Description: This session from the 1995 SOA Boston Meeting covers a general overview of basic interest rate models, the meaning of an arbitrage free model, the considerations in choosing an appropriate interest rate scenario generator, the uses, alternative approaches, the double mean reverting process model - DMRP, and how good the results are in terms of historical rates. From the Record of the Society of Actuaries, Vol. 21, No. 4b.

Description: This presentation considers the pricing of European derivatives in a Black-Scholes model with stochastic volatility. The presentation shows how Parseval’s theorem may be used to express those prices as Fourier integrals. This is a significant improvement over Monte Carlo simulation in many cases. The main ingredient in this presentation's method is the Laplace transform of the ordinary [constant volatility] price of a put or call in the Black-Scholes model, where the transform is taken with respect to maturity [T]. The presentation derives these formulas and then applies them to the case where volatility is modelled as a Markov chain in continuous time, the so-called 'Markov regime switching model'. This model has been used previously in stochastic volatility modelling, but mostly with only N=2 states. The presentation shows how to use N=3 states without difficulty, and how larger number of states can be handled. Numerical illustrations are given, including the volatility smile in a three-state model.

Description: The article warns against treating the results of a stochastic model with more deference than other actuarial analysis, and describes some hidden risks involved with choice of model and calibration.

Description: The research investigates the challenges associated with determining a “fair value” assessment for embedded options in two product types and incorporates the process into product pricing and liability valuation

Description: This paper addresses the evaluation of long-term returns when the short interest rate is modeled by a general diffusion process. By deriving the long-term return dynamics and invoking the Feyman-Kac formula, the long-term return is represented as the solution of a partial differential equation. A finite difference method is derived for the valuation of the long-term return. The method could be used for hedging interest rate risk, pricing and managing interest rate derivatives and interest rate sensitive insurance products. Numerical examples are included.

Description: Prior work indicates that a regime-switching stochastic model with randomized regime parameters creates a more plausible set of extreme paths than do the usual stochastic interest rate models. Generalizing the Black-Karasinski model by randomizing the mean reversion target provides an example. Better to understand such models, as well as to calibrate their parameters without trial and error runs of a stochastic model, this paper uses an asymptotic expansion to estimate the moments of the integrated stochastic process over time. Mention is made of a family of related asymptotic expansions that could be of wider interest.

Description: This article is based on an excerpt from a Milliman Research Report on universal life insurance with secondary guarantees, ULSG. It discusses stochastic profit analysis and the impact of investment volatility of two common variants of ULSGs - level specified premium and single-fund shadow account designs.

Description: 1998 Valuation Actuary Symposium. The panelists discussed how asset liability management is synonymous with a dynamic financial analysis and quantitative risk management. Topics covered were systems implementations including asset risk exposure optimization given a company-specific set of liabilities and total risk exposure optimization, which includes stochastic analysis of alternative asset and liability strategies.

Description: Feature article discussing numerical short cuts, rules of thumb, estimators, modeling tips, etc. The featured problem is fitting the intermediate points for a real-world stochastic model of interest rates.

Description: Producing a crystal-clear balance roll-forward on time, obtaining a fresh model output the first time or providing model results for your business partners with a tight turnaround all depend on one thing: how fast the model engine runs. Authors explored various reasons behind significant computational power consumption, and discussed different ways to speed the computational process such as in-force population compression, model use case consolidation, code simplification and streamlining, and infrastructure automation, etc.