3.3 Limitations and Prospects
The selection of technical indicators in this study is
primarily driven by the availability of data sourced
from Yahoo Finance; however, this reliance may
limit the comprehensiveness of the analysis. As a
result, the chosen indicators might not fully capture
all essential market dynamics, potentially
overlooking critical signs that could inform
investment decisions. While the focus on technical
indicators offers quantifiable and straightforward
metrics for model integration, it inherently narrows
the scope of the analysis. Other influential factors,
such as macroeconomic conditions—like interest
rates, inflation, and GDP growth—and investor
sentiment, which can be gleaned from social media
and news sources, are significant drivers of stock
market behavior. To enhance the model's predictive
accuracy and factor identification capabilities, it is
imperative to explore and incorporate these additional
parameters. For instance, integrating macroeconomic
indicators could provide a more holistic view of the
market context, allowing for better understanding of
how external economic factors impact stock
performance. Furthermore, incorporating sentiment
analysis could help capture the emotional and
psychological dimensions of investor behavior,
which often play a crucial role in market movements.
By expanding the range of indicators and factors
considered, future studies can develop a more robust
framework for stock prediction, ultimately leading to
improved investment strategies. This comprehensive
approach would not only enrich the analytical model
but also align it more closely with the multifaceted
nature of financial markets, increasing its
applicability and relevance to real-world trading
scenarios.
4 CONCLUSIONS
In summary, feature importance analysis provides
essential statistical support for factor selection in
stock selection. The results demonstrate that sectors
with distinct characteristics should assess tailored
groups of technical factors for achieving more
accurate price predictions. In volatile sectors, such as
semiconductors, short-term trends and volatility
emerge as critical indicators for effective decision-
making. Conversely, in more stable sectors like
public utilities, emphasis should be placed on broader
sector conditions to capture essential dynamics. As
the field evolves, the identification of more
meaningful and relevant factors that influence the
stock selection process is vital. A comprehensive
analysis that incorporates a wider array of related
factors can achieve deeper insights into stock
performance and selection strategies. This paper
presents a robust framework for determining which
factors to integrate into models for stock price
prediction and selection. Future research should focus
on quantifying additional factors, including
macroeconomic conditions, industry-specific trends,
and fundamental indicators, to further enhance this
feature importance analysis. On this basis, investors
can develop a more comprehensive understanding of
the factors driving stock prices, ultimately leading to
more informed and strategic investment decisions.
REFERENCES
Breitung, C., 2023. Automated stock picking using random
forests. Journal of Empirical Finance, 72, 532–556.
Feng, G., Giglio, S., Xiu, D., 2020. Taming the Factor Zoo:
A Test of New Factors. The Journal of Finance, 75(3),
1327–1370.
Gu, S., Kelly, B., Xiu, D., 2020. Empirical Asset Pricing
via Machine Learning. Review of Financial Studies,
33(5), 2223–2273.
Kakushadze, Z., 2016. 101 Formulaic Alphas. Wilmott,
2016(84), 72–81.
Leippold, M., Wang, Q., Zhou, W., 2022. Machine learning
in the Chinese stock market. Journal of Financial
Economics, 145(2), 64–82.
Nti, I. K., Adekoya, A. F., Weyori, B. A., 2019. Random
Forest Based Feature Selection of Macroeconomic
Variables for Stock Market Prediction. American
Journal of Applied Sciences, 16(7), 200–212.
Rapach, D., Zhou, G., 2021. Asset Pricing: Time-Series
Predictability. SSRN Electronic Journal, 3941499
Rasekhschaffe, K. C., Jones, R. C., 2019. Machine
Learning for Stock Selection. Financial Analysts
Journal, 75(3), 70–88.
Wolff, D., Echterling, F., 2023. Stock picking with machine
learning. Journal of Forecasting, 43(1), 81–102.
Yang, H., Liu, X., Wu, Q., 2018. A Practical Machine
Learning Approach for Dynamic Stock
Recommendation. SSRN Electronic Journal, 3302088.