Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Charles Chimedza has a total of 15 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Determinants of Women‘s Age at First Marriage in Zimbabwe: A Survival Analysis Approach

    by Evans Manjoro, Charles Chimedza, Isheunesu Munyira

    The study aims to examine the effect of select socio-demographic factors on women’s age at first marriage. The study employs survival analysis, specifically the Cox proportional hazard regression model and analysis of deviance using the Zimbabwe’s 2022 anonymised population and housing census data. The median women’s age at marriage in Zimbabwe is 21 years. Log rank tests show that there are statistically significant differences on median age at first marriage on all factor levels of the seven independent variables. Cox regression indicates that all the seven independent variables significantly influence women’s age at marriage. The most contributing factors to the age at first marriage are whether women had given live birth, province and level of education. The findings of this study suggest that policies that delay entry into first marriage must be province (region) based. Findings also suggest that improvement in educational attainment for women will play a significant role in delaying entry into first marriage.

  • Open Access

    Exploring the Power of Content and Visitor Sentiment: A Study of Web Traffic Dynamics in South Africa’s Residential Real Estate Landscape

    by Kola Ijasan, Charles Chimedza

    The real estate sector has increasingly shifted toward digital platforms, where content sentiment plays a crucial yet understudied role in driving user engagement. While sentiment analysis has been widely applied in retail and finance, its impact on real estate web traffic remains poorly understood, particularly in competitive digital marketplaces. This study examines the relationship between sentiment in web content and the traffic it attracts on residential real estate websites in South Africa. Specifically, it examines how different sentiments associated with the type of content (articles versus property listings) influence total monthly web traffic and user engagement. A quantitative analysis of six years (2017–2023) of scraped data from Property24, Remax, and Private Property employed R (rvest, sentimentr, and stats packages) for web scraping, sentiment analysis, and ANOVA testing to evaluate relationships between content sentiment, type (listings vs. articles), and web traffic metrics. The analysis revealed a significant impact of sentiment on web traffic, indicating that the sentiment of web content influences visitor numbers. Specifically, property listings generated a total of 16,780,623 monthly visitors, significantly surpassing the 13,407,521 visitors attracted by articles. This study contributes empirical evidence regarding the influence of content sentiment and content type on web traffic within the South African real estate market. It highlights the critical role of sentiment in shaping web traffic and potentially user engagement and provides actionable insights for real estate developers and marketers seeking to optimize their content strategies to improve user attraction and retention.

  • Open Access

    Minimum Covariance Distance-Based SMOTE Approaches for Zero-Inflated Datasets with High-Dimensional Heterogeneous Features in Big Data Analytics

    by Keith R Musara, Edmore Ranganai, Charles Chimedza, Florance Matarise, Sheunesu Munyira

    Big data in the credit risk landscape is often characterized by zero-inflated datasets, heterogeneity, and high dimensionality. These data aberrations adversely diminish the computational efficacy of the conventional predictive classifiers. To ensure accurate and reliable predictions, it is crucial to remedy these aberrations, as they may result in bias towards the majority class, sparsity, and computational complexity. The modified Euclidean distance (MED)-based synthetic minority oversampling technique (SMOTE) approaches have been suggested in contemporary literature as countermeasures for zero-inflated datasets coupled with heterogeneity. Despite their mathematical tractability, these approaches substantially fail to effectively capture correlations and variability among features. They are also susceptible to heavy-tailed error distributed data points (outliers) and collinearity, rendering them computationally suboptimal in high-dimensional data spaces. In this study, authors present a novelty of supplanting the MED with modified Mahalanobis distance (MMD) to the variants of SMOTE, enhancing their ability to adequately capture correlations, variability, and heterogeneous features. To mitigate the intricacies posed by these multifaceted data aberrations in high-dimensional data settings, the authors propose the fast minimum regularized coefficient determinant (FMRCD) approach to estimate the parameters of the MMD measure. Therefore, this paper enhances the robustness and computational efficiency of SMOTE-based approaches, by leveraging MMD computed intrinsically to the FMRCD approach, in conjunction with classical predictive classifiers. The empirical evidence suggests that our novelty, demonstrates superior predictions and offers computational stability edge over traditional approaches. These contributions circumvent overwhelming data complexities presented by zero-inflated datasets combined with high-dimensional heterogeneity in modelling big data phenomena.

