Re-estimated FR-BDF: New Features and an Assessment of Monetary Policy Tightening in France

Published on 13th of May 2026

Authors : Ugo Dubois, Bruno Ducoudré, Raphaël Martin, Anna Petronevich, Caterina Seghini, Camille Thubin, Harri Turunen

Working Paper Series no. 1044. This paper presents the updated and re-estimated version of the Banque de France’s semi-structural FR-BDF model, the institution’s core framework for quarterly projections and policy scenario analysis. The update reflects major statistical and structural shifts, including INSEE’s transition to a 2020 base year and the volatility during the COVID-19 period. The re-estimated model features a lower capital share, a weaker underlying productivity trend, a steeper price Phillips curve along with a flatter wage Phillips curve, while adding a credit block and an enhanced consumption equation that captures short-run income effects. The updated FR-BDF delivers stronger and faster real effects of monetary policy — with larger GDP and unemployment responses due to amplified transmission through long-term rates and the exchange rate — while nominal responses are more muted. Using the updated model, the recent monetary tightening is assessed to lower VA inflation by –0.6 pp and reduce GDP growth by 1 pp (at trough) under backward-looking expectations, whereas under forward-looking expectations the responses are more front-loaded and nearly three times larger. Under the more realistic hybrid expectations mode, VA inflation falls by about 1.2 percentage points and GDP growth by roughly 2 percentage points at the trough. These magnitudes remain broadly consistent with benchmark assessments reported in the literature.

Responses to 2022-2025 monetary policy tightening under different types of expectations

WP1044 — Note: The figure shows the responses of real GDP (left) and VA annualized price inflation (right) to a monetary policy tightening in the euro area as observed during 2022-2025, under different expectation regimes. The hybrid expectation mode is defined as a combination of two other modes, where the financial markets are forward-looking while the rest of the economic agents are backward-looking. Importantly, in the hybrid mode, the information about the monetary policy shock is diffused progressively (contrary to the forward-looking expectation mode, where the agents are fully perfect foresight) and so the expectations are updated as the new information arrives (staggered expectations). For reference, the Banque de France’s quarterly forecasting framework is based on backward-looking expectations.

Non-Technical Summary

Semi-structural macroeconomic models are key tools for central banks when producing economic projections and analysing policy scenarios. At the Banque de France, the FR-BDF model plays a central role in forecasting and in assessing the effects of economic shocks and policy measures on the French economy. However, such models need to be regularly updated to remain reliable. Since the last version of FR-BDF was released in 2019, the macroeconomic environment has changed significantly. The COVID-19 pandemic, the surge in inflation, and the sharp tightening of monetary policy have altered economic dynamics. In addition, the transition of French national accounts to a new statistical base year has modified the measurement and historical coverage of several macroeconomic series. Together, these developments call for a comprehensive update of the model.

This paper presents the revised version of the FR-BDF model used at the Banque de France. The update introduces several improvements designed to better capture recent developments in the French economy. In particular, the model now incorporates richer financial channels linking interest rates, credit conditions and balance sheets; a more detailed treatment of energy prices; and a stronger short-term link between household income and consumption. The model has also been fully re-estimated using the latest statistical data, while specific adjustments allow it to account for the exceptional economic fluctuations observed during the pandemic period. The revised model also serves as the structural foundation for the FR-BMEs forecasting tool currently used in the Bank of France interim projection exercises.

To assess how these changes affect the behaviour of the model, the paper compares the responses of key macroeconomic variables to a set of standard shocks—such as interest rate, foreign demand and fiscal shocks—with those obtained in the 2019 version. These comparisons show how the revised specifications modify the propagation of shocks in the model and help identify the mechanisms behind the changes in its macroeconomic responses.

Using the updated model, the paper evaluates the macroeconomic effects of the recent tightening of monetary policy. The results indicate that higher interest rates have sizeable effects on economic activity and inflation in France. However, the magnitude and timing of these effects depend strongly on how households, firms and financial markets form expectations about future interest rates. When economic agents react quickly to expected policy changes, the contraction in output and inflation is sharper and more front-loaded. When expectations adjust more gradually, the effects are smoother but remain persistent. Overall, the results confirm the importance of expectation formation in shaping the transmission of monetary policy. To better reflect real-world behaviour, the paper provides the evaluation of the effects in a hybrid framework that combines forward-looking financial markets with more gradual adjustments by households and firms and features progressive updating of expectations.

Keywords: Semi-structural Modeling, Expectations, Monetary Policy

Codes JEL : C54, E37

Download the full publication

Re-estimated FR-BDF: New Features and an Assessment of Monetary Policy... (PDF - 2.86 MB)

Updated on the 13th of May 2026