International Journal of Uncertainty, Fuzziness and Knowledge-Based Systems

Some qualitative studies have mentioned the impact of trading networks on the credit levels of countries or businesses. However, only a few studies quantified network information on simulated or actual data. This research studied the relationship between network characteristics and sovereign credit levels. Using a data set for 114 countries, we developed and compared the performance of five credit classification models with and without integrating network information.

We used several measures to quantify the trading network information, such as Trading Weight, Closeness, Betweenness, PageRank, and Modularity. The analysis showed that they were closely related to credit levels. In particular, the Trading Weight usually belonged to the top five critical features for almost all classifiers. Overall, we saw that the correct prediction of the network models was consistently higher than those without a network. It meant we could improve the sovereign credit classifications with available trading network information.

Benford’s Law is an interesting and unexpected empirical phenomenon — that if we take a large list of number from real data, the first digits of these numbers follow a certain non-uniform distribution. This law is actively used in economics and finance to check that the data in financial reports are real — and not improperly modified by the reporting company. The first challenge is that the cheaters know about it, and make sure that their modified data satisfies Benford’s law. The second challenge related to this law is that lately, another application of this law has been discovered — namely, an application to deep learning, one of the most effective and most promising machine learning techniques. It turned out that the neurons’ weights obey this law only at the difficult-to-detect stage when the fitting is optimal – and when further attempts attempt to fit will lead to the undesirable over-fitting. In this paper, we provide a possible solution to both challenges: we show how to use this law to make financial cheating practically impossible, and we provide qualitative explanation for the effectiveness of Benford’s Law in machine learning.

The a priori procedure (APP) provides minimum sample sizes to meet specifications for precision and confidence that, in turn, can enhance trust that the sample statistics to be obtained provide good estimates of corresponding population parameters. A recent advance by Chen et al.³ showed how to perform the APP with respect to Cohen’s $d$, which is the most popular effect size measure in the social sciences. However, a limitation is that Chen et al.³ assumed normal distributions that compose only a subset of the much larger family of skew normal distributions, which are used as distributions of random errors in many economical models such as stochastic frontier models that are popular and useful models to investigate production efficiency in microeconomics, see Hung.⁵ Therefore, it would be useful to generalize to skew normal distributions. We perform the mathematical derivations for the generalization and support them with simulations and worked examples. Finally, we provide a free and user-friendly computer program by which researchers can perform the APP to determine minimum sample sizes necessary to meet specifications for precision and confidence, with respect to Cohen’s $d$, for a matched sample under the large umbrella of skew normal distributions.

This paper revisits the causal relationship between education and earnings, or the returns to education, using Thailand’s socio-economic household survey data. We show that, with the minimum modeling assumption, the causal effect of education on earnings can be nonparametrically identified only when there is no unobserved ability affecting both education and earning, and when the selection mechanisms are conditionally independent to earnings conditioning on potential experience and education. The causal effects of education on earnings are estimated using a nonparametric approach with selection bias adjustment and parametric approaches with selection bias and ability bias adjustments. The results show different estimates for returns to education across model specifications, indicating the importance of the bias adjustments and parametric assumptions.

In many practical situations, we observe a special Burr probability distribution — e.g., Burr distribution describes the distribution of people by income, it describes the rainfall data, etc. In this paper, we provide a theoretical explanation for the ubiquity of Burr distributions: namely, we show that this distribution naturally follows from scale-invariance. To be more precise, the simplest distribution that can be obtained from scale invariance is the Pareto distribution — a frequent particular case of the general Burr family of distributions. Next simplest are all distributions from the Burr family. We also use the scale-invariance approach to come up with a more general class of distributions that will, hopefully, provide an even more accurate description of the corresponding phenomena.

In many practical situations, we need to change the spatial distribution of some goods. In such situations, it is desirable to minimize the overall transportation costs. In the 1-D case, the smallest transportation cost of such a change is proportional to what is known as the Wasserstein metric. The same metric can be used to describe the distance between two probability distributions. In the last decades, it turned out that this metric can be successfully used in many economic applications beyond transportation. These successes are somewhat of a mystery: in principle, there are many different metrics on the set of all probability distributions, and it is not immediately clear why namely Wasserstein metric is so successful. In this paper, we show that the Wasserstein metric naturally appears in decision making situations. This fact explains the usefulness of the Wassterstein’s metric in economic applications.

In general, the probability of the next state depends not only on the current state, but also on the whole pre-history. However, in practice, many real-life processes — including economic processes — can be well described by so-called Markov models, i.e., models in which the probability of the next state depends only on the current state. In this paper, we explain the empirical ubiquity of such models.

Several stochastic frontier production analysis problems are not linear regression problems; two error components (inefficiency and noise) may not have symmetric dependence. In nonlinear problems, the data are divided into two or more regimes according to turning points or change points; the relationships between the output and all the inputs are allowed to differ between regimes. In the problem of two-error dependence, the inefficiency component generally plays a more dominant role than the noise component, thereby the output of each firm is mainly influenced by the inefficiency; in such cases, the symmetric dependence is not suitable. It is, therefore, desirable to come up with a new nonlinear asymmetric copula-based stochastic frontier model (SFM). Two nonlinear structures, namely kink and threshold and skew-normal copula, are suggested to SFM to deal with these problems. The Monte Carlo simulations and analysis of a real data set are employed to evaluate the accuracy and performance of proposed models. The results show that the proposed models show a higher performance when compared with the conventional linear SFM with symmetric two-error dependence.

Forecasting volatility is an essential task in the financial market, especially in portfolio optimization. To improve the prediction accuracy of the volatilities of assets we use a hybrid ANN-EGARCH model then combining with extreme value theory and Copula models to perform out-of-sample forecasting returns for six indices in Asia stock markets then we simulate one-day-ahead returns of these indices. We use EGARCH model to capture the leverage of return shocks due to COVID-19. Based on ANN-EGARCH-EVT-Copula models, we solve our portfolio optimization consisting of these six indices with different copula models. Using different performance measures to evaluate the efficiency of the models we show that under the Sharpe ratio and Sortino ratio the Gumbel copula gives better performance whereas with Average Drawdown and Max Drawdown measures, the Gaussian copula model is a best model for optimizing the portfolio.

This paper proposes a new three-parameter inverse Weibull-type distribution via a similar way of constructing a skew normal distribution. The additional shape parameter increases the distribution’s modelling capability as other three-parameter inverse Weibull-type distributions. The mathematical properties of this inverse extended Weibull distribution are studied and proven. In the empirical study of four insurance datasets from Australia, Denmark and the USA, we show that the proposed distribution performs very well compared with other three-parameter competitors.