«Food Price PaSS-through in the euro area the roLe oF aSymmetrieS and non-LinearitieS by Gianluigi Ferrucci Rebeca Jiménez-Rodríguez and Luca ...»
Wo r k i n g Pa P e r S e r i e S
no 1168 / aPriL 2010
the euro area
the roLe oF
by Gianluigi Ferrucci
and Luca Onorante
WO R K I N G PA P E R S E R I E S
N O 116 8 / A P R I L 2010
FOOD PRICE PASS-THROUGH
IN THE EURO AREA
THE ROLE OF ASYMMETRIES
AND NON-LINEARITIES 1by Gianluigi Ferrucci 2, Rebeca Jiménez-Rodríguez 3 and Luca Onorante 4 NOTE: This Working Paper should not be reported as representing the views of the European Central Bank (ECB).
The views expressed are those of the authors and do not necessarily reflect those of the ECB.
In 2010 all ECB publications feature a motif taken from the €500 banknote.
This paper can be downloaded without charge from http://www.ecb.europa.eu or from the Social Science Research Network electronic library at http://ssrn.com/abstract_id=1578209.
1 The views expressed in this paper are those of the authors and do not necessarily reflect those of the European Central Bank, its Governing Council or the Eurosystem. The authors wish to thank Hans-Joachim Klöckers, Angela Maddaloni, José Marín Arcas, Aidan Meyler, Julian Morgan, an anonymous referee and participants at an internal ECB seminar for their valuable comments, suggestions and discussions. All remaining errors are the sole responsibility of the authors.
2 Corresponding author: European Central Bank, Kaiserstrasse 29, D-60311, Frankfurt am Main, Germany; Tel.: +49 69 1344 5714;
Fax: +49 69 1344 7602; e-mail: email@example.com 3 Department of Economics, University of Salamanca, Campus Miguel de Una
Address Kaiserstrasse 29 60311 Frankfurt am Main, Germany Postal address Postfach 16 03 19 60066 Frankfurt am Main, Germany Telephone +49 69 1344 0 Internet http://www.ecb.europa.eu Fax +49 69 1344 6000 All rights reserved.
Any reproduction, publication and reprint in the form of a different publication, whether printed or produced electronically, in whole or in part, is permitted only with the explicit written authorisation of the ECB or the authors.
Information on all of the papers published in the ECB Working Paper Series can be found on the ECB’s website, http://www.
Keywords: food commodity prices; inflation; non-linearities; pass-through.
JEL classification: C32; C53; E3; Q17.
ECB Working Paper Series No 1168 4 April 2010 Non-technical summary This paper analyses the transmission of commodity price shocks along the food price chain in the euro area. Conventional wisdom holds that increases in commodity prices are passed through, at least partially, to retail prices. Yet, formal statistical tests typically struggle to find a robust food price pass-through in the euro area. What explains this puzzle? One hypothesis is that the existing studies look at the wrong food commodity data: the international commodity prices that are at the heart of most empirical investigations are a poor approximation for the commodity cost pressures faced by euro area producers. This is because they do not account for the distortions induced by the Common Agricultural Policy in Europe. A second hypothesis is that these studies generally neglect that the pass-through may be non-linear and may depend on the sign, size and volatility of the impulse. These effects have been shown to matter for the transmission of oil price shocks to both real and nominal variables, but so far they have been neglected in the context of food prices.
We investigate these hypotheses using a novel database of farm-gate and internal market prices of food commodities collected in the European Union. These prices take implicitly into account the presence of the Common Agricultural Policy in Europe. We also model several departures from the linear pass-through benchmark and compare alternative specifications with aggregate and disaggregate food data.
Our investigation highlights a number of interesting results. Contrary to the existing literature, we find evidence of a statistically and economically significant food price pass-through in the euro area, provided that the Common Agricultural Policy is put into the picture. We also find that asymmetries and non-linearities are statistically and economically significant, and hence have to be accounted for in order to precisely measure the impact of a commodity price shock on consumer prices. A model-based decomposition of the factors driving the rise in food prices at the consumer level between 2006 and mid-2008 shows that commodity prices were a key determinant of the increase in food consumer prices, thus solving the pass-through puzzle highlighted in the literature. Finally, the disaggregate approach highlights important differences in the structure of pass-through for the various food items, which are lost when aggregate indices are used.
A few implications of our findings for the monitoring, modelling and forecasting of food prices in the euro area are worth mentioning. The Common Agricultural Policy plays an important role in the transmission mechanism of food price shocks in the euro area and the novel database adopted in this paper may provide valuable information for the assessment of near-term food price developments. Moreover, models of pass-through for the euro area should be preferably estimated at a disaggregated level and should ideally allow for non-linear pass-through.
It is widely believed that commodity price shocks pass through, at least partially, to final consumer prices. Policymakers and economic observers, for example, have relied on this argument to explain the surge in retail food prices observed in many developed and developing economies between 2006 and mid-2008, as the opening quotes indicate. However, somewhat in defiance of these accounts, a number of empirical investigations have struggled to find an economically and statistically significant pass-through between international food commodity prices and final consumer prices for the euro area (see IMF, 2008a; Benalal et al, 2004; Chauvin and Devulder, 2008).4 By contrast, there is evidence of a robust pass-through between the two variables for other major economies outside the euro area (see IMF, 2008b).
