Using machine learning methods to predict financial performance: Does disclosure tone matter?

Mousa, Gehan A.; Elamir, Elsayed A. H.; Hussainey, Khaled

doi:10.1057/s41310-021-00129-x

Cited by 22 publications

(10 citation statements)

References 73 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…RF [ 67 ] is another popular ML algorithm based on the concept of ensemble learning. This concept involves combining multiple classifiers to improve model performance.…”

Section: Machine Learning Methodsmentioning

confidence: 99%

On the development of an information system for monitoring user opinion and its role for the public

et al. 2022

View full text Add to dashboard Cite

Social media services and analytics platforms are rapidly growing. A large number of various events happen mostly every day, and the role of social media monitoring tools is also increasing. Social networks are widely used for managing and promoting brands and different services. Thus, most popular social analytics platforms aim for business purposes while monitoring various social, economic, and political problems remains underrepresented and not covered by thorough research. Moreover, most of them focus on resource-rich languages such as the English language, whereas texts and comments in other low-resource languages, such as the Russian and Kazakh languages in social media, are not represented well enough. So, this work is devoted to developing and applying the information system called the OMSystem for analyzing users’ opinions on news portals, blogs, and social networks in Kazakhstan. The system uses sentiment dictionaries of the Russian and Kazakh languages and machine learning algorithms to determine the sentiment of social media texts. The whole structure and functionalities of the system are also presented. The experimental part is devoted to building machine learning models for sentiment analysis on the Russian and Kazakh datasets. Then the performance of the models is evaluated with accuracy, precision, recall, and F1-score metrics. The models with the highest scores are selected for implementation in the OMSystem. Then the OMSystem’s social analytics module is used to thoroughly analyze the healthcare, political and social aspects of the most relevant topics connected with the vaccination against the coronavirus disease. The analysis allowed us to discover the public social mood in the cities of Almaty and Nur-Sultan and other large regional cities of Kazakhstan. The system’s study included two extensive periods: 10-01-2021 to 30-05-2021 and 01-07-2021 to 12-08-2021. In the obtained results, people’s moods and attitudes to the Government’s policies and actions were studied by such social network indicators as the level of topic discussion activity in society, the level of interest in the topic in society, and the mood level of society. These indicators calculated by the OMSystem allowed careful identification of alarming factors of the public (negative attitude to the government regulations, vaccination policies, trust in vaccination, etc.) and assessment of the social mood.

show abstract

“…RF [ 67 ] is another popular ML algorithm based on the concept of ensemble learning. This concept involves combining multiple classifiers to improve model performance.…”

Section: Machine Learning Methodsmentioning

confidence: 99%

On the development of an information system for monitoring user opinion and its role for the public

et al. 2022

View full text Add to dashboard Cite

show abstract

“…In addition, prior studies argue that narrative reporting can help investors without financial backgrounds better understand the information presented in financial statements and make the picture much clearer to stakeholders (Arslan-Ayaydin et al, 2016;Boudt et al, 2018). Moreover, over 70% of financial documents in the last decade have been in the form of narrative disclosures, which attest to the importance of textual disclosures (Li, 2010;Mousa et al, 2022). Therefore, Narrative reporting in the UK studying NDT is critical to understanding how non-financial information and the way in which it is presented influence real decisions/activities.…”

Section: Literature Review and Hypotheses Development 21 General Back...mentioning

confidence: 99%

Does negativity matter under the principle-based approach? Evidence from narrative reporting in the UK

Bassyouny,

Machokoto

2023

IJAIM

View full text Add to dashboard Cite

Purpose This paper aims to investigate the association between negative tone in annual report narratives and future performance in the UK context. Under the principle-based approach in the UK, managers tend to bias the tone of narrative reports upward, as the reporting regime is more flexible than the rule-based approach in the USA. Consequently, any negative disclosure not mandated by regulators conveys credible information about a firm’s prospects. Design/methodology/approach This paper uses a sample of UK FTSE all-share non-financial companies from 2010 to 2019. The authors use the textual-analysis approach based on Loughran and McDonald (2011)’s wordlist (LM) to measure the negative tone in UK annual reports. Findings The results show a significant negative association between negative tone and future performance. Moreover, our further analyses suggest that only the negativity in the executive section of the annual disclosures correlates significantly with future performance. In summary, this study suggests that negativity does matter under the principle-based approach and can be used as an indicator of future performance. Originality/value In contrast to the literature arguing that only positivity has the power to affect a firm’s outcomes under the principle-based approach, the authors provide new empirical evidence suggesting that negativity also matters within the UK context and can be used as an indicator for future performance. Also, to the best of the authors’ knowledge, this is the first study to identify which section of the annual report is more informative about a firm’s future performance.

show abstract

“…One approach measures text tone by calculating the ratio of the difference between positive and negative tone words to the total vocabulary [61,62]. The other approach calculates the ratio of the difference between positive and negative words to the sum of positive and negative words [63][64][65][66][67][68]. The latter method is widely used in current literature.…”

Section: Plos Onementioning

confidence: 99%

Financial crisis early warning of Chinese listed companies based on MD&A text-linguistic feature indicators

Zhang,

Liu,

Niu

2023

PLoS ONE

View full text Add to dashboard Cite

Nowadays, the international situation is severe and complex, and the structural issues within the Chinese economy are prominent. Consequently, the financial risks faced by Chinese listed companies continue to escalate. Hence, it is of great practical significance to furnish effective early warnings for financial crises in listed companies. In this paper, we first employ web crawler technology and natural language processing technology to assess four text-linguistic features in the Management Discussion and Analysis (MD&A) section of the annual financial reports of listed companies in China from 2011 to 2020. These features are text tone, forward-looking, readability and similarity. Subsequently, we combine these features with traditional financial indicators and explore thirteen mainstream machine learning models to comparatively analyze their effectiveness in predicting financial crises in listed companies. The empirical findings of this research reveal that MD&A text readability and similarity indicators contribute valuable incremental information to prediction models, whereas text tone and forward-looking indicators exhibit the opposite effect. The latter two indicators can be manipulated more effortlessly by management, as this study’s empirical findings indicate no evidence of their contributions in incremental informational value. In fact, the forward-looking indicator even introduces additional noise to the prediction. Finally, by comparing the early warning effects of thirteen machine learning models, it is found that RF, Bagging, CatBoost, GBDT, XGBoost and LightGBM models maintain stable and accurate sample recognition ability. In general, this paper constructs a more effective financial crisis early warning model by exploring the MD&A text-linguistic feature indicators, thereby offering a fresh research perspective for further investigations in this field.

show abstract

Using machine learning methods to predict financial performance: Does disclosure tone matter?

Cited by 22 publications

References 73 publications

On the development of an information system for monitoring user opinion and its role for the public

On the development of an information system for monitoring user opinion and its role for the public

Does negativity matter under the principle-based approach? Evidence from narrative reporting in the UK

Financial crisis early warning of Chinese listed companies based on MD&A text-linguistic feature indicators

Contact Info

Product

Resources

About