Professional Documents
Culture Documents
(a)
(b)
is determined on the assumption that the
farmer knows what and are.
Let us examine what happens to profits if the
farmer believes the price of the output is greater
or less than the actual price. The farmer is
optimistic about . That is holding
constant. The farmer will use more than the
optimal level and incurs losses. The net effect
will be reduction in profits.
(c)
Pessimistic Scenario :
(b)
Figure 1. Determination of profit maximizing input level.
ARIMA MODELING
ARIMA MODELS
Estimates ARIMA ( 2,1,1) ARIMA (2,1,2) ARIMA (1,1,1)
X(t) = .5449 X(t-1) - X(t) = .2547 X(t-1) - X(t) = - .1568 X(t-1)
.2357 X(t-2) + Z(t) - .9631 X(t-2) + Z(t) - + Z(t) + .2815 Z(t-1)
.4630 Z(t-1) .1587 Z(t-1) + .9975
Z(t-2)
WN Variance 10.24 9.23 10.69
MA Coefficient 2 .9975(.0070)
t computed 142.5
Ljung Box Q stat for 20.03 X2 (20) 13.24 X 2 (20) 28.49 X 2 (20)
the p-value = .4560 p-value=.8667 p-value= 0.983
residuals
Stationarity
This implies that the AR coefficients must satisfy
certain conditions.
AR (1) = 0.2547
AR (2) = -0.9631
To check for stationarity: we must apply the 3
conditions to the estimated AR coefficients;
stationarity conditions apply only to the AR
coefficients.
We find that all three conditions are met, and
hence the model is stationary. Moving average
models are always invertible.
Invertibility
To check the model for invertibility, we must
apply certain condition to the estimated MA
coefficient ; invertibility conditions apply only to
MA coefficients.
Figure 28. Residual plot.
Figure 29. Correlogram of the residuals.
FORECASTING
Davao
Approximate 95 Percent
Prediction Bounds
Step Prediction sqrt(MSE) Lower Upper
1 52.69314 3.03882 46.73716 58.64911
2 52.93209 4.50863 44.09533 61.76885
3 52.25034 5.71349 41.05211 63.44857
4 52.28194 6.58501 39.37556 65.18831
5 53.38199 7.25358 39.16525 67.59874
6 54.06717 7.93189 38.52096 69.61338
7 53.61762 8.66001 36.64432 70.59093
8 53.27861 9.29700 35.05682 71.50040
9 54.06061 9.80076 34.85149 73.26974
10 55.02172 10.28554 34.86242 75.18102
OUTPUT
• Forecast models were obtained using UBJ modeling
technique for selected vegetables and fruits
- Phase I
Tomato – Davao del Sur
Misamis Oriental
Ampalaya – Pampanga
Nueva Ecija
Tarlac
Marinduque
Ilocos Sur
Eggplant – Nueva Vizcaya
Pampanga
Squash – Cavite
Onion – Ilocos Norte
OUTPUT
- Phase II
Banana – Mindoro Oriental
Batangas
Isabela
Coffee Robusta – Davao City
Compostela Valley
Carabao Mango - Cebu
• Forecast prices were made for the selected
commodities