Research Article
Copula Analysis of Dependencies Between Extreme Exchange Rates and NSE20 Price Index
Rosemary Wanjiru Ng’ethe,
Thomas Mageto,
Joseph Mungatu
Issue:
Volume 9, Issue 3, June 2023
Pages:
50-59
Received:
12 July 2023
Accepted:
26 September 2023
Published:
1 November 2023
Abstract: Exchange rates within an economy affect international trade as they influence the price of goods and services sourced from another country and the attractiveness of the local produce to international consumers. This brings forth an interdependence relationship between exchange rates and market value of share products listed in Security Exchange Markets. This study evaluates the dependence structure between extreme exchange rates of the Kenyan Shilling against the US Dollar and the Nairobi Securities 20 price index using archimedean copulas. The Peak Over Threshold method was used to determine extreme values of the daily log returns of the KSH/USD exchange rate whose dependence structure was analyzed against the NSE20 price index. Parameter Estimation was via the Maximum log-Likelihood Estimation technique. Descriptive statistics showed that the minimum and maximum Ksh/Usd exchange rates were at Ksh. 79.44 and Ksh. 116.07 respectively. The highest NSE 20 price index was at Ksh.5500 with the lowest value of Ksh. 1724. This study found a negative correlation between Ksh/Usd extreme exchange rate data and the NSE20 price index. The Clayton copula was found as the best archimedean copula in modeling the dependence structure as it had the lowest standard error and a parameter estimate close to zero.
Abstract: Exchange rates within an economy affect international trade as they influence the price of goods and services sourced from another country and the attractiveness of the local produce to international consumers. This brings forth an interdependence relationship between exchange rates and market value of share products listed in Security Exchange Mar...
Show More
Research Article
Bayesian Spatio-Temporal Models for the Incidence of Malaria Using Time Dependent Covariates
Evalyne Nduvi Musyoka*,
Samuel Mwalili,
Boniface Malenje
Issue:
Volume 9, Issue 3, June 2023
Pages:
60-66
Received:
23 October 2023
Accepted:
3 November 2023
Published:
11 November 2023
Abstract: This research study focuses on the Spatial and temporal Modelling of malaria incidences in Kenya, taking into account Time- dependent covariates. Malaria remains a significant public health concern in Kenya, with varying rates of infection across its 47 counties. Environmental factors such as temperature, rainfall, humidity and elevation play a crucial role in influencing Malaria transmission. Despite numerous malaria control efforts and initiatives the burden of the disease persist. The main objective of this study was to formulate Bayesian Spatio-temporal models for malaria incidence, with a particular emphasis on incorporating time-dependent covariates. The availability of data collected over time from various counties, as provided by the malaria project Atlas, was essential for achieving this goal. The Besag-York-Molli ́e (BYM) Spatio-temporal Model were formulated and implemented using Bayesian approach. Bayesian inference technique, coupled with Markov Chain Monte Carlo (MCMC) algorithms, was used to fit the models to the data. We also conducted convergence diagnostic of MCMC algorithm in order to check if the algorithm has converged and how reliable the posterior estimates are. In the analysis under Bayesian model choice and comparison of spatio-temporal model, spatial model with time dependent covariates and Spatio-temporal model with time dependent covariate were fitted. We found out that Spatio-temporal model with Time Dependent covariates was the best model. The resulting model and maps will be valuable for identifying disease hotspots, allocating resources for disease prevention and mitigation, and guiding policy decisions to reduce the burden of malaria. To ensure the validity of the Bayesian analysis, MCMC diagnostics were applied, including the Geweke Test, Gelman-Rubin statistics, and trace plots. These tests confirmed that the MCMC chains had converged to a common distribution, indicating the reliability of the obtained results.
Abstract: This research study focuses on the Spatial and temporal Modelling of malaria incidences in Kenya, taking into account Time- dependent covariates. Malaria remains a significant public health concern in Kenya, with varying rates of infection across its 47 counties. Environmental factors such as temperature, rainfall, humidity and elevation play a cru...
Show More
