set c   crops       /corn, soybean, wheat/
    obs obervations /1*10/;

table observ(obs,c) variance coefficients
         corn       soybean     wheat
 1      162.46       135.60     105.49
 2      151.98       142.75      99.61
 3      151.94       135.11      98.21
 4      142.30       148.35     124.28
 5      145.80       134.32      99.23
 6      151.49       146.37     105.30
 7      119.68       116.97     106.25
 8      109.10       110.86     107.79
 9      137.29       140.77     125.99
 10     129.66       124.74     128.00


scalar riska inverse of the risk aversion coefficient /.00035/;

variables exputil expected utility
          x(c)  crop levels
          y(obs)  levels of terminal wealth;

positive variable x;

equations obj         expected utility
          wealth(obs) wealth equations
          land        land constraint;

obj..         sum(obs,-(1/10)*exp(-riska*y(obs))) =e= exputil;
wealth(obs).. sum(c,observ(obs,c)*x(c))  =e= y(obs);
land..        sum(c,x(c)) =l= 1280;

model port using /all/;
solve port using nlp maximizing exputil;