The study is an extensive examination that aims to forecast the critical factors impacting football match outcomes, with a particular focus on win/loss determination in the Indian Super League (ISL). The study makes use of three different machine learning algorithms: AdaBoost, Support Vector Machine (SVM), and Classification and Regression Trees (CART), using data from a total of 377 matches. The main goal of the study is to identify and select the most influential features that influence a match's outcome. Through the use of these algorithms, we want to increase forecast accuracy and offer insights into the crucial elements that may be strategically controlled to boost team performance. The study assesses each algorithm's prediction power and analyzes how well it performs in feature selection. The findings show notable differences in feature importance between the models, highlighting the advantages and disadvantages of each model. This work contributes to the wider use of machine learning in sports analytics while also deepening our understanding of football performance drivers.
The export of beauty and skincare products from India to international markets is rapidly growing with the rise in the global demand for natural, herbal, and Ayurvedic solutions. However, the beauty and skincare sector is volatile and unpredictable due to changing consumer preferences, beauty trends, global competition, and trade policies. Accurate forecasting of Indian beauty and skincare exports is essential for practitioners to optimize business strategies and maintain India’s competitive edge in this growing market. The present study compares the prediction performance of GRNN, MLP-ANN, and ARIMA methods in forecasting export volumes. The study utilizes historical export data from 2007 to 2024 to analyze trends and patterns in India’s beauty and skincare exports. The findings demonstrate that the machine learning model, GRNN, outperforms MLP-ANN and ARIMA, in capturing complex, non-linear data, resulting in more accurate and reliable export forecasts. This research provides valuable insights for policymakers, exporters, and businesses by offering precise predictions that can facilitate strategic decision-making, optimize supply chains, and support market expansion.
The study assesses the relative out-of-sample performance of different portfolio optimization strategies across four mean-risk frameworks and a benchmark naïve (1/N) strategy using weekly price data of a stock index, foreign currency, gold, natural gas, and crude oil from December 1997 to December 2023. For each mean-risk framework, we employ two optimization strategies: risk minimization and Sharpe ratio maximization. Using various risk-adjusted and economic measures, the out-of-sample performance analysis of all the strategies suggests that the Sharpe ratio maximization strategy of the mean-CVaR framework is the best performing model, while the variance minimization model of the mean-variance framework performs worst.
Accurate forecasts of geopolitical events are essential for security, foreign, and macroeconomic policy. Among human-based forecasting methods, predictions of collectives have established themselves as particularly accurate and useful. In particular, prediction polls and prediction markets have become well-studied and established methodologies. This article evaluates the discrimination and calibration of a prediction market on geopolitical events conducted in 2023 and 2024. It makes two contributions to the literature. First, it is the first article to provide evidence of the forecasting accuracy of a real-money prediction market on geopolitical events. Second, it provides one of the first comparisons of a prediction market’s forecasting accuracy with those of prediction polls for geopolitical events. This way, it contributes to a still small but growing literature that tries to establish the conditions under which prediction polls or prediction markets generate more accurate forecasts.
The National Collegiate Athletics Association (NCAA) introduced rule changes prior to the 2023–24 season. According to the Football Foundation Rules Committee, “…the most significant 2023 football rule changes involve adjustments to the timing and clock rules,” with the purpose to shorten the length of the game and to “…moderately reduce the number of plays per game.” We examine the impact of these changes on actual and expected scoring during the season. The rule changes led to lower average scoring, which was not fully encompassed at the start of the season in the financial (betting) market. The early season volatility in the market subsided as the season progressed resulting in a generally efficient market.
This study employs a panel data analysis to explore the determinants of cocaine and heroin prices within the European Monetary Union (EMU) from 2002 to 2021. Using economic and governance indicators, our approach provides a nuanced understanding of how governance affects drug market dynamics. The main objective of this study is to investigate and provide empirical evidence for the relationship between governance performance and the pricing of illicit drugs. Additionally, the study highlights that different aspects of governance have varying effects on specific types of drugs. The empirical evidence shows that stronger governance structures are associated with higher drug prices, as higher risk leads to higher prices. Moreover, the findings reveal that the rule of law impacts drug prices in general, while corruption specifically affects heroin prices. This research provides a unique contribution by linking governance performance directly to the pricing of illicit drugs within the context of the European Monetary Union. Unlike existing studies that focus predominantly on the medical, psychological, or criminal aspects of drug use, our study emphasizes economic and governance factors influencing drug prices, offering a novel perspective for policymakers and stakeholders in the fight against drug trafficking. To the best of our knowledge, this is the first model for illicit drug pricing.
