How the model works,
what it assumes, and why it exists.

RetireCalc is designed to answer retirement decision questions, not just summarize a portfolio. The engine is built around tradeoffs like retirement timing, Social Security timing, Roth conversions, withdrawal order, spending flexibility, and survivor exposure.

The output is strongest when you compare scenarios. A single projection is useful, but the real value comes from asking what changes if you retire earlier, spend less, convert more, or claim Social Security later.

The core engine runs one year at a time from current age through the end of the plan. For each year it builds income, applies spending targets, calculates taxes, determines draws, applies Roth conversions if enabled, and updates account balances.

The model keeps account buckets separate. 401(k), Roth, taxable, and cash are not blended together. That matters because tax treatment, draw flexibility, and conversion sources differ by bucket.

The engine also tracks year-specific details such as bracket name, IRMAA tier, MAGI, RMDs, conversion room, and per-bucket balances so you can inspect where stress appears over time.

Federal ordinary income tax, long-term capital gains tax, state tax, and IRMAA are modeled explicitly. Social Security taxation is calculated separately rather than treated as a flat percentage of benefits.

IRMAA is especially important because retirement plans often look safe on balances but become inefficient on after-tax income. RetireCalc exposes which tier a plan lands in and how Roth conversions or RMD growth affect future surcharges.

The current app includes Colorado and a broader state tax framework. The methodology is designed to show how tax drag compounds, not just how much tax is paid in the current year.

The engine supports fixed spending as well as a smile-spending curve. That means spending can decline through middle retirement and then rise again later as healthcare costs become more important.

This matters because flat inflation-adjusted spending assumptions often overstate mid-retirement needs while understating late-life healthcare pressure.

Withdrawal order is modeled directly because sequence matters. A plan that works with one draw order can become less tax-efficient or less resilient with another.

RetireCalc lets you test taxable-first, 401(k)-first, Roth-last, or custom orders, which is one of the biggest practical gaps in simpler retirement tools.

Roth conversions are modeled as a separate decision layer after spending is funded. The engine can target bracket room or IRMAA room and show the amount converted, the tax caused by the conversion, and how that tax was funded.

This makes it possible to compare present-year tax pain against future RMD relief, future MAGI reduction, and survivor-tax protection.

Monte Carlo is there to test resilience, not to replace the deterministic plan. The deterministic path helps you understand the mechanics of your strategy. Monte Carlo helps you understand how fragile or durable that strategy is under different market paths.

That combination matters. Users need to know both what the base plan is doing and how much variance they can survive.

The advisor has access to the projection outputs and is meant to evaluate decisions with numbers, not generic finance language. It can run scenario tests, compare multiple strategies, and suggest setting changes that can be applied directly in the UI.

The right standard for recommendations is not 'interesting commentary.' The right standard is 'what should I do next, and what changed because of it?'

RetireCalc is not a guarantee engine. Future returns, inflation, health costs, tax law, and personal behavior all move in ways that models cannot fully predict.

It also is not a substitute for legal, tax, or fiduciary advice. Its job is to make the decision surface clearer, faster, and more transparent.