MAKE MOST OF THE KNOWLEDGE NETWORK, JOIN ACADEMIC RESEARCH FOUNDATION

In this article, we examine the strict and second order periodic stationarities, the existence of higher order moments and the covariance structure of the periodic asymmetric power GARCH (p; q) process under general and tractable assumptions.

In this note, we statistically evaluate the performance of students participating in different in-person and online teaching formats. To this end, we introduce and prove the validity of a novel nonparametric bootstrap procedure based on matching estimators. Our approach supports various didactic evaluation methods as presented in the pedagogical literature. The results of the empirical analysis tend to show better performance of students attending lectures in the traditional in-person format. In general, students slightly prefer the in-person format, but are also in favour of a combination of in-person and online lectures. Based on these results, it seems appropriate to propose courses that combine both in-person and online teaching formats. However, in order to avoid student performance deficits, the planning and didactic approach of online lectures has to be adapted accordingly and requires further investigation.

Keywords: Bootstrap, Matching Estimators, Teaching-Learning Settings.

This paper presents a simulation study of a generalized Cox approach for modeling credit risk in the context of a firm exposed to extrem CO2 emissions. The study uses a Poisson process to model the random events associated with such excessive CO2 emissions, and a shot noise process to capture the impact of these emissions exceedances exceedances on the firm's hazard process. The simulations show the eectiveness of the generalized Cox approach in capturing the impact of extrem CO2 emissions exceedances on credit risk, and the sensitivity of the results to changes in the model parameters.

Keywords: credit risk, Emissions Trading System, GHGs, default proba-bility.

In this paper we propose a new approach to explore the stochastic frontier models which uses the power of Bayesian quantile regression. Compared with usual models based on regression in the conditional mean, our proposal inherits the advantages of quantile regression, such as robustness of estimators as it does not need to assume any distribution to the data nor assume homoscedasticity. Moreover, it also brings more details to the analyst since that several quantiles provide more information about the stochastic frontier. In addition, our proposal allows a better comparison of technical eciency estimation among rms through analysis of several quantiles in the stochastic frontier.

Keywords: Bayesian quantile regression, Gibbs sampling, Stochastic Frontier, Technical efficiency.

The impact of Coronavirus disease (COVID-19) is globally felt and understanding the spread or growth of the virus is one of the ways to flatten the curve of the virus. There is need to understand the spread of this virus in terms of future projections in other to put in adequate measures to curtail the virus. The performance of ARIMA and ARFIMA models in forecasting daily new cases of the disease in Lagos State, Nigeria, was evaluated in this study. The stationarity of the data was tested using the KPSS and ADF tests. To achieve stationarity, the data was subjected to integer and fractional differencing. ARIMA (2,1,1) and ARFIMA (1,0.79,1) were identified using the ACF and PACF plots. The adequacy of the identified models was assessed using the Ljung-Box Chi-Square test. The forecasting performance of both models was compared using Absolute Percentage Squared Error (APSE) and the results show that ARFIMA (1, 0.79, 1) model has a better forecasting performance.

This paper discusses the availability of digital transaction methods introduced by the government and banking sectors in India as steps toward achieving a cashless society. The data for this study is sourced from the Reserve Bank of India's website, covering the period from 2011 to 2022. The study focuses on the growth rate of digital transactions through NEFT and RTGS, analyzing the impact of demonetization and COVID-19 using dummy variable regression. Additionally, it provides predictions for the number and value of transactions in future years. The Compound Annual Growth Rate (CAGR) analysis is conducted using the inverse semi-logarithmic regression model. The study finds that NEFT and RTGS transactions experienced a significant annual growth rate of 31.46% and 8.93% respectively during the study period. Holt-Winters forecasting is utilized to predict the value of cashless transactions, estimating the values for NEFT in January and February 2023 to be approximately 27086 and 27353 billion rupees respectively. The forecast method demonstrates 97% and 98% accuracy for digital transactions via NEFT and RTGS respectively. The paper provides recommendations to stakeholders in India's cashless transaction ecosystem, including network providers collaborating with the government to expand coverage in remote and rural areas, enhancing network reliability through infrastructure upgrades, introducing specialized data packages for digital payment apps, partnering with payment service providers for bundled services, and ensuring secure transactions through encryption. Banks are also encouraged to promote digital payment adoption by offering incentives such as cashback rewards, reduced costs, and user benefits.