The unprecedented surge in prices since 2021 led the ECB to embark on the fastest tightening cycle in the euro area’s history. This column analyses the lags and strength of monetary policy transmission to output and inflation over the two decades in the euro area. It shows that the peak effects on real GDP and headline inflation are observed within 12-18 months of the interest rate rise. The impact on inflation has been stronger and more persistent in the recent tightening cycle, reflecting the forceful monetary policy response and the higher flexibility of prices.
The Covid-19 pandemic-induced supply chain disruptions and the Russian invasion of Ukraine led to an abrupt shift to a high-inflation environment in the euro area after nearly a decade of below-target inflation. Furthermore, this shift led to a return to conventional interest rate-setting as the main monetary policy tool following the deployment of several unconventional monetary policy measures during the effective lower bound era (ELB). Given the prominent role of the strength of monetary policy transmission in the ECB’s reaction function, adopted during the recent tightening campaign, uncertainty regarding transmission lags has been front and centre in policy discussions (e.g. Lane 2024).
Debate surrounding the lags and strength of monetary policy pass-through to aggregate output and prices dates back to the famous dictum of Friedman (1961). However, ‘long and variable lags’ have been questioned in the recent literature (e.g. Buda et al. 2023, Grigoli and Sandri 2023), suggesting that monetary policy can affect economic activity within months. Price formation, however, appears to respond more sluggishly to monetary policy announcements. Allayioti et al. (2024) and Aruoba and Drechsel (2024) show that it takes six to 12 quarters for a monetary policy shock to fully transmit to prices in the euro area and the US, respectively.
Lags are variable
In a recent paper (Zlobins 2025), I document the transmission lags to output and inflation in the euro area jointly and assess how sensitive they are to different modelling choices.
Evidence from a linear Bayesian structural vector autoregression (SVAR) in Figure 1 suggests that peak effects on real GDP and headline Harmonised Index of Consumer Prices (HICP) inflation are observed within 12-18 months after the interest rates rise. This result is robust to a wide array of stability tests: frequency of time series used in the estimation of models (monthly/quarterly), identification strategy of monetary policy shock (Cholesky decomposition/sign and zero restrictions/multiple high frequency monetary policy shocks, such as Altavilla et al. 2019, Jarocínski and Karadi 2020), alternative estimator of impulse response functions (local projections), and when controls for omitted variable bias are introduced in the model. While in some cases a sizeable pass-through of monetary policy to real activity can be detected in the short run, the impulse responses always remain hump-shaped, thus favouring the ‘long and variable’ hypothesis.
Figure 1 Monetary policy pass-through to output and headline inflation
Note: Figures show impulse response functions from a Bayesian SVAR to a conventional monetary policy shock, identified via mixture of high frequency information with narrative sign restrictions as in Grüning and Zlobins (2023). The shock has been normalised to generate a five basis point increase in the three-month EURIBOR. The solid line shows the median response while the dashed region denotes the 68% credible sets.
Speaking of the ‘variable’ part of this hypothesis, a medium-scale SVAR – deployed to account for the omitted variable bias – makes it possible to pin down transmission lags to a wider set of economic indicators, including different HICP items. A recent surge in inflation has been characterised by strong and very persistent price pressures in the services sector, causing serious concern among policymakers (e.g. Schnabel 2023).
Figure 2 Monetary policy pass-through to core and services inflation
Note: Figures show impulse response functions from a medium-scale Bayesian SVAR to a conventional monetary policy shock. The shock has been normalised to generate a five basis point increase in the three-month EURIBOR. The solid line shows the median response while the dashed region denotes the 68% credible sets.
Evidence in Figure 2 suggests that transmission lags to core and services inflation are substantially longer when compared to headline inflation as the peak impact takes more than two years to materialise. Thus, the bulk of the impact from monetary tightening of 2022 and 2023 is still in the pipeline and will contribute to the softening of services prices in the coming years, minimising the risks of sticky services inflation for medium-term price stability.
Is this time different?
The time-varying aspect of the ECB’s monetary policy is particularly important since the euro area in the last two decades has undergone structural changes (services deepening) and confronted large shocks (the Great Recession, the Covid-19 pandemic, and the energy cost crisis due to Russia’s invasion of Ukraine). These developments have had ambiguous consequences for the slope of the Phillips curve, as well as the regime change in monetary policy itself, as it relied on a suite of non-standard measures for the better part of the last decade due to the effective lower bound, resulting in a build-up of excess liquidity.
Figure 3 Time-varying effects of monetary policy
Note: Figures show impulse response functions from the TVP-SVAR-SV to a conventional monetary policy shock over the period from Q3 2002 to Q3 2023. The shock has been normalised to a five basis point increase in the three-month EURIBOR in each period, allowing the estimated elasticities to be comparable over time. The y axis shows the year of the impulse response function, the z axis shows the impact on the macroeconomic indicators and is expressed in percentage points, while the x axis shows the number of quarters since the shock.
In order to pin down potential changes in the monetary transmission mechanism in the euro area, the SVAR is extended to feature time-variation in both the parameter space and shock volatilities (TVP-SVAR-SV). Non-linear estimates in Figure 3 indicate that the response of output to monetary policy actions has been broadly stable over time, with a slight decrease towards the end of the sample. More importantly, the impact on inflation has been much stronger and more persistent in the recent tightening cycle, implying that the trade-off between output and inflation stabilisation – sacrifice ratio of monetary policy – has been historically low. This is partly due to a much higher persistence of monetary policy shocks in the post-pandemic period, as demonstrated by the reaction of the three-month EURIBOR, reflecting a forceful response of the ECB to the inflation surge.
Why is this time different?
These empirical observations are then rationalised using a medium-scale dynamic stochastic general equilibrium (DSGE) framework, containing most of the relevant nominal and real rigidities. Simulations via the structural model point out two ingredients which have contributed to the stabilisation properties of monetary policy in the recent tightening cycle. First, a post-pandemic inflation surge has been marked by a substantial increase in the repricing frequency (Cavallo et al. 2023, Montag and Vallenas 2023, Dedola et al. 2024), implying an upward shift in the slope of the Phillips curve. Ceteris paribus, the steeper the Phillips curve, the lower the sacrifice ratio of monetary policy stabilisation, since more flexible price-setting lessens the level of nominal rigidity in the economy and thus dampens the real effects of monetary policy. Second, a forceful and persistent monetary policy response (also as evidenced by empirical estimates in Figure 3) to the inflation surge contained an upside de-anchoring of inflation expectations, preventing the incorporation of second-round effects into prices by firms. At the same time, frictions related to wage-setting or real rigidities have likely only had minor implications for the effectiveness of recent policy rate hikes.
Conclusion
The main lesson provided by this inflationary episode is that large supply-side related disturbances can limit the ability to ‘look through’ them, as conventional wisdom would suggest. Instead, central banks can leverage the lower sacrifice ratio during such inflation surges and act forcefully to keep inflation expectations anchored to the target.
Source : VOXeu