Using a sample of 196 stocks, this study investigates the intraday market efficiency of the National Stock Exchange of India (NSE), a market that is entirely order-driven. The return autocorrelation and variance ratio tests suggest that the hourly returns of stocks at NSE are not serially correlated. Hourly order imbalances (OIBs) are highly persistent up to four lags, and can help in return prediction. Investors appear to follow short-horizon OIBs to conduct counter-vailing trades, and remove serial dependence in short-term returns. A simple order imbalance based trading strategy appears to offer abnormal returns; however, these returns vanish once the trading costs are factored in. Overall, the results indicate that the de-facto market making at NSE is effective.
This paper examines the price discovery process in the European Union Emission Trading Scheme (EU-ETS) – the largest carbon market across the world – for its third and fourth commitment periods. In particular, we examine the two leading carbon exchanges: European Energy Exchange (EEX: Spot and Futures) and European Climate Exchange (ECX: Futures). We examine the information transmission process in the EU-ETS for the three pairs, namely, (I) EEX spot-EEX futures, (II) EEX futures-ECX futures, and (III) EEX spot-ECX futures. To this end, we employ all three pair-wise bivariate vector error correction models (VECM) and price discovery measures, that is, component share (CS), information share (IS), and information leadership share (ILS) measures. We show that all three-price series substantially contribute to the price discovery. Moreover, the speed of adjustment and price discovery is comparable to the developed equity markets. The ability of carbon prices to incorporate the risk-premia related to climate-risk considerably depends on the pricing efficiency of carbon – one of the major objectives of the Kyoto Protocol and EU-ETS. Thus, these results have significant implications for policymakers, regulators, and academics in the forthcoming carbon markets from emerging economies (e.g., China, India).
Employing a unique NFL gambling dataset that includes both spread and money line data, we examine the disconnect in profitability between similar betting strategies across the two markets. If a naïve bettor wagered $110 on the favorite in every game against the spread (money line) he or she would, on average, lose $4.50 ($4.53) per game. Conversely, if the same bettor wagered $110 on the underdog every game he or she would, on average, lose $3.11 per game in the spread market, but lose less than 1 cent per game in the money line market. Further examination shows that a bettor could earn a 2.17% return betting the money line on the underdog if the closing spread was 7 or less and greater than 3, and 6.55% if the spread is 3 or less. As such, our results challenge the market efficiency of the NFL betting market and have important implications for sportsbooks and bettors.
This paper investigates the changes in financial assets and markets from December 1st, 2021, to April 30th, 2022, during the start of the Ukraine War. These dates roughly correspond to the prelude to the War in December 2021 to a few weeks after Russian troops withdrew from the Kyiv area on April 7th, 2022. We used the Goldstein 1992 Results Table to create Positive and Negative Geopolitical Risk bigrams (Goldstein, 1992). With these bigrams, we collected over 3.6 million tweets during our research period in seven different languages (English, Spanish, French, Portuguese, Arabic, Japanese, and Korean) to capture worldwide reaction to the Ukraine War. Using various sentiment analysis methods, we constructed a time series of changes in the daily Geopolitical Risk sentiment. We explored its relationship to 39 financial assets and markets at various time lags. We found through Granger causality that the geopolitical risk time series contained predictive information on several assets and market changes.
The real estate market plays an important role in the economies of many countries, and the future trend of the market has long been a topic of concern to both academics and practitioners. This paper attempts to study the effectiveness and superiority of two univariate time series models, Autoregressive Integrated Moving Average (ARIMA) model and exponential smoothing, to forecast the Macau residential property price index during the COVID-19 pandemic. Based on 1- and 2-year holdout samples during the shock of COVID-19 in Macau, the results show that the out-of-sample forecasting performances of both models are better than the baseline model of classical decomposition. There is also evidence that the ARIMA models outperform the Winters three-parameter exponential smoothing models in the two out-of-sample periods. Therefore, in the context of unprecedented events such as COVID-19, the ARIMA method is more effective than the Winters exponential smoothing method in making rapid and accurate adjustments when the Macau residential property price index is significantly affected. Our findings provide important implications for relevant government departments, home buyers and sellers, and property market participants in their selections of reliable models to forecast future property market behavior.
