Predicting gold market trends using Newton’s Divided Difference method

Gold is a vital financial asset often used as a hedge in times of economic instability. This study explores the application of Newton’s Divided Difference (NDD) method to forecast gold market trends using historical price data from March 2022 to April 2025. Both monthly (36 data points) and quarterly (12 data points) datasets were analyzed to assess the impact of data resolution on model accuracy and efficiency. Polynomial interpolation models of degree 3 and 7 were constructed using Newton’s Forward and Backward Divided Difference techniques. Error metrics such as Absolute Error (AE) and Percentage Absolute Relative Error (PARE) were used to evaluate prediction accuracy. The results revealed that while monthly data offered finer detail, quarterly data produced more stable and interpretable forecasts. This study confirms that NDD is a practical method for financial trend prediction, offering a simple yet effective alternative to more complex algorithms.