Article Dans Une Revue Asian Journal of Probability and Statistics Année : 2025

Exchange Rate Risk Measurement of Kenyan Commercial Banks: By Integrating Value-at-Risk and Extreme Value Theory

Résumé

Exchange rate risk is a critical concern for financial institutions particularly in emerging markets like Kenya where currency volatility poses significant challenges to economic stability. Kenyan commercial banks operate in an environment heavily influenced by fluctuations in the Kenyan Shilling (KSH) against the US Dollar (USD) facing considerable risks that disrupt financial planning, profitability and overall market stability. The increasing volatility in exchange rates driven by global economic uncertainties and domestic macroeconomic pressures has amplified the need for robust and reliable frameworks to assess and manage this risk. Traditional Value-at-Risk (VaR) methods such as Historical Simulation and Monte Carlo Simulation are widely employed to measure potential losses from adverse exchange rate movements yet these models often fail to adequately capture extreme market conditions leaving institutions exposed to rare but severe events. This gap underscores the importance of integrating advanced statistical techniques to improve the precision and reliability of risk assessment frameworks. This study presents a complete framework for measuring and managing exchange rate risk in Kenyan commercial banks through the innovative integration of Value-at-Risk (VaR) methodologies and Extreme Value Theory (EVT). The primary objective was to measure exchange rate risk using VaR methods namely Monte Carlo Simulation and Historical Simulation and integrate EVT particularly the Generalized Pareto Distribution (GPD) into the VaR framework to capture the likelihood and magnitude of extreme currency fluctuations. The final objective was to verify the validity of the integrated VaR-EVT model through rigorous backtesting procedures ensuring robust risk estimates. To achieve these objectives, the study employed a quantitative research methodology focusing on comprehensive daily KSH/USD exchange rate data from January 2019 to December 2023 comprising 1,262 observations for exhaustive analysis of risk measurement approaches under various market conditions. The research methodology combined secondary data from Kenyan commercial banks and financial reports with advanced statistical modeling techniques. VaR methods were used to quantify potential losses under normal market conditions while EVT was incorporated to model extreme events that fell outside traditional VaR assumptions. The reliability and accuracy of the combined VaR EVT framework were assessed using multiple robustness checks and backtesting procedures. Our results demonstrate that conventional VaR methods underestimate tail risk by 23-42% during extreme market events, while our integrated VaR-EVT framework provides superior risk estimates across all confidence levels. The paper includes detailed methodology, extensive empirical results with multiple robustness checks, practical implementation guidelines and policy recommendations. This research offers a comprehensive and structured approach to assessing exchange rate risk addressing critical limitations in existing methodologies. By integrating EVT into the VaR framework, the study enhances the ability of financial institutions to anticipate and manage the impact of extreme exchange rate events. The findings provide valuable insights for risk managers, financial analysts and policymakers in Kenya equipping them with advanced tools to mitigate exchange rate risk and strengthen the financial stability of the commercial banking sector. This research contributes both methodological advancements in financial risk management and practical insights for banking operations in emerging markets where extreme currency fluctuations are increasingly prevalent, thereby advancing academic literature in statistical applications to financial risk assessment.

Fichier non déposé

Dates et versions

hal-05193838 , version 1 (31-07-2025)

Identifiants

  • HAL Id : hal-05193838 , version 1

Citer

Barasa Elika Joan, Moses Kololi, John Sirengo. Exchange Rate Risk Measurement of Kenyan Commercial Banks: By Integrating Value-at-Risk and Extreme Value Theory. Asian Journal of Probability and Statistics, 2025, 27 (8), pp.17-27. ⟨hal-05193838⟩
50 Consultations
0 Téléchargements

Partager

  • More