Sentiment Index Research Articles

Climate change and crude oil prices: An interval forecast model with interval-valued textual data

Climate change compels the development and enforcement of policies and regulations designed to diminish carbon emissions, imposing substantial implications on the energy sector. Given the contribution of crude oil prices to carbon emissions, developing precise forecasting methods is imperative. However, existing studies often overlook the inherent uncertainty in price movements by focusing solely on point forecasting. To address this limitation, this paper constructs a threshold autoregressive interval-valued model with interval sentiment indexes for climate change (TARIX) to analyze and forecast interval-valued crude oil prices. We have found that the interval climate sentiment index, derived from social media, can significantly enhance the accuracy in forecasting interval crude oil prices. Moreover, we propose an interval-based trading strategy that can effectively reduce volatility and enhance returns. Our empirical results demonstrate that our interval-valued forecast model outperforms traditional forecasting methods in terms of forecasting accuracy and profit generation.

Russia-Ukraine War and Price Volatility of Global Commodities: The Role of Public Sentiments

We analysed how public sentiments have affected global commodity market volatility during the Russia-Ukraine war. Using principal component analysis, we created a sentiments index from 30 carefully selected Google trends search keywords related to the war. We tested the predictability of the sentiments index against market volatility. Our results show that while public sentiments increase commodity market volatility, incorporating the sentiment index into our predictive model significantly improves its precision.

Deep-learning-based stock market prediction incorporating ESG sentiment and technical indicators

As sustainability emerges as a crucial factor in the development of modern enterprises, integrating environmental, social, and governance (ESG) information into financial assessments has become essential. ESG indicators serve as important metrics in evaluating a company’s sustainable practices and governance effectiveness, influencing investor trust and future growth potential, ultimately affecting stock prices. This study proposes an innovative approach that combines ESG sentiment index extracted from news with technical indicators to predict the S&P 500 index. By utilizing a deep learning model and exploring optimal window sizes, the study explores the best model through mean absolute percentage error (MAPE) as an evaluation metric. Additionally, an ablation test clarifies the influence of ESG and its causality with the S&P 500 index. The experimental results demonstrate improved predictive accuracy when considering ESG sentiment compared to relying solely on technical indicators or historical data. This comprehensive methodology enhances the advantage of stock price prediction by integrating technical indicators, which consider short-term fluctuations, with ESG information, providing long-term effects. Furthermore, it offers valuable insights for investors and financial market experts, validating the necessity to consider ESG for financial assets and introducing a new perspective to develop investment strategies and decision-making processes.

Understanding Intraday Oil Price Dynamics during the COVID-19 Pandemic: New Evidence from Oil and Stock Investor Sentiments

This study employed intraday stock market and oil investor sentiment data related to news and social media (i.e., the Thomson Reuters MarketPsych Indices [TRMI] sentiment index) to gauge investors’ interest in the West Texas Intermediate (WTI) crude oil futures market during the recent health crisis. We proposed an original nonlinear empirical framework by considering oil price dynamics’ complexity and its potential interaction with investor sentiment. The analysis revealed three noteworthy findings. First, we observed evidence of nonlinearity in the relationship between excess returns on WTI crude oil futures and investor sentiment data. Second, the causality direction moved only from oil and stock market investor sentiment to oil returns. Third, the impacts of oil and stock market sentiment data on crude oil returns (i.e., volatility) were always negative. Furthermore, sentiment data related to social media showed a more pronounced cross-correlation than that of news.

