How LanternArc works
Every future, pressure-tested. Every number, checked.
Most retirement tools compress your plan into one hopeful percentage and hide the assumptions behind it. LanternArc keeps the questions, and the evidence, separate and in plain sight. Here's what's under the hood.
Three questions, not one.
A single "success rate" quietly averages away the trade-offs that decide a retirement. LanternArc answers three distinct questions and never collapses them into one score:
Success Rate
Will the portfolio outlast you, across thousands of market histories?
Lifestyle Confidence
Can you keep the life you planned, without forced, lasting cuts?
Floor Safety
Do the essentials stay protected, even through a bad decade?
One headline, many lenses.
Markets don't behave like a tidy average. Your headline numbers come from Standard Projection; three more risk lenses then re-run the same plan under a specific worry — clustered downturns, stretched valuations, or the market's actual past — so you can see which assumption your result depends on.
Lenses answer different questions — they are not second opinions to shop between. When one moves your numbers, the useful response is to ask which assumption changed, not to switch until they recover.
It checks its own work.
Headline Success Rate counts the simulated futures that last. During each run, SmartPoS guides guardrail decisions, with PEMC correction near uncertain decision thresholds. A slower nested simulation is retained for internal validation on published test states; it is not a planning mode you select.
Guardrails that flex like you would.
No real retiree spends on autopilot into a downturn. LanternArc can model dynamic guardrails that ease back in lean years and spend more in strong ones, so your plan is judged the way you'd actually live it, not as a rigid withdrawal robot marching a fixed number off a cliff.
Run the whole engine from your own agent.
LanternArc exposes its full engine over MCP. Point Claude (or any MCP client) at it and ask in plain language. Your agent runs the scenarios, compares strategies, and reads the results back to you. Planning by conversation.