Rare metals are the essential minerals for semi-conductors, PDP-LCD and other specialized metals. Supply and demand plans based on the reasonable demand forecasting are required because exogenous variables such as regional imbalance, productivity and ...
Rare metals are the essential minerals for semi-conductors, PDP-LCD and other specialized metals. Supply and demand plans based on the reasonable demand forecasting are required because exogenous variables such as regional imbalance, productivity and price variation may have bad influences on the stable supply system.
Conventional demand forecasting models have estimated the demand functions with the simplified Regressive Model. However, probabilistic characteristics are not included in the nonstationary time series and the spurious regression data are applied to the policy establishment process.
In this research, therefore, uncertain factors of the supply and demand of chromium, one of the rare metals, are analyzed and an adequate demand forecasting model is suggested by focusing on the domestic demand forecasting system and the relationship between demand and price. Thus, a demand function is determined in the consideration of various parameters affecting the demand forecasting of chromium.
A Vector Error Correction Model (VECM) of the rare metal is constructed by considering the stability of time series, time balancing, unit root test, cointegration test of variables and demand of chromium. The VECM consists of demand forecasting model reflecting dynamic behavior of cointegration relations instead of the current stationary model and analyzes the long/short term chromium demand through the analysis of price, GDP and the index flexibility of industrial product.
Many variables such as rapid economic growth, sectional demand and external impact on the domestic market may affect domestic chromium demand, and structural variation of the chromium demand are simulated for the suggested model to estimate the extent of the variation analysis. The simulation results, however, demonstrate that the error correction model is not effective for the radical variations related to the jump phenomena or limit cycles of chromium demand.
Therefore, Heteroskedastic Time Series Model is implemented for the demand forecasting model to consider the nonstationary time series. Based on the chromium time series data, characteristics and applicability of AR model, MA model, ARIMA model, ARCH model, GARCH model, IGARCH model and EGARCH model are investigated.
Heteroskedastic Time Series Model, of which application has been limited to Economic and financial analysis, is currently applied to the demand forecasting of rare metals to explain jump phenomena and limit cycles. The demand forecasting model constructed through this research and the applicability study of the model for rare metals is expected to contribute to establish domestic supply policy of rare metals as well as theoretical and practical studies on the demand forecasting model.