MSCI全球指数与韩国股指波动对韩国贸易的影响

This study aims to analyse the impact of the stock market on Korea’s import and export trade, mainly through the MSCI Global and KOSPI indices as stock market proxies, and Korea’s total imports, total exports, import price indices, economic sentiment indices, and business surveys indices as proxies for import and export trade. Using a time-varying vector autoregression (TVP-VAR) model, it was found that the MSCI Global and KOSPI indices have a short-term positive effect on Korea’s total imports and total exports, but the KOSPI has a negative effect on total exports. For price indices, the MSCI Global Index has a positive effect on the import price index, while the KOSPI has a negative effect at a 1-month lag. For the export price index, the MSCI Global Index has a negative impact at 1-month lag but a positive impact at 3-month lag. For the Economic Sentiment Index, both indices have a negative impact at the 1-month lag. For the business survey index, both have a positive impact. In addition, the study analyses data from three different points in time: the US subprime crisis in 2008, the European credit crisis in 2015 and the New Crown epidemic in 2020. The results show that in all three financial events, the MSCI Global Index had a positive impact on Korea’s total exports and imports in the short term, while the KOSPI had a positive impact on total imports and a negative impact on total exports. For the import price index, the MSCI Global Index has a short-term positive impact while the KOSPI has a negative impact. For the export price index, both indices have a negative impact during financial events. For the economic sentiment index, the MSCI Global Index has a negative impact in the first lag and turns to a positive impact in the second lag, while the opposite is true for the KOSPI. Finally, in all three financial events, both indices had a short-term positive impact on the business survey index.

Why Sentiment Analysis-Based Metrics are Essential for Measuring Channel Performance on YouTube: An Experimental Study

YouTube is a universal social medium that lets its users share videos and write comments on shared videos. Comments of the users on Youtube videos can be useful for YouTube channel owners, and it is worth analyzing them in terms of sentiment. The rate of like/dislike on a YouTube video is not sufficient for estimating the attitude of users towards it. This study proposes a three-step method for this attitude: In the first step, a sentiment analysis task is applied to the user comments on the videos of the YouTube channel "iJustin". In the second step, a new metric named Sentiment Index (SI) has been proposed and the Sentiment Indexes of the videos have been calculated. In the third step, an analysis is performed to show if the SI metric is time-independent. As a result, it was seen that most of the comments left on the videos (89%) have been written within the first 30 days after the video was published. Our experiments revealed that comments left more than 30 days after publishing the videos change the average SI values of the videos by only 0.4%, and in this respect, the SI metric is negligibly affected by the time parameter.

Investors’ Psych and Stock Return in Nigeria during the Covid-19 Pandemic Era

Investors’ attitudes were severely altered by the Covid-19 crisis, which had an intolerable effect on volatile returns in Nigerian stock market and other stock markets worldwide. Psychological trauma was experienced by stock investors due to the inability to foresee returns following the worldwide shock of Covid-19 and the subsequent drop in stock investments brought on by the unpredictability of macroeconomic behavior. This study analyzed the link effect of investor psyche on stock returns in Nigeria between March 2019 and December 2022. For this research, Sentiment Index (SMI) model was employed for the connection between investors psyche and returns on the Nigerian stock market. It was found that, investors psyche during Covid-19 pandemic negatively affected stock returns, resulting from negative concerns about their survival and financial security than the stock market, thus triggered sharp decline in stock investment. Thus, study concluded that Covid-19 outbreak impacted stock returns negatively. The research recommends that the Nigerian stock exchange commission should increase disclosure of market information, which is related to the magnitude and potentials of stock returns and degree of volatility at all times. The concerned volatility and potentials during the Covid-19 era were also influential to enhance investors’ confidence.

Research on the Impact of Investor Sentiment on IPO Underpricing of GEM

Abstract IPO underpricing appears when its closing price far exceeds issuance price on the first day of stock issuance, which is a common phenomenon all over the world. This paper first expounds the related concepts of investor sentiment and IPO underpricing, and then introduces the two perspectives of behavioral finance. Based on these two perspectives, this paper expounds the impact path of investor sentiment on IPO underpricing. Combined with the characteristics of investor sentiment, the proxy indicators of the composite indicators are selected, and the comprehensive indicators of investor sentiment are obtained by using the principal component analysis method, which are combined with the financial indicators that can reflect the market conditions. Taking the data of GEM stocks issued from January 2015 to December 2021 as the research sample, SPSS is used for stepwise regression analysis. From the results of the empirical research, there is a positive correlation between investor sentiment and IPO underpricing. Finally, this paper puts forward suggestions from two aspects: first, we should give correct and reasonable guidance to investors, and implement access policies; Second, in terms of macro policies, we should strengthen the audit of listed companies, establish an authoritative information disclosure mechanism, and put the expansion of GEM on the agenda. JEL classification numbers: G14. Keywords: GEM, Investor sentiment, IPO underpricing, Behavioral finance.