This paper analyzes the effect of firm-specific (FSD) and industry-specific determinants (ISD) on supply-chain-performance (SCP), export performance (EP) and SCP’s mediating effect on the relationship between FSD, ISD, and EP. It develops a theoretical framework from literature and empirically validates using the Indian automobile industry segments (IAIS) data. The sample frame consists of firms in ISIS between 2010–11 and 2020–21. The paper employs factor analysis for construct validity, panel-data-fixed-effect models to analyze the relationships, and bootstrap for cross-validation. It reveals that FSD and ISD directly influence both SCP and EP. SCP completely mediates the relationship between FSD, ISD, and EP.
Forecasting trends in stock indices is considered a difficult task in financial time series forecasting. Accurate forecasts of stock price trends can generate profits for investors. Due to the complexity of stock market data, developing effective forecasting models is very challenging. We are trying to predict stock prices for the next few days. This will become the basis for knowing the right time to invest or exit positions and generate profits. With the introduction of artificial intelligence and the increase in computing capacity, programmed forecasting methods have proven to be more effective at predicting stock prices. In this work, we used supervised machine learning algorithms such as linear regression model, SVR, XGBoost, and random forest. Thus, these models are evaluated using standard strategic indicators such as the EMR. A low value of the indicator shows that the models are effective in predicting stock prices.
This paper analyses the relationships between the volatilities of five major stock markets (S&P 500, CAC 40, DAX, FTSE 100, and Nikkei 225) and five cryptocurrencies (Bitcoin, Dash, Ethereum, Monero, and Ripple), (WTI), and gold. The GARCH model, which describes the volatility of financial assets and cryptocurrencies, was used. A significant and higher volatility spillover was observed across these market pairs. The conditional correlation between Bitcoin and other cryptocurrencies is time-varying, but the conditional correlations between crypto-currencies and gold and all assets are negative during the period (2017-2018) and positive. At the beginning of the COVID-19 crisis, the conditional correlation between cryptocurrencies, stock indices, and WTI increased, which confirms the impact of COVID-19 related contagion between them.Our findings show that cryptocurencies and gold are considered hedges for the international investors during the period 2017-2018.
This study conducts a comprehensive time series analysis of motor-vehicle fatalities in the USA spanning from 2019 to 2021, revealing a troubling upward trajectory. Factors such as over-speeding, impaired driving, reduced road traffic enforcement during the pandemic, and instances of driving under the influence have significantly contributed to the surge in fatal crashes during this period. Utilizing the Seasonal Autoregressive Integrated Moving Average (SARIMA) model, this research forecasts the trajectory of motor vehicle deaths in the USA. The forecast suggests a continuation of the upward trend, emphasizing the urgency of addressing the escalating fatalities. In response to the burgeoning global trend of increasing accidents and fatalities, this study advocates for the implementation of broader preventive measures worldwide. Proposed strategies encompass the crucial role of policy implementation and road safety measures in curbing the rising toll of road accidents, particularly in the USA.Furthermore, this study extends the existing 7E model (Education, Engineering, Enforcement, Exposure, Examination of Competence and Fitness, Emergency Response, and Evaluation) by introducing the eighth ‘E’—Empathy—in the context of road safety. This augmentation creates the 8E model, offering a more encompassing framework adaptable on a global scale. The inclusion of empathy underscores the significance of considering human emotions, behaviours, and societal impact in crafting effective road safety initiatives.