What explains the low estimates of food price pass-through in the euro area and why are formal empirical investigations at odds with common sense? This paper attempts to shed some light on the issues. Improving the understanding of the dynamic relationship between commodity prices and the retail prices of food products has clear benefits for inflation forecasting and the implementation of monetary policy, as food represents almost one-fifth of euro area consumption and food prices have been a key driver of the sharp rises and falls in headline inflation in recent years. Energy prices have also been a major influence on overall inflation over the period, but with half the weight of food, they have attracted undoubtedly much more attention, both 1 “Outstanding Issues in the Analysis of Inflation”, Speech by Ben S. Bernanke for the Federal Reserve Bank of Boston's 53rd Annual Economic Conference, Chatham, Massachusetts, June 9, 2008.
2 See the Editorial of the September 2008 issue of the Monthly Bulletin of the European Central Bank.
3 “Fears over food price inflation” by Jenny Wiggins, Financial Times, London, 23 May, 2007.
4 In Benalal et al (2004), food commodity prices appear as a determinant in the processed food price equation for the euro area, but only with a long lag of six months and a relatively small coefficient, implying an economically trivial impact on the dependent variable. The equation for unprocessed food prices does not include food commodity prices as a determinant. In Chauvin and Devulder (2008), processed food prices do not depend on food commodity prices. They depend on non-oil import prices, but only with a lag of two quarters and with an economically miniscule impact, relative to unit labour cost, which is the main explanatory factor. As in Benalal et al (2004), the unprocessed food price equation does not depend on food commodity prices, nor does it incorporate import prices.
ECB Working Paper Series No 1168 6 April 2010 in the literature and in the economic profession. Furthermore, some of the factors that have underpinned the rises and falls in commodity prices in recent years are still in place (structural shift in food demand due to rising income in developing countries, demographic growth) or can quickly materialize again (bad weather, bad crops) and while the recent peaks may not represent a new norm, the volatility of commodity prices may be expected to remain high in international markets going forward.
Looking at the pass-through mechanism in this context comes clearly at a premium.
Two hypotheses may help explain why the empirical literature has been unable to capture the full extent of the food price pass-through in the euro area. One is that the international prices that are typically at the heart of the existing studies of passthrough may not be the right gauge of commodity cost pressures in the euro area.
Being quoted in global markets, such prices do not account for the presence of the Common Agricultural Policy (CAP) in the European Union (EU). Certain features of the CAP introduce price distortions for an important range of food agricultural products of which the EU is a large producer, resulting in a disconnection between the prices in the EU and those quoted in international markets. For these products, international prices may provide a poor guide to the true input cost pressures faced by producers and retailers along the food production chain. They may give a wrong signal when tested in formal models of pass-through for the euro area.
A second hypothesis is that the pass-through may be non-linear. It may depend, for example, on the sign, size and volatility of the impulse. Such effects have been shown to matter for the transmission of oil price shocks to both real and nominal variables (see, for example, Hamilton, 1996; and Jiménez-Rodríguez and Sánchez, 2005), but so far they have been neglected in the context of food prices.
In line with these hypotheses, this paper contributes to the literature in two important dimensions. First, it uses a recent database of farm-gate and internal market prices for food commodities collected in the EU, which take implicitly into account the role of the CAP. The database, explained in further detail in Section 2, is produced by the European Commission (EC) and contains detailed price information for several food agricultural commodities produced directly in the EU. In comparison with the prices of internationally traded food commodities normally used in previous studies, the Commission’s database collects the prices of food items observed inside the EU, thereby capturing the presence of the CAP, which has the effect of shielding EU agricultural commodity prices from developments abroad.
Second, it employs econometric techniques aimed at capturing possible nonlinearities in the transmission of food price shocks. We take this possibility into account by comparing a linear model of pass-through with models including various non-linear transformations of the prices of food commodities, which have been successfully used to link oil prices and real activity in the past (see, for example, Bernanke et al 1997; Hamilton and Herrera, 2004; and Jiménez-Rodríguez and Sánchez, 2005; among others).
An additional contribution of the paper is that it compares alternative model specifications with aggregate and disaggregate food data. There is a long tradition for assessing the relative merits of the two approaches in economic modelling, going back to the works of Theil (1954) and Grunfeld and Griliches (1960). More recent attempts include Hubrich (2005) and Benalal et al (2004), which compare the two approaches in the context of forecasting euro area inflation. This literature highlights several rationales for using aggregate indices of disaggregate variables, instead of the aggregate variable of interest directly. One is that the disaggregate approach allows a more flexible modelling of the idiosyncratic properties of the data, for example by using different dynamic structures and information sets for the various food components. A second rationale is that it allows to measure individual pass-through patterns for different commodities and to analyse the food items that are more
2. Link between food commodity prices and inflation The dichotomy between flexible commodity prices and sticky industrial and retail prices lies at the heart of most formal accounts of pass-through. Commodity prices, ECB Working Paper Series No 1168 8 April 2010 which are set in competitive, flexible markets, respond immediately to general macroeconomic news, whereas intermediate and final consumer prices, which are set contractually by producers and retailers, take more time to react. Because commodity prices are more flexible, they can generally be expected to lead the adjustment along the price chain, regardless of the source of the initial shock. For example, a cost-push shock that originates in commodity markets and that is transmitted through the production chain will only affect final selling prices with a lag. Likewise, the first signs of a demand-pull shock might be visible in commodity markets, and affect final good markets only with a delay (see Blomberg and Harris, 1995; and Furlong and Ingenito, 1996).
Building on this simple intuition, the theoretical literature has described several channels through which movements in commodity prices provide early warning signals about inflation. Models of cost mark-up pricing á la Kalecki (1971), for example, emphasise the role of idiosyncratic shocks that originate in the markets for certain commodities, such as a drought that reduces the supply of a crop pushing up prices. To the extent that these commodities are used as intermediate inputs for the production of final consumption goods, such shocks may be eventually transmitted to intermediate and final prices, if they are not absorbed in profit margins or through advances in factor productivity (see Bloch et al, 2004).