Evaluation of Effectiveness of Arima Model Predictions in Investment Portfolio Formation and Management
Iryna Brolinska
Orcid profile and Grigorij Žilinskij
Orcid profile 
Published Online: Jun 25, 2025
Page range: 108 - 122
Received: Oct 21, 2024
Accepted: Jun 06, 2025
DOI: https://doi.org/10.2478/jec-2025-0009
Keywords
© 2025 Iryna Brolinska et al., published by Sciendo
This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
Research purpose.
The increasing array of financial assets and investment opportunities nowadays is making investors consider new ways of investment portfolio formation and management. Many choose to take advantage of a wide variety of forecasting models in order to enhance investment portfolio performance results. However, each forecasting model has its application peculiarities, strengths, weaknesses and distinctive features. This paper offers a methodological basis and evaluation of the application of the ARIMA forecasting model in investment portfolio formation and management. The purpose of the research is to evaluate the effectiveness of ARIMA model predictions in investment portfolio creation and management.
Design / Methodology / Approach.
The period of two years, from May 2023 to April 2025, was chosen as the interval for conducting this research. The dataset for this experiment comprised five years of monthly securities prices, therefore the ARIMA model application and reinvestment occurred on a monthly basis. For an accurate evaluation of the ARIMA model-based investment portfolio, it has been compared with an ordinary mean-variance investment portfolio which has been created and managed in the same way. Additionally, the accuracy of the forecasts and their influence on portfolio performance has been evaluated.
Findings.
The findings of this study suggest that, given the monthly reinvestment and securities chosen, the implementation of the ARIMA model in investment portfolios is not recommended. This conclusion is based on the relatively low performance of the portfolio and the significant time and effort required to implement the ARIMA model, leading to its inefficiency.
Originality / Value / Practical implications.
The practical implications of this study suggest that investors should consider alternative forecasting models, different trading frequencies, or other financial assets to achieve better results. If investors opt to implement the ARIMA model for investment portfolio formation and management, it is important to note that the forecasts have shown to be accurate for short-term predictions, particularly in securities with strong trends and low-price volatility.