APPLICATION OF GME TSALLIS WITH A KINK REGRESSION MODEL FOR INTELLIGENT FORECASTING OF CRUDE OIL PRICES VIA THE NONLINEAR RELATIONSHIP WITH GOLD

Saad Obaid Jameel Al Masoodi

Авторы

Saad Obaid Jameel Al Masoodi Bucharest Academy of Economic Studies, Bucharest, Romania

Ключевые слова:

GME Tsallis, kink regression model, Brent crude oil price, gold price, Wald test, in-sample prediction, nonlinear inflection points

Аннотация

This study stems from an important research need to understand the mechanisms through which safe-haven assets influence fluctuations in energy markets, particularly at a time when economic transformations and geopolitical changes are accelerating, This study aims to reveal the nature of the relationship between the price of crude oil (Brent) on the one hand and the price of an ounce of gold on the other, by employing a kink regression model with an unknown threshold, as developed by Hansen (2017). This is done using the generalized maximum entropy method based on the Tsallis scale, as it is a smart estimation measure; Tarkhamtham, Yamaka, and Sriboonchitta (2018) have demonstrated that it shows clear superiority when the tails are non-normal. A practical application was conducted on real monthly data covering the period from February 2015 to December 2025, i.e., 131 months These real-world observations were obtained from official sources: oil prices were obtained from the U.S. Energy Information Administration, and gold prices were obtained from the official websites of the World Bank. After conducting an in-depth analysis of the data, the results showed that there was a significant kink point when the gold price reached 2,254$ per ounce, and that the relationship between the gold and oil prices: it was a strong positive relationship (β_1^-=0.696) and then shifted (changed) to a negative relationship (β_1^+=−0.441) before and after the kink point. The Wald test value reached 16.04 at the 1% level according to the table established by Hansen (2017), and the mean absolute percentage error (MAPE) was 4.48% in the forecast within the sample, and that this finding confirms the usefulness of the proposed framework for identifying nonlinear structural relationships and for providing reliable forecast values for oil prices.

Библиографические ссылки

D. T. Pele and M. Mazurencu-Marinescu-Pele, “Using high-frequency entropy to forecast Bitcoin’s daily value at risk,” Entropy, vol. 21, no. 2, p. 102, 2019, doi: 10.3390/e21020102.

E. A. Drzazga-Szczęśniak, P. Szczepanik, A. Z. Kaczmarek, and D. Szczęśniak, “Entropy of financial time series due to the shock of war,” Entropy, vol. 25, no. 5, p. 823, 2023, doi: 10.3390/e25050823.

B. E. Hansen, “Regression kink with an unknown threshold,” Journal of Business & Economic Statistics, vol. 35, no. 2, pp. 228–240, 2017, doi: 10.1080/07350015.2015.1073595.

A. Golan, G. Judge, and D. Miller, Maximum Entropy Econometrics: Robust Estimation with Limited Data. Chichester: John Wiley & Sons, 1996.

P. Tarkhamtham, W. Yamaka, and S. Sriboonchitta, “The generalized maximum Tsallis entropy estimator in kink regression model,” J. Phys. Conf. Ser., vol. 1053, p. 012103, 2018, doi: 10.1088/1742-6596/1053/1/012103.

P. Tarkhamtham and W. Yamaka, “High-order generalized maximum entropy estimator in kink regression model,” Thai Journal of Mathematics, pp. 185–200, 2019.

H. Tong, Non-linear Time Series: A Dynamical System Approach. Oxford: Oxford University Press, 1990.

K. S. Chan and R. S. Tsay, “Limiting properties of the least squares estimator of a continuous threshold autoregressive model,” Biometrika, vol. 85, no. 2, pp. 413–426, 1998.

E. T. Jaynes, “Information theory and statistical mechanics,” Physical Review, vol. 106, no. 4, pp. 620–630, 1957.

C. E. Shannon, “A mathematical theory of communication,” Bell System Technical Journal, vol. 27, no. 3, pp. 379–423, 1948.

C. Tsallis, “Possible generalization of Boltzmann-Gibbs statistics,” J. Stat. Phys., vol. 52, no. 1–2, pp. 479–487, 1988.

W. Bank, “Commodity Markets Outlook: Pink Sheet Data,” World Bank Group, Washington, DC, 2025. [Online]. Available: https://www.worldbank.org/en/research/commodity-markets

C. D. Lewis, Industrial and Business Forecasting Methods. London: Butterworth Scientific, 1982.

H. Han et al., “A hybrid time series forecasting method based on GARCH and CEEMDAN-GCN model,” Journal of Cloud Computing, vol. 13, no. 1, p. 2, 2024.

R. F. Engle, “Autoregressive conditional heteroscedasticity with estimates of the variance of United Kingdom inflation,” Econometrica, vol. 50, no. 4, pp. 987–1007, 1982.