EUVIMEDEUVIMED entdecken
Open Access · CC BY-SA 4.0

Forecasting bitcoin prices using the ARIMA–GARCH model: A volatility-based time series approach

Hizbul Watan, Rini Oktavia, Fitria Lestari

Results in Mathematical Modeling · 01.06.2026 · 1 (1) · 22-28

DOI: 10.55749/rmm.v1i1.178 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Results in Mathematical Modeling, online erschienen am 01.06.2026. Verfasst von Hizbul Watan und 2 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY-SA 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Bitcoin is one of the most actively traded digital assets and is characterized by high price volatility, making accurate forecasting essential for investment decision-making. This study aims to determine the best ARIMA–GARCH model for forecasting Bitcoin prices and to analyze future price movements based on historical weekly data. The dataset consists of 157 weekly Bitcoin closing prices collected from August 12, 2018, to August 8, 2021, obtained from Investing.com. The forecasting process was conducted for the subsequent five-week period from August 15, 2021, to September 12, 2021. The analysis began with stationarity testing using the Augmented Dickey–Fuller (ADF) test and Box–Cox transformation. Since the original series was non-stationary in the mean, first-order differencing was applied. Several ARIMA models were identified using the ACF and PACF plots, followed by diagnostic checking and ARCH-LM testing to detect heteroskedasticity effects. The presence of volatility clustering justified the implementation of the GARCH model. Model selection was based on parameter significance and the minimum Akaike Information Criterion (AIC) value. The results indicate that the ARIMA(0,1,2)-GARCH(1,3) model is the best model for forecasting Bitcoin prices. Forecasting results show a gradual decline in Bitcoin prices over the next five periods. The model achieved a Mean Absolute Percentage Error (MAPE) value of 3%, indicating excellent forecasting performance. These findings demonstrate that the ARIMA–GARCH approach is effective for modeling and forecasting highly volatile cryptocurrency price movements.

Zitieren

APA

Watan, H., Oktavia, R., & Lestari, F. (2026). Forecasting bitcoin prices using the ARIMA–GARCH model: A volatility-based time series approach. Results in Mathematical Modeling, 1(1), 22-28. https://doi.org/10.55749/rmm.v1i1.178

Vancouver

Watan H, Oktavia R, Lestari F. Forecasting bitcoin prices using the ARIMA–GARCH model: A volatility-based time series approach. Results in Mathematical Modeling. 2026;1(1):22-28. doi:10.55749/rmm.v1i1.178

BibTeX
@article{Watan2026,
  author = {Watan, Hizbul and Oktavia, Rini and Lestari, Fitria},
  title = {Forecasting bitcoin prices using the ARIMA–GARCH model: A volatility-based time series approach},
  journal = {Results in Mathematical Modeling},
  year = {2026},
  volume = {1},
  number = {1},
  pages = {22--28},
  doi = {10.55749/rmm.v1i1.178}
}
Kostenlos starten →