Digital Assets and the Global Economy: How the Use of Statistical Models Can Help Bitcoin Price Prediction

The purpose of the study is to analyze the potential of statistical modeling in predicting the prices of the Bitcoin cryptocurrency and its impact on the economy. In the course of the article, answers were received to such questions as: What is the impact of macroeconomic events on the dynamics of the Bitcoin price? How quickly does the cryptocurrency market stabilize after the falls? How effective is statistical modeling to solve the problem of predicting the price of Bitcoin? Which model shows the best results? What measures of regulation and control of the cryptocurrency market are necessary at the stage of its formation in the Russian Federation?

Materials and methods. Historical data on average monthly Bitcoin closing prices and macroeconomic events such as the COVID-19 pandemic and the Russian-Ukrainian conflict were collected and analyzed. The paper uses statistical models, including ARIMA and LSTM, to predict future Bitcoin prices based on historical data. The accuracy of the models was calculated based on such indexes as the mean absolute error (MAE) and the mean square error (MSE).

Results. Analysis of the impact of macroeconomic events showed that during the crisis, the attractiveness of Bitcoin increased and investors used this asset as a new investment tool. During the analysis of the consequences of the Russian-Ukrainian conflict for the cryptocurrency market, its reaction to geopolitical events was revealed according to the increased liquidity indexes in the market. In the process of modeling the dynamics of the average monthly Bitcoin price, the model with parameters (1, 1, 0) at MAE = 15.03% was recognized as the best ARIMA model. The LSTM neural network model on a similar data set showed a MAE error equal to 2.57%.

Conclusion. The analysis shows that itcoin was the most attractive investment tool during the crisis, which led to a sharp increase in its price in 2021. The Russian-Ukrainian conflict has also affected its price, causing a significant decline in 2022. However, statistical modeling methods predict an increase in the price of Bitcoin in the first half of 2023, and governments may consider regulating or controlling its use to reduce risks associated with the cryptocurrency market. The recommended measures are the introduction of regulations, the introduction of transaction taxes, the development of national digital currencies, public education and the prevention of criminal activity.

1. Veb-skreyping CryptoCMD = Web scraping CryptoCMD [Internet]. 2023. Available from: https://github.com/guptarohit/cryptoCMD.

2. Abdi F. and Ranaldo A. A simple estimation of bid-ask spreads from daily close, high, and low prices. Review of Financial Studies. 2017; 30; 12: 4437–4480. DOI: 10.1093/rfs/hhx084.

3. Boubaker S., Goodell J.W., Pandey D.K. and Kumari V. Heterogeneous impacts of wars on global equity markets: evidence from the invasion of Ukraine. Finance Research Letters. 2022; 48: 102934. DOI: 10.1016/j.frl.2022.102934.

4. Corwin S.A. and Scultz P. A simple way to estimate bid-ask spreads from daily high and low prices. The Journal of Finance. 2012; 67; 2: 719– 760. DOI: 10.1111/j.1540-6261.2012.01729.x.

5. Cryptocurrency Price Prediction [Internet]. 2023. Available from: https://github.com/abhinavsagar/cryptocurrency-price-prediction.

6. Latif Navmeen., Selvam Joseph Durai., Kapse Manohar., Sharma Vinod and Mahajan Vaishali. Comparative Performance of LSTM and ARIMA for the Short-Term Prediction of Bitcoin Prices. Australasian Accounting, Business and Finance Journal. 2023; 17(1): 256–276. DOI: 10.14453/aabfj.v17i1.15.

7. Sparkes M. Will bitcoin help or hinder Ukraine? New Scientist. 2022; 253; 3377: 8. DOI: 10.1016/S0262-4079(22)00409-2.

8. The Ruble, the Russian national currency, lost more than 30 % of its value against the dollar [Internet]. 2023. Available from: www.cnbc.com/2022/02/28.

9. Toai T.K., Senkerik R., Zelinka I., Ulrich A., Hanh V.T.X., Huan V.M. ARIMA for ShortTerm and LSTM for Long-Term in Daily Bitcoin Price Prediction. In: Rutkowski, L., Scherer, R., Korytkowski, M., Pedrycz, W., Tadeusiewicz, R., Zurada, J.M. (eds) Artificial Intelligence and Soft Computing. ICAISC 2022. Lecture Notes in Computer Science. 2023: 13588. Springer, Cham. DOI: 10.1007/978-3-031-23492-7_12.

10. Wiseman P. and Mchugh D. Economic dangers from Russia’s invasion ripple across the globe [Internet]. 2022. Available from: https://abcnews.go.com/US/wireStory/economic-dangers-russias-invasion-ripple-globe-83197306.