Masters Theses

Date of Award

3-1970

Degree Type

Thesis

Degree Name

Master of Science

Major

Agricultural Economics

Major Professor

James G. Snell

Committee Members

Charles Sappington, David W. Brown

Abstract

The objectives of this thesis were twofold. It was an attempt, first, to clarify the nature of economic relationship among various factors affecting the Brazilian cocoa price level in the world market, and secondly, to predict Brazilian cocoa prices in the international market. The period covered by the study was from 1946 to 1966, and secondary data were used in the analysis. Least square regression was applied to fit four linear structures to the data representing the variables assumed to be the most important economic factors in the determination of Brazilian cocoa prices. In the first equation the cocoa prices deflated by an appropriate price index appeared in relation with Brazilian cocoa production, the cocoa production from other areas, and the cocoa stock in consuming areas. In the second one, Brazilian cocoa production and cocoa inventories held by consuming countries appeared in relation with deflated Brazilian cocoa prices. In the remaining two structures the dependent variable was not deflated and the price index appeared as an extra variable. In this connection Brazilian cocoa price was stated as a function of Brazilian Cocoa production, the rest of the world cocoa production, the level of cocoa stock held by consuming areas, and the price index. In the last structure Brazilian cocoa production, the cocoa stock level, and the price index were used as independent variables. The rationality for using such economic factors was based on the brief analysis of the world cocoa economy carried out by this study. The following were the structures obtained by this analysis: Y₁=56.646-0.032891 X₂(t̳-̳1̳) + 0.0082769 X₃ - 0.011968 X₁(t̳-̳1̳) (0.01663) (0.02961) (0.06271) Y₁= 43.296- 0.041702 X₃ (0.01622) Y'₁1 = -14.553-0.04971 X₂(t̳-̳1̳) +0.90478 X4-0.04409 X₁(t-1)+0.012049 X₃ (0.01374)(0.22781)(0.04739)(0.02143) Y'₁ = -2.7673-0.048754 X₃-0.50215 X4-0.011009 X₁(t̳-̳1̳) (0.01740)(0.26007)(0.06083) where: Y₁ is the annual average wholesale New York spot price of Brazilian cocoa deflated by an appropriate price index; Y'₁ is the simple annual average wholesale New York spot price of Brazilian cocoa; X₁ is the level of cocoa production in Brazil; X₂ is the production of cocoa beans in other areas; X₃ is the volume of cocoa beans in stock in consuming areas; and X₄ is the wholesale price index calculated by the Bureau of Labor Statistics. On the basis of the statistical significance of the increases in the variations explained by the variables used in the equations, the structure below was selected as the one for attaining the objectives of the study: Y'₁ = -14.553 - 0.04971 X₂(t̳-̳1̳) +0.90478 X₄-0.04409 X₁1(t̳-̳1̳)+0.012049 X₃ (0.01374)(0.22781)(0.04739)(0.02143) R² = 0.6281 Se =6.6990 The ability of the model in predicting the Brazilian cocoa prices for 1967 and 1968 was considered reasonable. Inserting data into the equation above and solving for Y'₁, the Brazilian cocoa price level was estimated at 22.46 cents while the actual price was 26.35 cents per pound in 1967, an underestimate of 3.89 cents; and at 23.39 cents while actual price in 1968 was 32.85 cents per pound, an underestimate of 9.46 cents per pound. In dealing with future forecasts caution must be taken since from the coefficient of determination of 0.6281 it may be inferred that more price variation could be explained by the effects of other variables not included in the analysis. Further studying about the behavior of Brazilian cocoa price on monthly basis would result in a refinement of the model achieved in the present study.

Download
Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Share

COinS