Analytics and predictive modeling
Models we build, document and stand behind, with the range of outcomes shown rather than a single number.
Most of what sponsors call judgment is a question the data can answer, if someone builds the model. We build them, document the assumptions and show the range of outcomes rather than a single number.
Employee preference measurement
MaxDiff and choice-based conjoint surveys, estimated with hierarchical Bayes methods, that show which pay, retirement, health and time-off elements employees actually trade off, what each is worth to them in dollars, and the mix that raises perceived value at the same cost. Delivered through the firm's Survey, Evaluate, Transform method.
Predictive workforce modeling
Retirement, turnover and preference shift projected as the workforce ages, so a package designed today still fits in three years. Competitive offer modeling against the employers a client actually loses people to.
Experience studies
Mortality, retirement, turnover and claims experience measured on the plan's own population, weighted against standard tables by credibility, and carried straight into the valuation assumptions.
Stochastic projections
Funded status, contributions and ESOP repurchase obligations run across thousands of economic scenarios, with fan charts, stress cases and the probability of each outcome, in place of a single deterministic path.
Benchmarking and expense analytics
Peer comparison for universities, institutions and employers that separates structural cost from one-time items and shows where a line item is high for a reason and where it is simply high.
Data quality and reconciliation
Participant data reconciled across plan, payroll and recordkeeper sources before any model runs, with a person-level gain and loss reconciliation each year so results can be explained, not only reported.
Every model is delivered with its documentation, its assumptions and a second actuary's review. Where a result rests on survey data or a simulation, we say so, and we show the uncertainty around it.