Predictive Analytics in Finance Using the Arima Model. Application for Bucharest Stock Exchange Financial Companies Closing Prices
Stock markets can be volatile, thus accurate predictions can greatly help investors and stakeholders to make wise financial choices. The main goal of this paper is to test how well the Autoregressive Integrated Moving Average (ARIMA) model can capture and predict changes in closing prices. The ARIMA...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Sciendo
2024-12-01
|
| Series: | Studies in Business and Economics |
| Subjects: | |
| Online Access: | https://doi.org/10.2478/sbe-2024-0042 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Stock markets can be volatile, thus accurate predictions can greatly help investors and stakeholders to make wise financial choices. The main goal of this paper is to test how well the Autoregressive Integrated Moving Average (ARIMA) model can capture and predict changes in closing prices. The ARIMA model is the combination of autoregressive (AR) and moving average (MA) processes of an integrated or differenced time series model. Moreover, the selected model is part of the time series analysis under prediction algorithms, the purpose of the research being to predict the prices of the selected shares. Our analysis compiles daily trading data of financial companies on the BSE using a quantitative methodology. We preprocess the dataset to ensure reliability and accuracy, then conduct an exploratory data analysis to identify underlying patterns and correlations. Next, we use the ARIMA model, carefully optimizing the parameters through selection and a validation process, to forecast the closing prices over a 30-day period, we also evaluate the model’s performance. The preliminary results show that the ARIMA model seems to be efficient at predicting closing stock prices accurately. The model appears to understand the patterns and fluctuations in the stock market data, which gives useful information about future price changes. The forecasts we generated show similarities with the real market results, capturing important patterns, and making it a viable option for forecasting market performance. |
|---|---|
| ISSN: | 2344-5416 |