  • Open Access

    The Occupational Indemnity Insurance Modelling: Brighton Mahohoho XGBoost Probabilistic Automated Actuarial Reserving-Pricing-Underwriting

    by Brighton Mahohoho, Charles Chimedza, Florance Matarise, Sheunesu Munyira

    This paper introduces the IFRS 17-Compliant Brighton Mahohoho Probabilistic Framework for Inflation-Adjusted Frequency-Severity Modeling in Occupational Indemnity Insurance, integrating AI-driven actuarial methodologies for loss reserving, risk pricing, and underwriting. Objectives: The framework ensures IFRS 17 compliance while enhancing actuarial accuracy and operational efficiency. Methods/Analysis: A simulation-based dataset of policy, claims, premiums, inflation adjustments, and underwriting data is generated. Claim frequencies and severities are modeled using Poisson and Gamma distributions, with inflation adjustments incorporated into reserves. XGBoost is applied for Automated Actuarial Loss Reserving (ALR) and Automated Actuarial Risk Pricing (ARP), while a weighted average approach estimates Automated Actuarial Loss Reserve Risk Premiums (AALRRP). Findings: Model accuracy is validated through MAE, MSE, RMSE, residual analysis, and scatter plots. IFRS 17 metrics—Contractual Service Margin (CSM), Fulfillment Cash Flows (FCF), Risk Adjustments, and Liabilities—are simulated, with sensitivity analysis ensuring robustness. Policyholders are segmented into underwriting clusters, incorporating expenses, outgo, and revenue to derive the Automated Net Actuarial Underwriting Balance (ANAUB). Novelty/Improvement: This integrated AI-driven actuarial framework significantly advances IFRS 17-compliant pricing and reserving, offering enhanced predictive accuracy, regulatory alignment, and improved risk assessment in occupational indemnity insurance.

  • Open Access

    Robust Distance-Based SMOTE Approaches for Skewed Fat-Tailed Distributed Datasets with Heterogeneous Features

    by Keith R. Musara, Edmore Ranganai, Charles Chimedza, Florence Matarise, Sheunesu Munyira

    Skewed fat-tailed distributed (class-imbalanced) datasets with inherent heterogeneous features often induce overwhelming aberrations in numerous machine learning (ML) algorithm applications, particularly in credit risk assessments. To mitigate these complexities, data-level (DL) approaches have been suggested in the literature as remedies. Notably, the benchmarks of DL approaches in the literature consist of variants of the synthetic minority oversampling technique (SMOTE) customized to effectively handle various data aberrations using distance measures. To counter the limitations of the Euclidean distance (ED) in capturing heterogeneous features, their computational efficiency is enhanced through the utilization of the modified ED (MED). However, the MED substantially fails to account for correlations between features. To circumvent these shortcomings, some studies have suggested using the Mahalanobis distance (MD), which accurately accounts for correlations in features. Nevertheless, this measure considerably falls short due to its inability to capture heterogeneous features, as well as its susceptibility to outliers. Therefore, this study proposes a modified Mahalanobis distance (MMD) designed to adequately capture heterogeneous features. Additionally, the MMD parameters were estimated using the median absolute deviation (MAD), which efficiently discards outliers. The study evaluated the computational efficiency of SMOTE-based approaches combined with the MMD, computed intrinsically to MAD approach, using the most widely employed ensemble-based ML approach, the random forest (RF) algorithm. The findings indicate that our novelty significantly outper-formed conventional approaches. This investigation enriches the understanding of generalizable predictive performance in the credit risk landscape.

  • Open Access

    The Advanced Actuarial Data Science Based AI-Driven Solutions for Automated Loss Reserving Under IFRS 17 in Non-Life Insurance

    by Brighton Mahohoho, Charles Chimedza, Florance Matarise, Sheunesu Munyira

    This study introduces an AI-driven Automated Actuarial Loss Reserving Model (AALRM) designed to meet IFRS 17 standards for non-life insurance. The model leverages advanced machine learning techniques to improve accuracy, efficiency, and adaptability in loss reserves, with a specific focus on inflation-adjusted frequency-severity modeling. A unique aspect of this research is the integration of bancassurance services, enabling automated management for both microfinance and car insurance on a unified platform. This includes a no-claims bonus system that categorizes policyholders into four tiers—base, variable, final, and high-bonus—resulting in more precise risk assessments and enhanced customer retention. Among eight evaluated machine learning algorithms, the Random Forest (RANGER) outperformed others for estimating Aggregate Comprehensive Automated Actuarial Loss Reserves (ACAALR). The model’s effectiveness was validated through stress tests, scenario analyses, and comparisons with traditional methods like the Chain Ladder. Additionally, the study introduces a novel Robust Automated Actuarial Loss Reserve Margin (RAALRM) with adaptive bounds, addressing traditional limitations in reserve margin calculations. This AI-integrated approach significantly improves predictive accuracy, operational efficiency, and strategic decision-making, offering a scalable solution for the insurance industry.