Scholars in the intersection of operational research, strategy, and finance have extensively examined the effects of event studies in finance, especially that of a strategic nature, such as that of planned as well as unexpected corporate events and respective abnormal returns on the stock market. Nonetheless, there is still a research gap on the extent of the forecastability of this abnormal behaviour, especially when predictions may provide crucial information to both investors and issuers, and therefore drive effectively investment decisions. In this study we forecast the value effect of SEOs and Stock Splits, across developed and emerging economies. The selection of these nations, namely the United States (benchmark), Brazil, and India, was based on their Gross Domestic Product (GDP) and the impact of their stock markets on economic growth. Data consist of 2,043 strategic financial decisions with historical information from the New York Stock Exchange (NYSE), Bombay Stock Exchange (BSE), National Stock Exchange of India (NSE) and Brazil Stock Exchange (B3) from 2010 to 2020. Linear regression (benchmark), random forests, gradient boosting machines, support vector regression and neural networks methods are empirically evaluated, with non-linear models performing better than the benchmark. A trading simulation is also incorporated to complement model outcomes and determine whether these predictions could be capitalised through effective decision making in the investment spectrum. Finally, the effects of the COVID-19 pandemic were also analysed for SEOs in the NYSE, and significant differences were discovered in March and April 2020. Results indicate how negative abnormal returns were exacerbated by COVID-19’s systemic impact during March and rebounded in April.
The rapid expansion of cryptocurrency trading has become a defining feature of contemporary financial markets, attracting a constantly growing group of participants, now surpassing 106 million worldwide. This research focuses on the psychological and behavioral foundations of trading behaviors, investigating how individual psychological states and lifestyle choices impact cryptocurrency trading activities. Using Ordinary Least Squares (OLS) regression, we examine the influence of various factors such as Loneliness, Negative Emotions, Fear of Missing Out (FOMO), Socialization, Healthy Lifestyle Habits, Entertainment Spending, and Sense of Achievement on the frequency of cryptocurrency trades. Our study also includes an analysis of gender differences through Levene’s T-test, thereby increasing the depth of our predictive model. The results of this study aim to fill a gap in existing literature by quantifying the degree to which individual psychological profiles and behaviors can predict trading activities, thereby providing detailed insights into the emotional and cognitive dimensions of the digital trading world. This research not only contributes to the field of behavioral finance but also provides a foundation for developing strategic interventions tailored to various trader segments, ultimately fostering a deeper understanding of the complex dynamics that characterize the crypto market’s volatile landscape.
Prediction markets are a form of collective intelligence that leverage market mechanisms to incentivise large numbers of individuals to make forecasts about future uncertain events. Since their origin in the 1980’s, they have been the subject of a small but steady stream of academic research. Proponents suggest that they have several advantages over comparable information aggregation mechanisms such as polls or expert groups. More recently the rise of blockchain, cryptocurrencies and decentralised finance (DeFi) has excited new interest in prediction markets. The characteristics of this triad of technologies has particular resonances with prediction markets. This research identifies the potential impact of blockchain technology on prediction market design and performance with a view to informing a research agenda to investigate those potential impacts.
Using an international sample of 60 green funds from 01 January 2010 to 31 December 2022, this study compares the managerial skills of green and matching conventional funds. Additionally, the study separately assesses how fund size and age affect the managerial abilities of both types of funds. The results suggest that fund’s green characteristic impacts managerial skills since green fund managers show better managerial skills than their conventional counterparts. Fund size influences managerial skills as small mutual funds are mainly responsible for the positive stock-selection skills in green and conventional fund managers. This paper also controls the results for the fund age effect and concludes that young and green funds possess better managerial skills than their older and conventional counterparts. The study reflects that green mutual fund, along with other green financial instruments, contribute to the global movement called “Sustainable Finance”, which aims to invest in public and private projects, therefore supporting mitigation and containment of climate change risks.
It is essential to recognize that dynamics of bankruptcy events vary across regions and legal frameworks. In this context, the paper aims to fill the critical gap in literature by presenting an analysis of machine learning (ML) models for early detection of bankruptcy probability among Indian companies operating under the Insolvency and Bankruptcy Code (IBC) of 2016. This study distinguishes itself by leveraging an extensive dataset covering the period from FY 2016 to FY 2022, encompassing 65,583 entries for 7,008 unique corporations, including 257 bankrupt entities. This paper employs various predictive variables, including traditional financial ratios, Altman Z-scores, and comprehensive financial statement data, employing a scenario-based approach over a one-year forecasting horizon. The findings support the notion that ML models, particularly XGBoost, outperform traditional logistic regression models and Altman Z-scores in accurately predicting bankruptcy among Indian corporates. These findings align with the trend in the literature favoring ML models for enhanced predictive power, offering valuable insights for financial institutions and policymakers in India’s corporate landscape.
