Mixed-frequency machine learning: Nowcasting and backcasting weekly initial claims with daily internet search volume data

We propose an out-of-sample prediction approach that combines unrestricted mixed-data sampling with machine learning (mixed-frequency machine learning, MFML). We use the MFML approach to generate a sequence of nowcasts and backcasts of weekly unemployment insurance initial claims based on a rich trove of daily Google Trends search volume data for terms related to unemployment. The predictions are based on linear models estimated via the LASSO and elastic net, nonlinear models based on artificial neural networks, and ensembles of linear and nonlinear models. Nowcasts and backcasts of weekly initial claims based on models that incorporate the information in the daily Google Trends search volume data substantially outperform those based on models that ignore the information. Predictive accuracy increases as the nowcasts and backcasts include more recent daily Google Trends data. The relevance of daily Google Trends data for predicting weekly initial claims is strongly linked to the COVID-19 crisis.     

Machine learning and the cross-section of emerging market stock returns

This paper compares various machine learning models to predict the cross-section of emerging market stock returns. We document that allowing for non-linearities and interactions leads to economically and statistically superior out-of-sample returns compared to traditional linear models. Although we find that both linear and machine learning models show higher predictability for stocks associated with higher limits to arbitrage, we also show that this effect is less pronounced for non-linear models. Furthermore, significant net returns can be achieved when accounting for transaction costs, short-selling constraints, and limiting our investment universe to big stocks only.     

Detecting the hidden asymmetric relationship between crude oil and the US dollar: A novel neural Granger causality method

The relationship between financial series is not always easy to detect due to their underlying asymmetry and nonlinearity. Both characteristics are not usually considered simultaneously, which may lead to many drawbacks in financial analysis. Hence, we develop a novel neural Granger causality method from both asymmetric and nonlinear perspectives and further revisit the response and impact of crude oil on the exchange rate. Our findings reveal the unidirectional nonlinear and asymmetric effect of crude oil on the exchange rate; that is, positive and negative oil prices can have a substantial impact on exchange rate shocks. Interestingly, this influence seems to strengthen after the Russia–Ukraine conflict. Besides, we also use simulation technology to evaluate the rationality and effectiveness of our proposed methods. Investors, policymakers, and scholars may be interested in our findings regarding the oil-dollar relationships; as well as interested in applying our methodology to other contexts.     

Neural networks for estimating Macro Asset Pricing model in football clubs

The recent crisis caused by COVID-19 directly affected consumption habits and the stability sof financial markets. In particular, the football industry has been hit hard by this pandemic and therefore has more volatile stock prices. Given this new scenario, further research is needed to accurately estimate the value of the shares of football clubs. In this paper, we estimate an asset pricing model in football clubs with different compositions of risk nature using non-linear techniques of artificial neural networks. Usually, asset pricing models have been estimated with linear methods such as ordinary least squares. Our results show a precision higher than 90% for all the estimated models, which far exceeds those shown by linear methods in the previous literature. We find that the residual represents about 40% of the variance of the price-dividend ratio. Long-term risks follow in importance, and above all, the habit component and its behaviour in the face of changes. The importance of the residual component exists due to a low correlation between the asset price and consumer behaviour, but to a much lesser extent than that shown in previous studies. The estimation carried out with artificial neural networks, both the Deep Learning methods and especially the Quantum Neural Network, opens up new possibilities to estimate more efficiently the pricing of financial assets in the football industry.     

Data-driven forecasting for operational planning of emergency medical services

Emergency medical services (EMS) play a vital role in delivering pre-hospital care. The operational efficiency of such services is critical and adequate demand forecasts can contribute to such a goal. But for that, the available data need to be well characterized before being used. Previous studies have failed to address some important aspects of this need, such as exploring a comprehensive list of contextual data to decide which are relevant to explain the EMS demand behavior. Moreover, modern forecasting techniques have been explored in the EMS context, including neural networks, but the computational complexity inherent to the methods and their use was not discussed. Finally, it is also unclear how different demand patterns can be when predicting the volume of emergency calls considering the priority level and the number of dispatches according to vehicle type. This study proposes a generic data-driven forecasting method to address these shortcomings and to support operational decisions. The results obtained with the proposed method indicate that each priority call and vehicle type shows different patterns, which suggests that such differentiation should contribute to better resource allocation. At the same time, the operational impact of the demand shared by neighboring zones proved to be significant at bases near the border. The models developed resulted in important decision tools that can be used to predict the dynamic demand of EMS on an hourly or shift basis. Additionally, the method adds value for decision-makers that want to plan not only when and how many but also where resources are demanded, avoiding assumptions that impact the operational performance.     

The use of ICTs and income distribution in Brazil: A machine learning explanation using SHAP values

This study explores the complex relationship between information and communication technologies (ICTs) and socioeconomic characteristics. We employ a cutting-edge explainable machine learning approach, known as SHAP values, to interpret an XGBoost and neural network model, as well as benchmark traditional econometric methods. The application of machine learning algorithms combined with the SHAP methodology reveals complex nonlinear relationships in the data and important insights to guide tailored policy-making. Our results suggest that there is an interaction between education and ICTs that contributes to income prediction. Furthermore, level of education and age are found to be positively associated with income, while gender presents a negative relationship; that is, women earn less than men on average. This study highlights the need for more efficient public policies to fight gender inequality in Brazil. It is also important to introduce policies that promote quality education and the teaching of skills related to technology and digitalization to prepare individuals for changes in the job market and avoid the digital divide and increasing social inequality.     

Growth vs value investing: Persistence and time trend before and after COVID-19

Value investing and growth investing allow economic experts to adopt different investment strategies depending on their chosen specialty; the two investment types have been conditioned by the pandemic, changing the trend of investments and their results. This research aims to analyze the behavior and trends of the different investment strategies before and after the health crisis. We use methodologies based on fractional integration and cointegration to analyze the persistence and trend of the series and their relationship in the long run. We find that the shock is long-lived and causes a change in trend; however, we find no evidence of mean reversion. In addition, we use multivariate wavelet analysis to analyze the correlation between both time series, concluding that a growth-based investment strategy is more successful than a value-based investment strategy. We use neural networks to corroborate our results.     

基于城市轨道交通网络的广告营销站点选择研究

城市轨道交通网络背景下的广告营销已成为一种流行的产品推广方式，快速、准确筛选出传播范围广、影响力高的站点，运营方和投资商对其投入大量广告可以产生较高的宣传力度和回报收益。影响力最大化的提出，为解决站点筛选提供了一种新的思路，首先提出一种融合节点结构信息和网络拓扑结构信息的网络表示学习模型，并将其用于学习网络节点的特征向量；进一步将网络节点的特征向量输入聚类算法中有效筛选候选种子集，同时结合贪婪策略从候选种子集中筛选出种子节点集合，进而有效提升网络节点影响力最大化问题求解效果。以武汉地铁网络为例进行验证，结果表明所提方法在传播范围上优于现有的影响力最大化方法，筛选高影响力站点具备合理性。