AUTHOR=Rao Ananth , M. V. Manoj Kumar , Moonesar Immanuel Azaad , Atalla Shadi , Prashanth B. S. , Joshi Gaurav , Soni Tarun K. , Le Thi , Verma Anuj , Marashdeh Hazem 

TITLE=Cross Country Determinants of Investors' Sentiments Prediction in Emerging Markets Using ANN

JOURNAL=Frontiers in Artificial Intelligence

VOLUME=Volume 5 - 2022

YEAR=2022

URL=https://www.frontiersin.org/journals/artificial-intelligence/articles/10.3389/frai.2022.912403

DOI=10.3389/frai.2022.912403

ISSN=2624-8212

ABSTRACT=Purpose: This paper presents a data model to assess investors' behavior to attract more investments that meet the United Nation's Sustainable Development Goal 3 and 8 . The proposed ANN model analyses the investor sentiments (IS) to capture the investors' behavior proxied by equity returns. The performance and generalization of the model have been experimentally demonstrated against three fast-emerging economies, i.e., China, India, and the UAE. Specific research objectives to attract investments for Health Sector and Growth in emerging markets viz., India, China, and the UAE are: 
1.	Are the USA-IS key IS predictors (proxied by equity returns)?
2.	How important are macroeconomic and socio-cultural factors in predicting IS?
3.	How important are economic crises and pandemic events in predicting IS?
4.	Is there contemporaneous relation in predicting IS across the three countries in terms of USA-IS, and if yes, is the magnitude of the impact of USA-IS uniform across the three countries IS? 

Design/Methodology: ANN model is applied to capture behavioural elements in the investors' decision making in these emerging economies. ANN framework uses Historical weekly data  of the AAII survey as a proxy of US-IS and respective countries' market indices as proxies for market return.  Other input variables include liquidity in the country captured through Stocks traded as % GDP, and stock turnover ratio as share of domestic shares traded besides value added by various sectors to country's GDP. Health Sector variables include: HDI, CHE as % GDP, per capita health expenditure in constant US$. 

Findings: The empirical findings confirmed the superiority of the ANN framework over traditional logistic model in capturing the cognitive behavior of investors. Health predictor - Current health expedntiure as percentage in GDP (CHE%GDP), USA IS predictor - Spread, and macro-factor GDP-annual growth % are the common predictors across the Pooled, China, India and the UAE tresting set that postively impacted the emerging markets' IS behavior. USA (S&P 500) return is the only common predictor across the Pooled, China, India and the UAE tresting set that negatively impacted the emerging markets' IS behavior. The magnitude of impact is however varying across the countries.