Predicting Bullish and Bearish Markets at the Nairobi Securities Exchange Using News-Based Sentiment Analysis
International Journal of Finance and Accounting · 30.04.2026 · 5 (1) · 241-258
DOI: 10.37284/ijfa.5.1.4905 ↗Kurz gefasst
Fachartikel in der Zeitschrift International Journal of Finance and Accounting, online erschienen am 30.04.2026. Verfasst von Jackson Barngetuny. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Stock market prediction in emerging markets presents unique challenges, including organisational inefficiencies, limited data, and heightened sensitivity to investor behaviour. At the Nairobi Securities Exchange, delayed information and behavioural biases affect price formation. These issues reduce the effectiveness of traditional forecasting methods. This study describes a sentiment-driven predictive framework that integrates quantitative stock data with qualitative financial news to strengthen market forecasting precision. This study examines whether financial news sentiment can give early signals of bullish (rising market) and bearish (falling market) trends, improving market forecast accuracy. Using a dataset spanning 2014 to 2024 that includes daily stock prices, trading volumes, and National Stock Exchange (NSE) index movements, the research integrates local and international news with historical market data. This approach captures both short-run fluctuations and wider trends. Sentiment is extracted from textual data using natural language processing (NLP) methods, including lexicon-based and machine-learning approaches. It is then combined with technical indicators (such as moving averages and momentum measures) and lagged variables (past values of observed indicators) to develop predictive models, including logistic regression, random forests, support vector machines, and long short-term memory (LSTM) networks. The results indicate that including sentiment, which refers to the overall emotional tone of news or social media content, boosts predictive performance. Models that combine sentiment and technical indicators—quantitative measures such as moving averages or momentum—outperform those based solely on historical price data. Positive sentiment is tied to bullish (rising) trends, while negative sentiment precedes bearish (declining) periods. News sentiment functions as a leading indicator, or early signal, in less efficient markets. The study gives practical recommendations for investors. These include using sentiment trends (general market mood from investor communication) for early market entry, reducing exposure during negative sentiment (pessimistic market mood), and confirming trades with technical indicators (tools analysing past prices to guide decisions). Investors can capitalise on delayed reactions (market response lagging behind news), prioritise sentiment-sensitive sectors (industries strongly influenced by investor mood), monitor sentiment intensity (strength of optimism or pessimism), combine sentiment with trading volume (number of shares traded), diversify across stocks, and use sentiment dashboards (tools summarising market mood) within daily investment routines. In summary, integrating behavioural insights, textual sentiment, and analytics improves market prediction and decision-making. Combining multiple signals lets investors make informed, timely, and risk-aware decisions. This approach supports greater market efficiency and transparency.
Zitieren
APA
Barngetuny, J. (2026). Predicting Bullish and Bearish Markets at the Nairobi Securities Exchange Using News-Based Sentiment Analysis. International Journal of Finance and Accounting, 5(1), 241-258. https://doi.org/10.37284/ijfa.5.1.4905
Vancouver
Barngetuny J. Predicting Bullish and Bearish Markets at the Nairobi Securities Exchange Using News-Based Sentiment Analysis. International Journal of Finance and Accounting. 2026;5(1):241-258. doi:10.37284/ijfa.5.1.4905
BibTeX
@article{Barngetuny2026,
author = {Barngetuny, Jackson},
title = {Predicting Bullish and Bearish Markets at the Nairobi Securities Exchange Using News-Based Sentiment Analysis},
journal = {International Journal of Finance and Accounting},
year = {2026},
volume = {5},
number = {1},
pages = {241--258},
doi = {10.37284/ijfa.5.1.4905}
}
EUVIMED