Robert Ștefan Constantin

Mr. Robert Ștefan Constantin

orcid

Research Focus & Metrics

Academic Career Length

2Years (2025-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
17 September 2025 in Sustainability
Verified
Journal Article
Open access
PDF file
ADAdriana AnaMaria Davidescu
Adriana AnaMaria Davidescu
EMEduard Mihai Manta
Eduard Mihai Manta
MFMargareta-Stela Florescu
Margareta-Stela Florescu
RCRobert Ștefan Constantin
Robert Ștefan Constantin
CMCristina Manole
Cristina Manole
The escalating frequency and severity of natural disasters present significant challenges to the stability and sustainability of global financial systems, with the US stock market being especially vulnerable. This study examines sector-level exposure and contagion dynamics during climate-related disaster events, providing insights essential for sustainable investing and resilient financial planning. Using an advanced econometric framework—dynamic conditional correlation GARCH (DCC-GARCH) augmented with Generalized Hyperbolic Processes (GHPs) and an asymmetric specification (ADCC-GARCH)—we model daily stock returns for 20 publicly traded US companies across five sectors (insurance, energy, automotive, retail, and industrial) between 2017 and 2022. The results reveal considerable sectoral heterogeneity: insurance and energy sectors exhibit the highest vulnerability, with heavy-tailed return distributions and persistent volatility, whereas retail and selected industrial firms demonstrate resilience, including counter-cyclical behavior during crises. GHP-based models improve tail risk estimation by capturing return asymmetries, skewness, and leptokurtosis beyond Gaussian specifications. Moreover, the ADCC-GHP-GARCH framework shows that negative shocks induce more persistent correlation shifts than positive ones, highlighting asymmetric contagion effects during stress periods. The results present the insurance and energy sectors as the most exposed to extreme events, backed by the heavy-tailed return distributions and persistent volatility. In contrast, the retail and select industrial firms exhibit resilience and show stable, and in some cases, counter-cyclical, behavior in crises. The results from using a GHP indicate a slight improvement in model specification fit, capturing return asymmetries, skewness, and leptokurtosis indications, in comparison to standard Gaussian models. It was also shown with an ADCC-GHP-GARCH model that negative shocks result in a greater and more durable change in correlations than positive shocks, reinforcing the consideration of asymmetry contagion in times of stress. By integrating sector-specific financial responses into a climate-disaster framework, this research supports the design of targeted climate risk mitigation strategies, sustainable investment portfolios, and regulatory stress-testing approaches that account for volatility clustering and tail dependencies. The findings contribute to the literature on financial resilience by providing a robust statistical basis for assessing how extreme climate events impact asset values, thereby informing both policy and practice in advancing sustainable economic development.

1 July 2025 in Proceedings of the International Conference on Business Excellence
Verified
Journal Article
Open access
PDF file
RCRobert Ștefan Constantin
Robert Ștefan Constantin
ADAdriana AnaMaria Davidescu
Adriana AnaMaria Davidescu
EMEduard Mihai Manta
Eduard Mihai Manta
Developing forecasting models capable of learning from small datasets is increasingly valuable for scenarios with limited computational resources and tight time constraints. In this case study, we processed monthly data on Romania’s inland gas consumption—a primary benchmark reflecting the country’s industrial growth, level of technological advancement, and reliance on non-renewable energy sources. This research tests the extent to which LightGBM, a gradient boosting framework, can predict seasonal patterns in monthly gas consumption. To aid the machine learning framework in better understanding the series pattern, we applied a first-order differencing to the data. By combining hyperparameter tuning, cross-validation, and tailored feature engineering (including lagged variables and rolling-window statistics), the analysis thoroughly evaluates LightGBM’s performance under data-scarce conditions. Model accuracy was assessed using Mean Absolute Error (MAE), Mean Absolute Percentage Error (MAPE) and Root Mean Squared Error (RMSE), demonstrating the extent of LightGBM’s predictive capacity across multiple horizons despite constrained data settings. These findings offer insights into the feasibility of employing fast-adapting, lightweight machine learning techniques for reduced time series datasets, while minimizing both computational effort and processing time.