Product Demand Forecasting with Neural Networks and Macroeconomic Indicators: A Comparative Study among Product Categories

In the fiercely competitive global corporate arena, the intricacies of demand forecasting in the retail sector have become a focal point. While previous research has delved into various methodologies, it consistently overlooks the distinct performances of forecasting models within different retail product categories. Understanding these variations in prediction performances is pivotal, enabling firms to fine-tune forecasting models for each category. This study bridges this gap by scrutinizing the prediction performances of models tailored to different product categories. Building on recent research, we incorporate external macroeconomic indicators like the Consumer Price Index, Consumer Sentiment Index, and unemployment rate, alongside time series data of retail sales spanning various categories. This amalgamated dataset is employed to train a Long Short Term Memory model, projecting future demand across product categories. We further extend the analysis by identifying features that contribute most towards explaining product demand and quantifying their strength. The fitted models yield comprehensive insights into their performances and pinpoint the product categories warranting more focused model development.

Do topic and sentiment matter? Predictive power of online reviews for hotel demand forecasting

Studies integrating textual data for forecasting have mainly focused on the overall sentiment reflected in text. Yet textual data convey various types of information, such as review topics, that can be beneficial when forecasting hotel demand. This study aims to combine topic modeling and sentiment analysis to improve forecasting performance of hotel demand. Specifically, the latent Dirichlet allocation (LDA) topic modeling technique and the long short-term memory (LSTM) model are employed to construct topic-based sentiment indices. The autoregressive integrated moving average (ARIMA) with explanatory variable–type models and mixed data sampling (MIDAS) models are adopted for the evaluation of predictive power. Results reveal that MIDAS forecasting with topic–sentiment and COVID-19 variables generates most accurate forecasts. The findings contextualize the application of online textual big data in hotel demand forecasting research. Hotel management can utilize these online data for short-term forecasting to facilitate crowd management and respond more effectively to unforeseen public health events.

More is better? The impact of predictor choice on the INE oil futures volatility forecasting

This paper aims to address the predictor choice issue in forecasting volatility of INE oil futures by a comprehensive comparative study with a large number of predictive variables and applying machine learning models along with their interpretability tools. The main finding is that the selection of predictors is crucial for improving volatility forecasting accuracy, but it is not always the case that including more predictive variables leads to better forecasting results, even for machine learning models. Specifically, this paper has five major findings: (1) A few variables can significantly improve forecasting accuracy independently, but their contribution is limited. (2) Increasing the number of predictors from specific categories (market sentiment indicators, crude oil futures prices from other exchanges, and energy market indicators) helps to enhance forecasting accuracy. (3) Low-frequency variables have a weak effect on improving the daily volatility. (4) Ensemble tree models perform better than traditional machine learning models based on variable selection with dynamic parameter optimization, even without much parameter tuning. The above findings still hold true under a series of robustness tests and economic value assessments. These findings provide substantial evidence for addressing the issues of model and variable choice in crude oil futures volatility forecasting.

Systemic Financial Risk Forecasting with Decomposition–Clustering-Ensemble Learning Approach: Evidence from China

Establishing a scientifically effective systemic financial risk early warning model is of great significance for prudently mitigating systemic financial risks and enhancing the efficiency of financial supervision. Based on the measurement of systemic financial risk and the network sentiment index of 47 financial institutions, this study adopted the “decomposition–reconstruction–integration” approach, utilizing techniques such as extreme-point symmetric empirical mode decomposition (ESMD), empirical mode decomposition (EMD), variational mode decomposition (VMD), hierarchical clustering, fast independent component analysis (FastICA), attention mechanism, bidirectional long short-term memory neural network (BiLSTM), support vector regression (SVR), and their combination, to construct a systemic financial risk prediction model. The empirical results demonstrate that decomposing and reconstructing relevant indicators before predicting systemic financial risks can enhance prediction accuracy. Among the proposed models, the ESMD-HFastICA-BiLSTM-Attention model exhibits superior performance in systemic financial risk early warning.

An Index of Skilled Analysts and Stock Market Returns

ABSTRACT This study constructs a monthly index of skilled analysts (ISA) regarding the coverage by such analysts and explores their predictive ability on the future excess returns of Chinese A-share market. The empirical results reveal that ISA significantly and positively predicts excess stock market returns, with R 2 s of 4.39% and 4.13% in- and out-of-sample respectively, which surpasses the predictive ability of most commonly used economic variables. High ISA precedes a higher investor sentiment index (ISI) and lower idiosyncratic volatility (IVOL) at firm level. The conclusion remains valid after controlling for a series of other predictor variables and performing robustness tests.

