How Fikata works: methodology, taxes, and known limitations

Most planning tools ask you to trust a single number. Fikata tries to show you how the number was produced by providing as many details and metrics as possible for each year in each path. We also document how a simulation runs, what the tax engine models, and what it doesn’t model yet, including which way each gap pushes your results.

That last part actually matters. Every planner contains some simplifications. Knowing whether a simplification makes your plan look better or worse than reality is what lets you decide how much to trust it.

How a simulation runs

• Real market history, not an average return. Your plan is replayed against actual sequences of US stock returns and inflation from 1920 through 2024, with every investment modeled as the S&P 500, so a bad decade early in retirement hurts the way it really would.
• Inflation matched with the returns. Each path’s own inflation drives your expenses, Social Security adjustments, contribution limits and tax brackets.
• Success rate is the share of paths where the money lasted: every year’s expenses, debt payments and taxes were covered through the end of the plan. It describes the results of the histories your plan was tested on. It is not a prediction of the future.
• Cash flows are automatic. Spare cash goes down a savings ladder (employer match, high-interest debt, emergency fund, tax-advantaged accounts, then taxable investments). Shortfalls are drawn from taxable, then tax-deferred, then Roth accounts, with optional bracket caps and Roth conversions.
• Every number is visible. Open any path and any year to see what produced it, test changes with What-if, or let Optimize search for the earliest retirement or the most wealth above a success-rate floor you set.

What the tax engine models

Fikata computes both federal and state tax data every year, on every path. That covers 2026 brackets and deductions, including OBBBA’s senior deduction and SALT cap and the cap’s reversion to $10,000 in 2030; capital gains stacked correctly on top of ordinary income; the Net Investment Income Tax; Social Security taxation; FICA and self-employment tax; required minimum distributions and Roth conversions; all 50 states and DC; rental depreciation, recapture and the home-sale exclusion; and the jump in tax when a spouse dies. Future years follow current law, indexed with each path’s inflation, and figures the law holds fixed stay fixed.

The gaps most likely to matter

These are the known gaps that affect the most plans, and which way each one pushes your results. You can use these to help interpret your results and which way they might skew.

• Every investment is modeled as the S&P 500DependsYour portfolio is treated as 100% US stocks, whatever your real actually holds. Your results will show higher long-run returns but deeper drawdowns than a more balanced plan.
• ACA subsidies and IRMAA Medicare surchargesOptimisticExtra income, such as a Roth conversion, can cost marketplace subsidies before 65 and raise Medicare premiums after. Neither is modeled, so a real conversion plan should usually be less aggressive than Fikata suggests.
• Tax on dividendsOptimisticReturns include dividends, reinvested, but a taxable account pays no yearly tax on them; its gains are taxed only when sold.
• Rental loss limitsOptimisticA rental loss offsets your other income in full. The law caps this at $25,000 and phases it out between $100,000 and $150,000 of income.
• Tax creditsConservativeThe child tax credit, saver’s credit and others aren’t applied, so families and lower-income savers will see tax overstated.
• Every spare dollar is savedOptimisticSurplus cash goes down the savings ladder in full. Real households usually let some of it drift into spending.

What you can do to make your results more accurate

A few inputs make a large difference and are easy to skip:

• Cost basis of taxable accounts and real estate. Enter what you paid. Left blank, Fikata assumes you have no unrealized gain today, which understates the tax when you sell.
• Land value for rentals. Left blank, Fikata assume 20% of the purchase price, but entering the exact number will allow us to calculate depreciation more accurately.
• Depreciation already taken on a rental. If you’ve owned a rental for years, enter the depreciation you’ve already claimed. Otherwise the tax on the sale is understated.
• Self-employment. Tick the self-employed box on income you earn for yourself, so it pays self-employment tax instead of FICA.
• “Other” income. Choose whether it’s taxable, tax-free, or tax-exempt interest. Unanswered, it’s treated as taxable.
• Expenses on a rental. Attach a rental’s costs to that property; that link is what makes them deductible.

How we keep the numbers right

• Sourced figures. Federal figures come from the IRS’s annual inflation-adjustment revenue procedure; state figures from the Tax Foundation’s annual tables, checked against state revenue departments where states index their own thresholds.
• Annual refresh. Federal and state figures are updated each year when they become available.
• Tested. Over a thousand automated tests check the arithmetic of each tax and planning rule, and guard the tax tables against bad data, such as a state schedule whose rates fall as income rises. Sample households, along with scenarios built to exercise particular features (a rental sale, Roth conversions, a spouse’s death), are re-run in full on every change and compared year by year with their stored results. A fix in one place can’t quietly move results somewhere else, and a change that does move them has to be understood before it ships.
• These pages are part of the product. When a gap closes, it comes off the list. When we find a new one, it goes on.