A common process across categories is used to estimate initial values for the first years. It relies upon the first year values of production, imports, exports, and losses, all of which are generated in the pre-processor and discussed elsewhere in this document, to compute an apparent level of demand or consumption. The reason for this is that demand data are much less available for food and agriculture than are production and trade data. An initial estimate of demand in each category is given as the sum of post-loss production and imports minus exports:

An initial portion of this first estimate of demand is set aside for the purposes of satisfying the need for growth in food stocks as underlying total food demand and supply change (using initial economic growth as a proxy for that annual stock growth). See also the section on stocks.  That adjusted demand then becomes the basis, along with production, for an estimate of food stocks.

where

agdstl is a parameter used to set desired stock levels for agricultural commodities.  It is set to be 1.5 times dstl , which is a global parameter that can be adjusted by the user

igdpr is the initial growth rate in GDP

Given the resulting initial model year value for food stocks, it is then possible to calculate a more precise increment of stock change needed and add that back into the demand.