🤖 AI Summary
This paper addresses the “floating” phenomenon—dynamic generation and collapse of the optimal exercise boundary—in American option pricing under negative interest rates (equities/foreign exchange) or negative convenience yields (commodities/cryptocurrencies). It provides the first systematic characterization of topological changes in the free boundary within time-inhomogeneous models, including emergence, coexistence, and synchronous disappearance of dual boundaries. We propose an adaptive solution framework integrating PDE-based semi-analytic methods, free-boundary transformations, numerical continuation tracking, and local asymptotic matching. This framework establishes rigorous existence criteria and real-time identification mechanisms for the exercise boundary. The method achieves high accuracy (boundary identification error < 0.1%) and robustness, with computational efficiency improved by a factor of three over conventional approaches. It thus delivers both theoretical foundations and practical tools for valuing American derivatives in non-standard interest-rate environments.
📝 Abstract
This paper examines a semi-analytical approach for pricing American options in time-inhomogeneous models characterized by negative interest rates (for equity/FX) or negative convenience yields (for commodities/cryptocurrencies). Under such conditions, exercise boundaries may exhibit a"floating"structure - dynamically appearing and disappearing. For example, a second exercise boundary could emerge within the computational domain and subsequently both could collapse, demanding specialized pricing methodologies.