Xipeng Huang
PhD Candidate, University of Waterloo
Room: M3 3127
A Control Approach to Dynamic Ultimate Forward Rates: Balancing Stability and Consistency
The Ultimate Forward Rate (UFR) is the pivotal regulatory anchor within the Smith-Wilson term structure model used by Solvency II for long-term valuation. Solvency II prioritizes stability through periodic and restricted updates to the UFR; however, this procedural rigidity results in a significant lag during market shocks. Conversely, endogenous UFR selection based solely on smoothness criteria (De Kort & Vellekoop, 2016) captures market signals but introduces excessive intertemporal volatility untenable for actuarial practice.
This paper proposes a dynamic control framework as a middle ground. We formulate UFR adjustment as an optimal control problem solved via Least Squares Monte Carlo (LSMC). Our framework balances smoothness with intertemporal stability, mitigating the volatility of endogenous models and the lag of regulatory rules. It proves robust across stochastic scenarios, and backtesting confirms that it provides a stable yet market-responsive anchor.