The Insurance and Pension Solutions Group (IPS) stable value fund wrap model is presented. The modeling approach is to be used for several two basic contract types involving stable value protection on Bank Owned Life Insurance funds (BOLI), and Company Owned Life Insurance funds (COLI). These products are fundamentally the same but differ with respect to tax treatment and/or the specific details regarding the redemption of the funds or portions thereof.
The model as tested, does not perform pricing, but rather makes an estimate of losses. The other side of the pricing equation, namely the estimation of received fees, is not treated in the model, although the loss estimation appears to be, for the most part, consistent with the detailed structure of these contracts. Certain simplifications have, however, been made to facilitate the modeling process. These simplifications do not appear to lead to major errors in loss estimation.
The stable value model is aimed at estimating the value of providing protection on a portfolio consisting of fixed income and equity instruments. The fund upon which stable value protection is provided is assumed to be under active management and thus to have a constant duration for the purposes of modeling. The protection provided by the stable value contract is written on any shortfall between the market value of the fund and a defined “book value” which exists when redemptions of the fund are made.
The book value of the fund is computed as a function of the “crediting rate”, which itself is a function of previous market and book values as well as the equity and/or fixed income indices that indicate the market value of the fund.
In reality, a two-tiered strategy is used to accelerate the convergence of the market value toward book value in the event that certain exposure limits are breached. When the book value exceeds market value by 10%, the portfolio must be reallocated to reduce the duration. If the book and market values continue to diverge, when the book value exceeds market value by 20%, the portfolio is reallocated to strictly money-market instruments, and therefore has a duration of 0.25.
Payouts under the contract occur only if redemptions of the protected fund are made by the policy holders. Thus it is necessary to model the expected redemptions of the fund. The model currently uses a probabilistic redemption model. The redemption probability is given by specifying an annual redemption rate. This annual redemption rate is then converted into a redemption probability for a time step
The model uses the redemption probability for a particular time step to make a random decision whether or not redemption occurs in that time step. If the decision to redeem is made, then the model values the payout which must be made if the entire protected fund is liquidated. Under this scheme, partial surrenders of the fund never occur.