How the Appleseed Engine Actually Works
A fiduciary-grade simulation engine built for the real complexity of tech wealth - where flat-math calculators quietly fall apart.
Most retirement calculators run on flat math and a single average return - fine for a rough sketch, dangerous for a real decision. Appleseed wraps a stochastic Monte Carlo model around a year-by-year federal and state tax calculator, with native support for tech equity and pre-IPO lockups.
Here is exactly how the engine works under the hood - the six systems that make its numbers trustworthy.
Five hundred futures, not one average.
Every plan is evaluated across 500 independent Monte Carlo paths. Within each path, the engine runs a complete annual close: income, vesting, spending, full tax computations, and a tax-aware withdrawal waterfall.
Because returns are drawn year by year, the engine natively captures sequence risk. Two paths with identical average returns can end in wildly different places - it depends on when the drawdown lands.
Real households do not spend blindly into a crisis. With adaptive spending on, the engine trims lifestyle or downsizes in a bad decade, showing the mathematical value of flexibility.
Most tools model a paycheck. We model the whole grant.
Appleseed follows the full life of a stock grant - from unvested RSUs, through pre-IPO lockups, to liquidity - instead of flattening it into a salary line.
Unvested value releases on its actual schedule, with annual refresh grants layered on top - the real overlapping shape of tech compensation.
Vested private shares grow with company valuation but sit in an illiquid bucket, unable to fund life until a modeled IPO or tender releases them.
At each vest, shares are sold to cover income-tax withholding at your rate, then statutory payroll tax stacks on top.
Taxes computed inside every year of every path.
Not sketched as one blended effective rate. Each layer is calculated where it actually lands - bracket by bracket, surcharge by surcharge.
All 50 states, with California's full marginal structure modeled bracket by bracket, plus long-term capital-gains brackets and the 3.8% NIIT above statutory thresholds.
HSAs can be preserved for tax-free compounding instead of current medical costs. 529 assets are automatically prioritized against modeled college costs.
Computed off modified AGI with the correct two-year lookback. A large equity vest at 63 triggers the exact Medicare premium surcharge at 65.
A home is the biggest lever - so it is a system, not a line item.
For many households a primary or second home is the single largest variable in the plan. Appleseed models its full carrying economics and every transition.
Schematic median liquid assets. Buying a more expensive home dips liquidity; selling and renting later frees the most capital, downsizing frees some, and keeping the home leaves the most wealth locked in property.
Full amortization schedules with interest and principal splits, appreciation, maintenance, and property tax.
Downsize, move, or rent at a specific age or retirement date. Selling costs are modeled and after-tax proceeds flow into the broader portfolio.
In failing paths, the engine can model a forced home sale and price a replacement rental scaled to the local market value of the home you exited.
We replace the generic market crash with your company's stock.
When you run the Dot-Com stress test, a dedicated model separates your employer equity from the broad market, answering the question a concentrated employee actually cares about.
Schematic, not sourced return data. This wider, fatter-tailed distribution is applied only when you run the Dot-Com stress test - it is not how concentrated equity is treated in a normal plan.
Run the Dot-Com stress test and your employer-stock position is simulated using the historical returns and volatility of an equal-weight basket of the 10 largest NASDAQ Composite companies on March 10, 2000, measured through May 2026.
If my company's stock does what stocks like it have historically done in a sector crash, does my plan survive?
It reads the model and explains what is driving the result.
Built into every plan, the Assistant translates success probability, percentile bands, and withdrawal sequencing into plain English - then helps compare what-if scenarios side by side.
Your plan looks strong in this run. The main reason is that projected income covers spending for a long stretch, giving the portfolio more time to grow before withdrawals begin.
Cash flow. Income is projected to run ahead of living expenses during the working years, creating room to keep saving.
Time in the market. A long runway before withdrawals gives invested assets more time to compound.
Retirement timing. Moving the retirement date changes both the savings runway and the number of years the plan needs to fund.
The biggest swing factors are spending, savings rate, retirement timing, and market assumptions. Running side-by-side scenarios makes those tradeoffs easier to see.
The Assistant is designed to explain the model, not make decisions for you. It shows what is helping, what is sensitive, and which what-ifs may be useful to compare.
See what a real engine says about your plan.
Build a fiduciary-grade plan in about five minutes. No spreadsheet required.