Shedding light on consumer sentiments: evidence from India

ABSTRACT This paper shows how location-based indicators can predict consumer confidence in India. We capture local economic activity using city-wise night-time luminosity (NTL) data. Using data on unit-level observations on consumer confidence from the Consumer Confidence Survey by the Reserve Bank of India from June 2016 to November 2021, we find that NTL is positively associated with consumer sentiments. Our results are robust to alternative definitions of consumer sentiments. Finally, we extend our study to analyse the role of NTL on several individual components of household sentiments, such as household perception and outlook on household income, spending, employment and general price levels. Overall, our results provide fascinating insights about using NTL as a measure of local economic indicators and its implications on households’ sentiment indicators.

The impact of investor greed and fear on cryptocurrency returns: a Granger causality analysis of Bitcoin and Ethereum

PurposeThis study probes the psychological interplay between investor sentiment and the returns of cryptocurrencies Bitcoin and Ethereum. Employing the Granger causality test, the authors aim to gauge how extensively the Fear and Greed Index (FGI) can predict cryptocurrency return movements, exploring the intricate bond between investor emotions and market behavior.Design/methodology/approachThe authors used the Granger causality test to achieve research objectives. Going beyond conventional linear analysis, the authors applied Smooth Quantile Regression, scrutinizing weekly data from July 2022 to June 2023 for Bitcoin and Ethereum. The study focus was to determine if the FGI, an indicator of investor sentiment, predicts shifts in cryptocurrency returns.FindingsThe study findings underscore the profound psychological sway within cryptocurrency markets. The FGI notably predicts the returns of Bitcoin and Ethereum, underscoring the lasting connection between investor emotions and market behavior. An intriguing feedback loop between the FGI and cryptocurrency returns was identified, accentuating emotions' persistent role in shaping market dynamics. While associations between sentiment and returns were observed at specific lag periods, the nonlinear Granger causality test didn't statistically support nonlinear causality. This suggests linear interactions predominantly govern variable relationships. Cointegration tests highlighted a stable, enduring link between the returns of Bitcoin, Ethereum and the FGI over the long term.Practical implicationsDespite valuable insights, it's crucial to acknowledge our nonlinear analysis's sensitivity to methodological choices. Specifics of time series data and the chosen time frame may have influenced outcomes. Additionally, direct exploration of macroeconomic and geopolitical factors was absent, signaling opportunities for future research.Originality/valueThis study enriches theoretical understanding by illuminating causal dynamics between investor sentiment and cryptocurrency returns. Its significance lies in spotlighting the pivotal role of investor sentiment in shaping cryptocurrency market behavior. It emphasizes the importance of considering this factor when navigating investment decisions in a highly volatile, dynamic market environment.

Are national or regional surveys useful for nowcasting regional jobseekers? The case of the French region of Pays‐de‐la‐Loire

AbstactIn this paper we develop nowcasting models for the Pays‐de‐la‐Loire's jobseekers, a dynamic French regional economy. We ask whether these regional nowcasts are more accurate by only using the regional data or by combining the national and regional data. For this purpose, we use penalized regressions, random forest, and dynamic factor models as well as dimension reduction approaches. The best nowcasting performance is provided by the DFM estimated on the regional and regional‐national databases as well as the Elastic‐Net model with a prior screening step for which the national data are the most frequently selected data. For the latter, it appears that the Change in foreign orders in the industry sector, the OECD Composite leading indicator, and the BdF Business sentiment indicator are among the major predictors.