  • Open Access

    A Comparative Modelling of Essential Characteristics of Volatility: Simulation and Empirical Study

    by Richard T. A. Samuel, Charles Chimedza, Caston Sigauke

    This study utilised the dynamics of five time-varying models to estimate six essential features of financial return volatility that are relevant for robust risk management. These features include pronounced persistence, mean reversion, leverage effect or volatility asymmetry, conditional skewness, conditional fat-tailedness, and the long memory behaviour of volatility decomposition into long-term and short-term components. Both simulation and empirical evidence are provided. Through the applications of these models using the S&P Indian index, the study shows that the market returns are characterised by these volatility features. Our findings from the long-memory behaviour revealed that although the response to shocks is greater in the short-term component, it is however short-lived. On the contrary, despite a high degree of persistence in the long-term component, market information or unexpected news arrival only has a low long-run impact on the market. Based on this, the long-run investment risks within the Indian stock market seem to be under control. Hence, our findings suggest that rational investors should try to stay calm with the arrival of unexpected news in the market because the long-run effect of such news will not be severe, and the market will eventually return to its normal state.

  • Open Access

    Simulation Framework to Determine Suitable Innovations for Volatility Persistence Estimation: The GARCH Approach

    by Richard T. A. Samuel, Charles Chimedza, Caston Sigauke

    This study rolls out a robust framework relevant for simulation studies through the Generalised Autoregressive Conditional Heteroscedasticity (GARCH) model using the rugarch package. The package is thoroughly investigated, and novel findings are identified for improved and effective simulations. The focus of the study is to provide necessary simulation steps to determine appropriate distributions of innovations relevant for estimating the persistence of volatility. The simulation steps involve “background (optional), defining the aim, research questions, method of implementation, and summarised conclusion”. The method of implementation is a workflow that includes writing the code, setting the seed, setting the true parameters a priori, data generation process and performance assessment through meta-statistics. These novel, easy-to-understand steps are demonstrated on financial returns using illustrative Monte Carlo simulation with empirical verification. Among the findings, the study shows that regardless of the arrangement of the seed values, the efficiency and consistency of an estimator generally remain the same as the sample size increases. The study also derived a new and flexible true-parameter-recovery measure which can be used by researchers to determine the level of recovery of the true parameter by the MCS estimator. It is anticipated that the outcomes of this study will be broadly applicable in finance, with intuitive appeal in other areas, for volatility modelling.


  • Open Access

    Framework for Simulation Study Involving Volatility Estimation: The GAS Approach

    by Richard T. A. Samuel, Charles Chimedza, Caston Sigauke

    In econometrics and finance, volatility modelling has long been a specialised field for addressing a variety of issues pertaining to the risk and uncertainties of an asset. This study presents a robust framework, through a step-by-step design, that is relevant for effective Monte Carlo simulation (MCS) with empirical verifications to estimate volatility using the Generalized Autoregressive Score (GAS) model. The framework describes an organised approach to the MCS experiment that includes "background (optional), defining the aim, research questions, method of implementation, and summarised conclusion". The method of implementation is a workflow that consists of writing the code, setting the seed, setting the true parameter a priori, data generation process, and performance assessment through meta-statistics. Among the findings, it is experimentally demonstrated in the study that the GAS model with a lower unconditional shape parameter value can generate a dataset that adequately reflects the behaviour of financial time series data, relevant for volatility estimation. This dynamic framework is intended to help interested users on MCS experiments utilising the GAS model for reliable volatility calculations in finance and other areas.


  • Open Access

    Framework for Simulation Study Involving Volatility Estimation: The GARCH Approach

    by Richard T. A. Samuel, Charles Chimedza, Caston Sigauke

    This study rolls out a robust framework relevant for simulation studies through the Generalised Autoregressive Conditional Heteroscedasticity (GARCH) model using the rugarch package. The package is thoroughly investigated, and novel findings are identified for improved and effective simulations. The focus of the study is to provide necessary simulation steps for volatility estimation that involve "background (optional), defining the aim, research questions, method of implementation, and summarised conclusion". The method of implementation is a workflow that includes writing the code, setting the seed, setting the true parameters a priori, data generation process and performance assessment through meta-statistics. This novel, easy-to-understand steps are demonstrated on financial returns using illustrative Monte Carlo simulation with empirical verification. Among the findings, the study shows that regardless of the arrangement of the seed values, the efficiency and consistency of an estimator generally remain the same as the sample size increases. The study also derived a new and flexible true-parameter-recovery measure which can be used by researchers to determine the level of recovery of the true parameter by the MCS estimator. It is anticipated that the outcomes of this study will be broadly applicable in finance, with intuitive appeal in other areas, for volatility estimation.