same extent.

Now, we are still facing several uncertainties regarding future inflation, especially in terms of how the nexus of profits, wages and productivity will evolve and whether the economy will be hit by new supply-side shocks. And it will take time for us to gather sufficient data to be certain that the risks of above-target inflation have passed.

The strong labour market means that we can take time to gather new information, but we also need to be mindful of the fact that the growth outlook remains uncertain. All of this underpins our determination to be data dependent and to take our policy decisions meeting by meeting.

Conclusion

Let me conclude.

Our policy decisions have successfully kept inflation expectations anchored, and inflation is projected to return to 2% in the latter part of next year. Considering the size of the inflation shock, this unwinding is remarkable in many ways.

Even though millions of businesses and workers have been independently striving to protect their profits and incomes, our 2% inflation target has remained credible and has continued to anchor the inflation process.

This speaks to the value of the policy frameworks that central banks have built up over the last 30 years, focusing on price stability and central bank independence. And it is why we will not waver from our commitment to bring inflation back down to our target for the benefit of all Europeans.

As the late footballer and manager Sir Bobby Robson said, "the first 90 minutes are the most important". Similarly, we will not rest until the match is won and inflation is back at 2%.