What impact will higher interest rates have on French economic growth? An overview of macroeconometric models
Informations
The year 2022 was marked by a very strong surge in inflation in the United States and the eurozone. At the end of October, the year-on-year inflation rate reached 7.7% in the United States, 10.6% in the eurozone and 7.1% in France, i.e. between 5 and 8 points above the inflation targets of the US Federal Reserve (Fed) and the European Central Bank. In response, the two central banks tightened monetary policy on a massive scale. The Fed’s key interest rate rose from 0% in March 2022 to 4% in November 2022. While the increase in the ECB’s key rate has been more measured for the time being, long rates on public debt in European countries have risen sharply, gaining between 250 and 300 basis points in one year in France and Germany, and even more in eurozone countries where the risk on public debt is perceived to be higher. This rise is close to that expected for short rates in 2023. OFCE forecasts that the ECB’s key rate will reach 3% in the third quarter of 20231.
1 See in the OFCE forecast table 2 of appendix 1 of the part Tour du monde de la situation conjoncturelle, Département Analyses et Prévisions, under the direction of E. Heyer and X. Timbeau.
Estimating the impact of this tightening on economic activity is difficult. The transmission of a monetary shock to the rest of the economy is the subject of a very rich literature using methods that are conceptually similar, or even equivalent, but whose results can vary greatly in practice. We are particularly interested here in the impact of an interest rate shock in macroeconometric models of the French economy. For this overview, we use three models: the Mésange model co-developed by DG Trésor and INSEE (see Bardaji et al., 2017), the FR BDF model of the Banque de France (see Lemoine et al. 2019 and Aldama and Ouvrard 2020 for the variants) and the specification of the OFCE emod model used in Heyer and Timbeau (2006).
What is a macroeconometric model?
Macroeconometric models represent the oldest class of macroeconomic models. They combine accounting relationships (or equations) and estimated behavioural equations to make predictions about the economy’s response to shocks. The major macroeconomic variables (wages, prices, household consumption, investment, employment) are expressed in the form of error-correction equations. In the long term, they converge towards a certain target, determined by economic theory. Thus household consumption expenditure will converge over the long term towards a certain fraction of household disposable income. On the other hand, short-term behaviour is left much more free in order to achieve good forecasting performance. The interest rate is essentially a long-term factor. The impact of an interest rate shock is initially limited and becomes more significant as the gap between the variables and their long-term targets closes.
The Mésange model
We consider the variant published in Bardaji et al. (2017). The results are summarised in Table 1. A monetary shock of 100 basis points (or 1%) results in a fall in GDP of 0.2% after one year, 0.8% after three years and 3% in the long term. This fall is explained in particular by the sharp drop in investment: -2.7% after 3 years (-3.4% for the GFCF of non-financial companies) and -5.5% in the long term, but all the components of aggregate demand are negatively affected, including exports, which fall by 3.3% in the long term. Surprisingly, monetary tightening translates into higher prices in the Mésange model. Market value added prices rise by 0.1% after one year, 0.8% after 3 years and by more than 6% in the long term! This rise in prices makes the economy less competitive, which explains the fall in exports. Two transmission channels are at work. The first is the direct negative impact of a rise in interest rates on business investment. In the Mésange model, the demand for capital and therefore investment depends in the long term on the cost of capital. The intuition is that of standard microeconomic theory: companies choose the combination of capital and labour that maximises their profit. An increase in the cost of capital encourages companies to substitute labour for capital and reduces investment. The user cost of capital is made up of the depreciation of capital, the long-term interest rate on government debt and risk premium terms between government bonds and corporate loans, while the long-term elasticity of investment to this user cost is estimated at 0.44. Assuming a capital depreciation rate of 10%, initial nominal rates at 0 and disregarding risk premiums, a 1% rise in the interest rate translates in the long term into a 5% fall in investment. The second channel, which is much less intuitive, plays a key role in this variant and explains in particular the response of prices and exports. An increase in the cost of capital represents an increase in production costs for companies. They pass on these higher costs in their selling prices, resulting in an inflationary effect and reduced competitiveness. This positive effect of a rise in interest rates on prices via the cost of capital channel was recently explored by Portier, Beaudry and Hou (2022). However, it should be emphasised that this effect is difficult to detect using more agnostic empirical methods (unrestricted VAR models, local projections). While positive effects in terms of the impact of a rise in rates on prices are sometimes obtained, the effect more often than not becomes either insignificant or clearly negative over longer horizons (see, for example, Miranda-Agrippino and Ricco, 2021).
Le modèle FR-BDF
Par rapport à *Mésange*, l’une des spécificités importantes de FR BDF est le traitement des anticipations des agents dans le modèle. Cette spécificité explique que deux taux d’intérêt interviennent dans la dynamique du modèle. Le taux d’intérêt de court terme, déterminée par la Banque centrale européenne, affecte les anticipations des agents tandis que le taux d’intérêt de long terme des obligations publiques joue sur la demande de facteurs de production à long terme. L’élasticité de long terme de l’investissement au coût du capital est de 0,5, légèrement supérieure à celle de Mésange. Le modèle n’incorpore pas de relations systématiques entre les taux longs et les taux courts. Pour obtenir l’effet d’un choc de taux dans le modèle, il convient donc d’additionner deux variantes analytiques distinctes, la première simulant l’impact d’une hausse permanente du taux court, la seconde simulant l’impact d’une hausse du taux long. Ces deux variantes sont disponibles dans Aldama et Ouvrard (2020). Les effets d’un choc de taux sont beaucoup plus faibles que dans Mésange. Au bout de 3 ans, le PIB en volume diminue de 0,3% contre 0,9% dans Mésange. Cela s’explique en particulier par une baisse bien plus faible de la FBCF des entreprises (-1,9% contre -3,4% au bout de 3 ans dans Mésange). Les effets sur les prix sont plus conformes à l’intuition keynesienne habituelle avec une baisse du déflateur du PIB de 0,2% au bout de 3 ans. L’amélioration de la compétitivité qui en résulte permet une hausse des exportations de 0,2% au bout de 3 ans (contre une baisse de 0,2% dans Mésange). Ces différences s’expliquent principalement par deux éléments. Tout d’abord, le canal de transmission du coût du capital vers les prix est neutralisé dans le modèle FR BDF. Si les prix de valeur ajoutée sont déterminés par le coût des facteurs de production et un markup constant comme dans Mésange, le coût du facteur capital qui entre dans l’équation de prix n’est pas le coût d’usage du capital mais le rendement marginal du capital. Ensuite, l’investissement réagit beaucoup moins fortement à court terme à la croissance de la valeur ajoutée dans FR-BDF, et se comporte de manière plus inertielle. Le choc négatif d’investissement se diffuse donc plus lentement.