For pension plans and annuity providers, capital market-based solutions represent one potential approach to managing excessive mortality and longevity risk exposures. The valuation of longevity-linked securities necessitates the application of stochastic methods to project future mortality developments, alongside an incomplete market premium principle that accounts for the market price of longevity risk. This paper examines the influence of the mortality model, process and parameter uncertainty, and the choice of premium principle on the market pricing of e-forward (life expectancy) contracts. Two model combination approaches for mortality forecasting are evaluated—Bayesian Model Ensemble and Stacking Regression with the elastic net as the meta-learner—alongside two widely used pricing principles: the Wang transform and the Proportional Hazard transform. Mortality and life annuity data for the total Portuguese population are employed to calibrate the models and generate illustrative empirical results.

--