Scholarlink

<p>Artificial Intelligence (AI) strengthens the reliability and speed of complex derivative valuation, aiding informed financial decisions under varying market conditions. This study explores the application of AI in valuing American-style put options through metaheuristic optimization and sensitivity analysis. Five distinct boundary methods are evaluated for a single option pricing problem using Particle Swarm Optimization (PSO) and Ant Colony Optimization (ACO), with performance benchmarked against existing published results regarding accuracy and computational efficiency. The exponential boundary under ACO is identified as the most effective. This optimal boundary is then used to analyze key option Greeks (Delta, Vega, Rho, Theta). Finally, Response Surface Methodology (RSM) is applied to examine the sensitivity of these Greeks for variations in input parameters, strike price (K), interest rate (r), volatility (σ), and time to maturity (T). Both numerical and statistical sensitivity analysis indicate that r, σ, and K are the most critical factors in American put option pricing. The findings demonstrate the robustness of AI-driven techniques in complex option pricing and offer practical insights for financial modeling and risk assessment.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

American put option valuation using ACO and PSO: performance comparison of boundary functions and sensitivity analysis via numerical ACO and statistical RSM methods

  • 期刊论文

  • 发表日期：

    2025年11月12日

查看全文
  • Afroza Akter,
  • A. B. M. Shahadat Hossain,
  • Salma Parvin

摘要

Artificial Intelligence (AI) strengthens the reliability and speed of complex derivative valuation, aiding informed financial decisions under varying market conditions. This study explores the application of AI in valuing American-style put options through metaheuristic optimization and sensitivity analysis. Five distinct boundary methods are evaluated for a single option pricing problem using Particle Swarm Optimization (PSO) and Ant Colony Optimization (ACO), with performance benchmarked against existing published results regarding accuracy and computational efficiency. The exponential boundary under ACO is identified as the most effective. This optimal boundary is then used to analyze key option Greeks (Delta, Vega, Rho, Theta). Finally, Response Surface Methodology (RSM) is applied to examine the sensitivity of these Greeks for variations in input parameters, strike price (K), interest rate (r), volatility (σ), and time to maturity (T). Both numerical and statistical sensitivity analysis indicate that r, σ, and K are the most critical factors in American put option pricing. The findings demonstrate the robustness of AI-driven techniques in complex option pricing and offer practical insights for financial modeling and risk assessment.