Food price volatility in developing countries – the role of trade and storage

International agricultural commodity markets have experienced high volatility in recent years which raised concerns about food security and poverty impacts in developing countries. Most existing research to understand price dynamics in developing countries uses either partial or general equilibrium models (typically with annual time resolution) or time-series analysis with monthly, weekly or daily frequency. While the former approach allows analyzing structural determinants of prices like transaction costs or trade policies, it does not allow to model short-term price fluctuations and volatility. The second approach using time-series models, however, has problems combining high-frequency price data with slow-moving structural variables like infrastructure, trade regime, policy interventions or annual harvest shocks. The work at hand overcomes the shortcomings of both approaches and contributes to the ongoing discussion on drivers of food price volatility by looking at structural causes of domestic food price instability in developing countries which are particularly affected by increasing international price volatility. We first develop an equilibrium trade model that is able to explain how several explanatory variables such as storage policy, production variability, trade policy, transaction costs and international market volatility affect domestic price variability. In doing so, the equilibrium model provides a theory-based prediction on the relevance and direction these variables which is then tested in the empirical model. The empirical model uses a comprehensive data set of grain prices in more than 70 developing countries and innovative approaches to measure policy involvement. It addresses nonlinearities with respect to different country types and impacts of stabilization policies. The empirical analysis employs a dynamic panel, estimated by system generalized method of moment (GMM) that successfully accounts for changes in volatility over time. The panel approach also accounts for country and crop fixed effect. The regression results support the evidence that international price volatility and institutional quality strongly impact on domestic price volatility. New evidence is provided with respect to heterogeneity between study countries. According to this, grain stocks are particularly price stabilizing in food importing countries that are most dependent on international trade. Furthermore, the spill-over of international volatility is almost double for importer countries. Non significance of production shocks for intra-annual volatility does not differ from the exiting empirical literature. Distinguishing between high and low intervention countries shows that market related variables as stocks, trade policies, general inflation, and institutional quality are unimportant or less important to explain volatility in countries with interventionist governments. Regression results for public storage and food aid distribution look inconsistent as we find the level of public stocks to increase volatility, while food aid distribution stabilizes prices. On the other hand, the level of public stocks does not necessarily capture the extent of market intervention and further research is required to clearly extract the effects. Most insightful are the findings with respect to trade policies and regional integration. Insulation policies significantly reduce domestic price volatility not only for exporters but also for trade switchers at relevant margins. Using a unique data set on bilateral trade agreements, we find the relative share of regional trade to have a dominant stabilization effect on all types of countries. These results emphasize the positive effect of regional integration on trade flows and trade policy volatility. From this, a clear policy recommendation towards regional market integration can be deduced.