Global gold prices forecasting using Bayesian nonparametric quantile generalized additive model
| Autor: | Yudhie Andriyana, Yollanda Nalita, Bertho Tantular, I Gede Nyoman Mindra Jaya, Annisa Nur Falah |
|---|---|
| Jazyk: | angličtina |
| Rok vydání: | 2023 |
| Předmět: | |
| Zdroj: | International Journal of Data and Network Science, Vol 7, Iss 3, Pp 1033-1044 (2023) |
| Druh dokumentu: | article |
| ISSN: | 2561-8148 2561-8156 |
| DOI: | 10.5267/j.ijdns.2023.6.002 |
| Popis: | Gold is one of the most attractive commodities and popular investments. Investment experts often recommend investing in gold because gold is one of the safest investments. It is a stable classic hedge, although the conditions of currency volatility or global markets are depreciated. However, the gold price fluctuations can be influenced by some other factors, such as the USD Index, which reflect and measure the strength of the US Dollar currency, and the Index of Dow Jones Industrial Average (DJIA) or a reflection of the political and economic conditions of the stock market. In this study, we conduct a global gold price forecast (USD) based on the USD Index, the DJIA Index, and the influence of time trends. Based on the data's characteristics, we face the fact that the data is nonlinear, contains outliers, and its pattern is not easy to specify parametrically. Due to the complexity of the model, we then propose a more flexible, robust modeling technique called the Bayesian Nonparametric Quantile Generalized Additive Model method. According to the results for the median case, the proposed method shows an accurate forecasting category due to the value of the Mean Absolute Percentage Error, MAPE less than 10 percent. |
| Databáze: | Directory of Open Access Journals |
| Externí odkaz: |
|