News-driven stock market index prediction based on trellis network and sentiment attention mechanism

Applying deep learning, especially time series neural network, to stock market prediction, has become one of the important applications in the quantitative finance field. However, due to the multi-correlation and volatility of the stock market, how to timely and accurately predict it has become a challenging issue. In order to cope with this challenge, a news-driven stock market index prediction model based on TrellisNet and a sentiment attention mechanism (SA-TrellisNet) is proposed. A sentiment analysis model based on CNN and LSTM is presented to obtain the sentiment index of massive news crawled from authoritative financial websites. Furthermore, a sentiment attention mechanism is designed for data fusion of stock data and news sentiment index as the input of the simple and efficient TrellisNet network for model training and prediction. The performance of our model is systematically evaluated using seven major international stock market indices including S&P500, NYSE, DJI, NASDAQ, FTSE 100, Nikkei 225 and SSE, and comparative experiments demonstrate that SA-TrellisNet is competitive to the other state-of-the-art methods in predicting stock market indices.

An Analysis of the Relationship between Business Confidence and Macroeconomic Factors: Türkiye Case

Investor sentiment plays an important role in the all level of economic activities. Confidence as an indicator of investor sentiment has become an important and interesting concept in recent years. Confidence indicators are a way in assessing economic developments in the short term. The aim of this study is to analyze the long and short term relationship between the Real Sector Confidence Index and macroeconomic/financial variables for the period of January 2007 – December 2021. The Johansen Cointegration Test results prove a long-term cointegration relationship. Vector Error Correction Models for both macroeconomic and financial variables are statistically significant at the 5% level. In the short run, Manufacturing Capacity Utilization Ratio, Official Reserves, Domestic Debt Stocks, Real Exchance Rate, US Dollar/TRL Buying Rate, Non-Residents’ Stock Portfolio and Volatility Index have a statistically significant effect on the changes in Real Sector Confidence Index. Thus, this study informs investors to predict the effects of changes in the confidence index on financial instruments bilaterally and contribute to their investment decision.

LEET: stock market forecast with long-term emotional change enhanced temporal model.

The stock market serves as a macroeconomic indicator, and stock price forecasting aids investors in analysing market trends and industry dynamics. Several deep learning network models have been proposed and extensively applied for stock price prediction and trading scenarios in recent times. Although numerous studies have indicated a significant correlation between market sentiment and stock prices, the majority of stock price predictions rely solely on historical indicator data, with minimal effort to incorporate sentiment analysis into stock price forecasting. Additionally, many deep learning models struggle with handling the long-distance dependencies of large datasets. This can cause them to overlook unexpected stock price fluctuations that may arise from long-term market sentiment, making it challenging to effectively utilise long-term market sentiment information. To address the aforementioned issues, this investigation suggests implementing a new technique called Long-term Sentiment Change Enhanced Temporal Analysis (LEET) which effectively incorporates long-term market sentiment and enhances the precision of stock price forecasts. The LEET method proposes two market sentiment index estimation methods: Exponential Weighted Sentiment Analysis (EWSA) and Weighted Average Sentiment Analysis (WASA). These methods are utilized to extract the market sentiment index. Additionally, the study proposes a Transformer architecture based on ProbAttention with rotational position encoding for enhanced positional information capture of long-term emotions. The LEET methodology underwent validation using the Standard & Poor's 500 (SP500) and FTSE 100 indices. These indices accurately reflect the state of the US and UK equity markets, respectively. The experimental results obtained from a genuine dataset demonstrate that this method is superior to the majority of deep learning network architectures when it comes to predicting stock prices.

Prediction of Economic Growth by Using Machine-Learning Algorithms through Sentiment Index Analysis in Economy of Pakistan

The neural network method's ability to mimic nonlinear communications without prior information is a major benefit for accurate function approximation. It differs from ARIMA and linear regression, which need linearity and may not work for all data sorts. Compared to (seasonal) ARIMA models, neural networks yield similar results. Neural networks also model and predict multivariate time series using causal and co-integration research. In contrast to the traditional economic forecasting approach, machine-learning techniques have prioritized precise forecasting without predetermined assumptions or conclusions. Owing to their enhanced predictive abilities and technological advancements, they provide more adaptability and have gained widespread acceptance in several fields. In terms of estimating US real estate values, machine-learning technology fared better than traditional econometric models, as shown. Total there were 131601 economic and financial news, which may include the phrases along with the names specified by the bank, and it may also include local media and mainstream. Machine learning uses the strength of data-driven algorithms and state-of-the-art computational techniques to improve decision-making in a range of industries. The findings of this research contribute to better forecasting practices and show how machine learning may provide more accurate forecasts in real-world scenarios.