We started from the original model calculating the prognoses within six months following
the approximation term after modification of the initial condition value by the obtained monthly
price prognoses The original model forecasts the aluminium price reliably within
the stable price course, when the price does not changed rapidly. Within the rapid increase or
decrease of stock exchanges, but also in the case of changes in the price course the forecasting
fails. Since the variability with rapid and sudden changes is typical of the commodity price
course, we judged the possibility of making the forecasting more accurate by using
the modification of the initial condition value by aluminium price. We analysed the use of the
initial condition drift concerning differently chosen stock exchanges for successful forecasting.