Managing Risk by Coordinating Investment, Marketing, and Production Strategies: Reply Donald A. Johnson and Michael D. Boehlje Perry correctly states that the distribution of a gross revenue variable differs from an additive distribu- tion of price and yield variables. However, our anal- ysis did not error by treating them as being the same. One of the key goals of the model was to allow disposal of produced crops through cattle feeding or a variety of marketing programs; practices Iowa farmers follow or might want to follow. Linear pro- gramming studies typically use transfer equations to eliminate the need to enter production coeffi- cients in every disposal activity. This is an efficient practice that reduces matrix density and processing costs. Such techniques become even more impor- tant in quadratic programming (QP), particularly if the code requires the entire matrix to load into memory (as the Rand QP code does). This technique causes an inconsistency when ap- Donald A. Johnson is senior economist, Caterpillar, Inc., Peoria IL; and Michael D. Boehlje is professor and head, Department of Agricultural and Applied Economics, University of Minnesota. plied to QP-production activities have variable output, while disposal activities use fixed amounts. Either this inconsistency has to be eliminated or all disposal activities have to be converted into gross revenue activities. However, the latter solution would eliminate using produced crops to feed cattle, an unrealistic choice for an Iowa farmer. Instead, the study removed yield variability by adjusting production costs to reflect the purchase or sale of enough output to keep that available for transfer equal to expected output. This increased production cost variability to compensate for the removal of yield variability. So production activi- ties are independent of marketing activities and our analysis is theoretically correct. The following steps were used: 1. Determine historical costs of production. 2. Determine how much historical yields differed from expected yields. 3. Value the difference between actual and expected yields at actual prices. Table 1. Cost of Growing an Acre of Corn Adjusted for Yield Variability Ex- Cost Actual Actual pected Yield Cost Adjusted Cost in Index Cost Yield Yield Difference Price Difference Cost Johnson (1) (2) (3) (4) (5) (6) (7) (8) (9) 1965 96 38.08 73.5 86.8 13.3 .93 12.37 50.45 50.16 1966 100 39.66 91.6 88.6 -3.0 1.13 -3.39 36.27 36.20 1967 100 39.66 80.2 90.4 10.2 .90 9.18 48.84 48.78 1968 100 39.66 98.2 92.2 -6.0 .82 -4.92 34.74 34.62 1969 104 41.25 111.9 94.0 -17.9 .93 -16.65 24.60 24.56 1970 108 42.84 86.7 95.8 9.1 1.16 10.56 53.39 55.38 1971 113 44.82 105.0 97.6 -7.4 .82 -6.07 38.75 37.90 1972 121 47.99 118.0 99.4 -18.6 1.02 -18.97 29.02 28.94 1973 146 57.91 113.3 101.2 -12.1 2.08 -25.17 32.74 32.68 1974 166 65.84 76.0 103.0 27.0 3.51 94.77 160.61 160.46 1975 182 72.19 100.9 104.8 3.9 2.51 9.79 81.98 81.82 1976 193 76.55 90.6 106.6 16.0 2.21 35.36 111.91 111.82 1977 200 79.33 120.9 108.4 -12.5 1.63 -20.38 58.95 58.90 Means 52.75 58.64 58.63 Variances 241.14 1,512.06 1,511.72 Note: (8) equals (2) plus (7); (7) equals (5) times (6). Actual yield for 1969 was incorrectly shown as 11.9 in Johnson. Western Journal of Agricultural Economics, 11(2): 235-236 Copyright 1986 Western Agricultural Economics Association