Examining ARIMA Model in Predicting Stock Return for Finnish Major Banks.
Karimi, Mahla (2023)
2023
All rights reserved. This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-2023081324714
https://urn.fi/URN:NBN:fi:amk-2023081324714
Tiivistelmä
This thesis aims to examine the ARIMA model in predicting stock return for Finnish major banks: OP, Nordea, and Danske. Investors invest their time and money in this market to generate income or capital appreciation. So, finding the right company for investment is crucial. According to CFA Institute (2003), the efficient market hypothesis states that the market promptly reflects in stock prices once new information comes into it. So, neither technical analysis nor fundamental analysis cannot generate more returns than when an investor randomly selects stocks. However, Investors have always been looking for ways to predict the stock market, and many believe that data science can help us turn the dream of predicting the stock market into reality. As time series analysis is used for variables that change over time, using time series analysis for tracking stock prices over time is a favoured approach. One widely used time-series model is the ARIMA model, which can be used to forecast future data.
This thesis reviews the traditional ways investors use to make a successful investment, such as researching the company, applying financial ratios, using Beta, and examining dividend history & yield. And then, the thesis builds random buying models and ARIMA models to compare the performance of these models to examine if time series models such as ARIMA can help investors predict stock prices in order to earn higher returns from their investments. Results show that although the time series model can be challenging to implement and might not be suitable for all investors, they can bring higher returns than random buying methods but also have higher risks of losing money. So, investors can use traditional ways for a successful investment and use time series models as an additional tool that can help them make a better decision based on the stock's historical data; This will help investors to pick stocks with the potential highest return keeping in mind that stock return cannot be predicted fully and there is always a risk in investing in this market.
This thesis reviews the traditional ways investors use to make a successful investment, such as researching the company, applying financial ratios, using Beta, and examining dividend history & yield. And then, the thesis builds random buying models and ARIMA models to compare the performance of these models to examine if time series models such as ARIMA can help investors predict stock prices in order to earn higher returns from their investments. Results show that although the time series model can be challenging to implement and might not be suitable for all investors, they can bring higher returns than random buying methods but also have higher risks of losing money. So, investors can use traditional ways for a successful investment and use time series models as an additional tool that can help them make a better decision based on the stock's historical data; This will help investors to pick stocks with the potential highest return keeping in mind that stock return cannot be predicted fully and there is always a risk in investing in this market.