Master of Science (MS)


Agricultural Economics

Document Type



The purpose of this thesis is to study the empirical linkages between nearby futures prices for coffee at the New York (NY) Intercontinental Exchange (ICE)/NY Board of Trade (NYBOT) and cash prices (producer and export prices) in selected Latin American countries. This theme was entertained in Fortenbery and Zapata (2004) and subsequently by Fortenbery and Zapata (2004) and Li and Fortenbery (2013). This thesis uses data from January 1990 to May 2013 and adds producer cash prices in addition to export prices, thus expanding the dataset by over a decade and adding local market cash prices relative to the first paper, and also adds Brazil (the largest Arabica coffee exporter) and Colombia (a producer of high quality coffee) to the country mix; all cash prices were provided by the International Coffee Organization. Cointegration methodology is used to study price linkages between nearby futures prices in New York and the above spot prices. Implications for speculative activity are derived in light of the more recent paper above. Cointegration tests suggest that nearby futures market prices in New York are strongly linked with export prices in Brazil and Guatemala as well as with producer prices in Brazil and Honduras. Weak linkages exist for the remaining series of producer and export prices. Finding strong linkages for some markets is consistent with anecdotal evidence on the partial use of ICE/NYBOT futures prices to set domestic producer level prices in some countries. When evaluating the price relationship between producer and export prices in each country, Brazilian producer and export prices were the only ones that resulted to be cointegrated with each other, suggesting that local prices for coffee at different market levels are strongly linked and causal in at least one direction for Brazil. Optimal lag lengths used for the cointegration tests imply that the information transmission between the cointegrated series is slower than expected when compared to some US commodity markets where price changes can be reflected in 1 to 3 days. Impulse response functions from error correction models and vector autoregressive models confirm the causal nature of the relationship between coffee futures prices and cash prices in the four countries. When considering the implications of this research, preliminary results from a very simple regression analysis suggest that, consistent with Li and Fortenbery (2013), increases in intertemporal spreads by noncommercial speculative activity significantly decreases price volatility. This adds to existing evidence that efforts to limit the size of speculative positions may not strongly contribute to commodity price stability. For countries where coffee contributes to significant economic activity, and in which there is a large number of small producers, the gains from using commodity futures markets for coffee price risk management has considerable appeal. Hedging in futures markets, for example, or distributing futures price information to local cash market participants in developing countries could lead to more stable coffee producer’s income and possibly higher prices.



Document Availability at the Time of Submission

Release the entire work immediately for access worldwide.

Committee Chair

Zapata, Hector