Longevity Standard is an independent analytical framework for lifetime income. This page explains the model behind the public scenario library and the custom work LS does for plan sponsors.
LS helps individuals, plan sponsors, and consultants look at lifetime income arrangements in terms of rules and trade-offs — how savings becomes a paycheck, what can change over time, and how different claim structures behave when conditions move against the baseline.
The core idea is simple: instead of asking "what income does my balance produce?", LS asks "what does a given level of income cost to fund, and what changes that cost?" That question is the basis for every comparison on this site. The frictionless baseline — a hypothetical arrangement with zero costs — is the reference point every real arrangement is measured against. Realized value is the share of that theoretical benefit a real product actually delivers.
LS does not provide financial, legal, or tax advice. It does not recommend products, providers, or actions. It does not forecast markets, inflation, or personal outcomes.
Every LS analysis works through four questions that apply to any lifetime income arrangement:
These four questions are a plain-language version of the four structural properties — risk sharing, adjustment mechanism, liquidity, and cost structure — that organize the LS framework. The scenario library works through them directly.
The scenario library is the public part of LS: an interactive library organized around the questions people actually ask about lifetime income. Each entry pairs a tool with a short essay — the tool lets you adjust the parameters yourself, the essay explains what you're looking at. The library is free and open.
Custom analytical work. LS also does structured analysis for plan sponsors, benefits committees, and consultants evaluating a specific plan or cohort. The most common version is a pool design analysis — a member-level look at a plan's actual age and balance distribution, including cost-of-income comparisons, outcome ranges, and an evaluation of the proposed arrangement against the four questions above. The public scenario library and the custom work use the same method; the custom work applies it to a specific plan.
The model runs on two calculation engines, both using standard actuarial methods.
The actuarial engine handles the population-level comparisons behind the public scenario library. It's built on Gompertz survival probabilities and present-value discounting, with a credibility adjustment for pool size. It compares three things directly: managing your own drawdown, a frictionless pooled arrangement, and a commercial product with a representative insurer cost added. It's deterministic, runs in under a second, and powers the interactive tools in the browser.
The pool design engine handles the member-level analysis behind custom engagements. It runs a Monte Carlo simulation across a plan's actual member cohort, using the same mortality assumptions as the actuarial engine, and produces member-level outcome ranges, comparisons of different payout rules, and an estimate of pool solvency under the plan's assumptions. This engine builds on Milevsky's 2022 open-access work on tontine design, with some adaptations for this use.
Both engines share the same mortality assumptions and produce consistent comparisons. Sources for the underlying methods are listed on the bibliography page.
LS is an educational and analytical tool. It simplifies real-world complexity in order to make mechanics legible and support specific decisions.
Results are illustrative, not guarantees. Real outcomes depend on things outside the model — policy changes, market conditions, provider decisions, and individual circumstances the model doesn't capture. The mortality assumptions are population-level and don't reflect individual health or lifestyle. Investment returns in the pool design engine are currently fixed rather than variable, which is a simplification we intend to improve. The discount rate used doesn't capture the risk of poor returns early in retirement. Insurer costs are estimated, not sourced from actual product pricing, which reflects proprietary assumptions we don't have access to. The frictionless baseline is a reference point, not something you could actually buy.
Some arrangements involve discretion or governance that can't be fully modeled. Where that's the case, LS says so explicitly rather than assuming more precision than the model has.
The LS actuarial engine and pool design engine are maintained on GitHub. Access by request.
[LS Engine] · [LS Pool Engine]
Access by request.