Kaggle forecasting competitions: An overlooked learning opportunity

We review the results of six forecasting competitions based on the online data science platform Kaggle, which have been largely overlooked by the forecasting community. In contrast to the M competitions, the competitions reviewed in this study feature daily and weekly time series with exogenous variables, business hierarchy information, or both. Furthermore, the Kaggle data sets all exhibit higher entropy than the M3 and M4 competitions, and they are intermittent.In this review, we confirm the conclusion of the M4 competition that ensemble models using cross-learning tend to outperform local time series models and that gradient boosted decision trees and neural networks are strong forecast methods. Moreover, we present insights regarding the use of external information and validation strategies, and discuss the impacts of data characteristics on the choice of statistics or machine learning methods. Based on these insights, we construct nine ex-ante hypotheses for the outcome of the M5 competition to allow empirical validation of our findings.     

Optimal age- and sex-based management of the queue to ventilators during the Covid-19 crisis

Triage protocols for intensive care units are based on priorities assigned to presents, but ignore patients about to arrive, so a priority newcomer may not find a ventilator and its associated nursing staff available because they are occupied by a lower-priority patient who however was present at the moment of assignment. Conversely, waiting too long leads to losing elderly patients who could have been saved by ventilators. As age and sex are major determinants of mortality by Covid-19 and have the merit, in contrast to other priority criteria, of being immediately available to health professionals, the criterion is the minimization of the mean mortality rate weighted by age- and sex-specific life expectancies. The dynamics is a queuing process involving mortality and return home flows and competition between ages. The result is the determination of an optimal threshold age that can guide triage.     

Border tax adjustments and tariff‐tax reforms with consumption pollution

We develop a model of a small open economy, where pollution per unit of consumption between domestically produced and imported quantities of the same good differs. We show that the first‐best policy combination calls for consumption taxes on all polluting goods, and border tax adjustment (BTA) measures, that is, tariffs or import subsidies. We identify conditions under which well‐known tariff‐tax reform policies for developing economies, such as a consumer‐price‐neutral piecemeal reform of trade and a consumption tax, and a consumer‐price‐neutral reform of all trade and consumption taxes improve welfare. We also evaluate whether reforms of trade taxes alone are superior to consumer‐price‐neutral reforms of trade and consumption taxes.     

Passport,please! Travels,travails and trade

A country’s visa policies are widely assumed to have economic consequences. In this short paper, I examine the effect of the ease with which a country’s citizens can enter foreign countries on international trade. Using a specification of the gravity model that avoids the endogeneity problems that typically arise when analyzing the association between ease of travel and the extent of bilateral interactions, I find that countries which issue powerful passports experience more international trade.     

CO‐OP OPEN MEMBERSHIP: ECONOMIC AND FINANCIAL EFFECTS

This paper aims to analyze the objectives pursued by cooperatives upon opening doors to new members and how this affects business activity and financial indicators. Surprisingly, the results show that accepting new partners makes no positive impact on the return on assets, but it does make a variable impact on financial indicators according to the type of cooperative. Distinguishing between agricultural and worker co‐ops, we conduct a cross‐sectional study of a sample of Galician cooperatives to find whether they apply this principle the same way regardless of membership size. Our results corroborate that cooperatives apply the principle differently. This not only allows us to extract other relevant information from accounting for cooperatives, but it also permits other agents like financial entities to obtain indicators that reflect the true company image more adequately.     

Sentiment and bias in performance evaluation by impartial arbitrators

This study investigates whether relative performance evaluations of labor output are biased in the presence of sentiment, even when the (supposedly independent) evaluators are external. Data from a field-experiment setting—involving a pro-sport League's best-player award—allows for empirical testing of this proposition. After controlling for within-match performance, the results show that the match officials provide significantly better evaluations to players celebrating a ‘milestone’ game—an occasion on which sentimental favorites exist. However, this sentiment bias holds only in certain cases, and appears to have weakened in more recent years.     

跨境资本流动对货币政策有效性的影响研究——基于系统性风险视角的分析

日益频繁的跨境资本流动势必引发系统性风险,进而对货币政策有效性造成冲击,严重威胁我国的经济金融稳定。本文构建四部门DSGE模型并利用脉冲响应函数及模拟技术检验跨境资本流动对我国货币政策有效性的影响。研究发现,跨境资本流动加速了我国系统性风险累积过程,价格型与数量型货币政策的有效性均受到削弱。因此,随着我国经济国际化进程的不断加深,仅仅依靠传统货币政策已经无法有效应对系统性风险,需要采取更多样的措施。     

Job Security and Risk‐Taking: Theory and Evidence From Professional Football

This article investigates how job security impacts risky decision making. In a theoretical model, we show how risk‐taking can be affected by job security. Agents with moderate job security become more risk averse as their job security increases. Conversely, agents with very high (or low) job security act in a more risk neutral manner. We test these predictions using data on head coaching decisions from the National Football League, finding that job security has a negative and statistically significant effect on risk‐taking.     

Detecting Accounting Fraud in Publicly Traded U.S. Firms Using a Machine Learning Approach

We develop a state-of-the-art fraud prediction model using a machine learning approach. We demonstrate the value of combining domain knowledge and machine learning methods in model building. We select our model input based on existing accounting theories, but we differ from prior accounting research by using raw accounting numbers rather than financial ratios. We employ one of the most powerful machine learning methods, ensemble learning, rather than the commonly used method of logistic regression. To assess the performance of fraud prediction models, we introduce a new performance evaluation metric commonly used in ranking problems that is more appropriate for the fraud prediction task. Starting with an identical set of theory-motivated raw accounting numbers, we show that our new fraud prediction model outperforms two benchmark models by a large margin: the Dechow et al. logistic regression model based on financial ratios, and the Cecchini et al. support-vector-machine model with a financial kernel that maps raw accounting numbers into a broader set of ratios.