Estimation of Optimal Storage Level in Korea Rice Industry : Application of Dynamic Stochastic Optimization Model

1
Estimation of Optimal Storage Level in Korea Rice
Industry Application of Dynamic Stochastic
Optimization Model
2004. 7. 23 Department of Agricultural Economics
Gyeongsang National University Jeong-Bin, Im
2
lt ? ? ? ? gt
1. Necessity of Grain Storage 2. Review of
the Previous Studies 3. Analytical Approach in
This Study 4. Model Specification and
Analytical Result 5. Summary and Conclusion
3
1. Necessity of Grain Storage
? Preparation for uncertainty in world grain
supply and demand - in order to stabilize
domestic market supply Food security/availabilit
y    - in order to stabilize domestic price and
income Market stability   ? Governments in
both developing and developed countries have
intervened in the grain market by means of
stockpiling schemes. ? Research Motivation
(1) Why is many governments grain storage level
higher than that in efficiency criterion?
(2) Why is the current storage decision sensitive
to current harvest level rather than carry-overed
stocks level? (3) What is the optimal grain
storage rule if policymaker put the different
welfare weight toward interest groups?
4
2. Review of the Previous Studies
? Two types of the previous study on optimal
storage level.    (1) Optimal inventory
problem Social planners dynamic optimization
model - stockpiling requires back
quantities from current consumption such that
the expected social welfare, as measured by
an objective function, is maximized
given the current state of the world.
(Gustafson (1958),Gardner (1979), Burt, Koo and
Dudly (1980)).    (2) Competitive private
storage private rational expectation -
Competitive private industry could carry socially
optimum stocks . (Wright and Williams
(1982,1984,1991), Miranda and Glauber (1993)
etc.).   ? Limit of existing research   - Only
focus on stabilizing domestic market   - Can not
explain high grain storage level in many
countries - Limitation on explaining many
governments storage behavior
5
3. Analytical Approach in This Study
? Political Preference Function(PPF) approach  
- Different welfare weights(or political weights)
toward interest groups.   - Optimal storage
level in Korean rice industry with policy
objective function. - Comparison with the
optimal storage rules derived from PPF Model and
conventional utilitarian social welfare
model( or competitive market model) 
6
4. Model Specification and Analytical Results
  A. Dynamic optimization model in PPF approach
7
? Equation (4-1) is generally transferred to
Equation (4-2) called Bellmans equation
8
B. Derive optimum storage level with PPF model
   ? Used parameter for solving the dynamic
optimum model      - Price elasticity of
demand - 0.117(Doo Bong, Han 2003)     -
8years(19952002) average price and consumption
data      - Harvest distribution average 5,195
ton, standard deviation 257 ton -
Welfare weight toward interest groups ?p
1.15, ?c 0.88, ?G 0.96(Jeong-Bin, Im 2003)
9
Optimal Storage Rule
ltFigure 4-1gt Optimal storage rule Equilibrium
storage and supply
Note SWG0 ?p 1.15, ?G 0.96, ?c 0.88, 
      SWG1 ?p ?G 1.12 gt ?c 0.76
      SWF ?p ?G ?c 1, and CMS
Competitive Market Storage.
10
ltTable 4-1gt Optimal Carryover Levels for
alternative policy objective function in Korea
Rice Industry
(Use average harvest and carried stock data
during 1995 2002)
Total Market Supply 6000 meter ton Total Market Supply 6000 meter ton Optimal Storage Level (meter ton) Optimal Storage Level (meter ton) Optimal Storage Level (meter ton)
Harvest Carried Stock SWG0 SWG1 SWF or CMS
5200 800 1,005 1,031 580
Note SWG0 ?p 1.15, ?G 0.96, ?c 0.88, 
      SWG1 ?p ?G 1.12 gt ?c 0.76
     SWF ?p ?G ?c 1, and CMS
Competitive Market Storage.
11
ltfigure 4-2gt  Optimal Storage Rule in 3-D ?p
1.15, ?G 0.96, ?c 0.88
12
ltFigure 4-3gt Optimal Storage Rule in 3-D ?p
?G ?c 1, or CMS
13
5. Summary and Conclusion

• Optimal grain storage level
• - depend on the relative magnitude of welfare
  weights toward interest groups
• - depend on the difference in marginal
  propensities of current harvest and carryover
  stocks
• (2) In PPF approach, we can explain many
  governments public storage behavior with high
  concerns for grain producer
• (3) Estimated optimal rice storage level derived
  by PPF approach is larger than derived by
  traditional utilitarian social welfare
  maximization problem or competitive storage
  market model.
•     

14
(4) Optimal storage level in this research is
similar to FAO level   - optimal rice storage
level is 19 level of average consumption during
1995 2002 year - storage for market
stabilization 11, reserved stock for food
security 8   ? Optimal rice storage level in
the previous research (1) Rice consumptions
11 13 is optimal storage level KREI(2003)
    (2) Rice consumptions 11 15 is optimal
storage level KREI(2003) if
considered the import trouble from international
market     (3) Storage Level recommended by FAO
      - Consumptions 17 19 level before
1997       - Consumptions 19 20 level at
1997       - Distribution Stock 12, Reserved
stock 78
 
15
Policy Implication for decreasing rice stocks
  ? Efforts for expanding rice consumption
- more elastic demand reduces the
incentive for holding stocks ? Efforts for
reducing production uncertainty   - less
production uncertainty reduces the incentive for
holding stocks