Towards a circular economy: Secondary raw materials price prediction based on their listings on global stock quotes

This study aims to develop and validate a statistical model of the purchase price and price prediction of secondary raw materials, depending on the price of their original counterparts on the global stock markets. The paper investigates the relationship between the prices of secondary raw materials and their original counterparts, estimating a linear model to represent this relationship. The derived equations were statistically validated, yielding models of purchase prices that are determined by stock prices with a time delay. The data indicates that the model for predicting the purchase price of copper cable scrap in Poland, considering stock quotes in the UK and the USA from four days prior, explains about 98 % of the variance of the considered purchase prices, as indicated by the coefficient of determination ( R 2 value). Similarly, models for predicting purchase prices of pieces of aluminum sector considering stock quotes in the UK from 21 days prior and for scrap nickel considering stock quotes in the UK from 19 days prior explain about 95 % and 91 % of the variance of the considered purchase prices, respectively. Detailed verification of these models indicates that they meet the necessary assumptions for linear regression, affirming their suitability for prediction. It is likely that the resulting forecasts will only contain relatively small errors, which was confirmed by empirical tests for models obtained for copper and aluminum. However, empirical tests for the nickel model only indicated small partial errors.