Good to know

A one-time license with no subscription, including 12 months of updates, plus a 30-day free trial that needs no card and no account. The license itself never expires and nothing ever locks: when the update year is up, every version you already have keeps working exactly as it did, and an update pass is optional and never automatic. macOS, Windows and Linux are all available now; the download section labels the current direct release separately from the Microsoft Store shelf version. Current launch and regular figures live on the pricing section. The retirement planning assumes the US tax system, so licenses are sold in the United States only. Anyone anywhere can still download and run the full trial. If it is not right for you, reply within 30 days for a refund (terms).

Nothing changes in the app you're running. Every version released inside your 12 months is yours permanently. It keeps opening, importing, projecting and exporting, with no feature switched off and no nag. What lapses is only your claim on newer builds, and it is checked when you install one, never while you work. An optional update pass extends coverage by another 12 months from the current coverage end: renewing early loses no days, and waiting does not create free coverage for the gap. It never renews automatically. If you never buy one, every covered version remains yours.

A rangefinder measures the distance to a target before you act. That's the job here: read where you stand, measure the gap to the mix you chose, turn that gap into exact Add / Reduce moves, and test the distance to and through retirement. RangefinderInvest never connects to a brokerage or places a trade; you place every order yourself.

No, and that's by design. You import a CSV transaction export from Fidelity or Schwab, or type in your monthly statement. RangefinderInvest never has your login, never moves money, and never places a trade. It shows illustrative Add / Reduce moves toward the target you set; you place every order yourself.

In a single SQLite file on your own machine. There's no Kalman-hosted portfolio cloud, product account, usage analytics or crash telemetry. You can see exactly where the file lives, move it to a synced folder or external drive, and keep automatic timestamped backups.

RangefinderInvest sends no accounts, holdings or plan anywhere on its own. Configured market-data lookups carry ticker symbols and date ranges. The about-daily update check is never sent sooner than 20 hours after the last successful automatic check; it carries the app version, OS, a first-run flag and a temporary random retry receipt. The server retains only the receipt's SHA-256 digest for up to 90 days so a retry is counted once. Neither request carries portfolio records, and both can be turned off.

Optional Assistant and MCP access is also off until enabled. A hosted provider or connected AI client receives the portfolio context and tool results you authorize; a local model can keep the flow on your machine. Ten app-defined tools can read and compute, and the only write is a proposed target model you must approve. Optional HTTP and local-network doors have separate off-by-default switches. Full detail is in the privacy policy.

Yes. One click downloads official daily share prices for all 16 TSP funds, including G, F, C, S, I and every Lifecycle (L) fund, from tsp.gov. The five core funds reach back to 2003; each Lifecycle fund begins at its launch. You enter your own balances because RangefinderInvest does not connect to a TSP account. Those holdings can then be tracked, compared, backtested and included in the retirement projection. There is no guided stay-versus-roll decision engine.

No. RangefinderInvest is a calculator for a plan you design. It shows illustrative Add / Reduce moves toward the target allocation you set and hypothetical projections from assumptions you enter. It is not financial, investment, or tax advice, and it doesn't recommend securities. Confirm prices, tax lots, wash-sale rules and suitability with your broker or advisor before acting; the app carries the same note at the point of each result.

The CSV importer auto-detects Fidelity and Schwab formats, and brokerage-aware ordering plus NTF (no-transaction-fee) filtering cover both. The model layer itself is brokerage-agnostic. Implement the same target with whatever tickers you hold, wherever you hold them.

Yes. It's the whole second half of the app. The Projection runs a seeded Monte-Carlo simulation of your household in today's dollars, starting from your real balances by tax class. It models working-years contributions, spending phases, Social Security, one-time flows, taxes (federal brackets, a flat state rate, the Social-Security provisional worksheet, ACA/IRMAA ceilings) and per-owner RMDs. On top sit the optimizers: fill-to-bracket Roth conversions, a "when can I retire?" age sweep, a home-affordability frontier, a spending guardrail that flexes only the spending you've marked discretionary, and a survivorship model that shows the widow's penalty honestly.

The same plan is scored under a forward Normal model, a fat-tailed Student-t model, and a historical replay of 1928–2025. The historical replay itself offers four sampling methods because how you reuse history changes the answer. Showing the spread makes visible how much the plan depends on one reading of market history.

Deliberately so. It runs in today's (real) dollars, so Social Security and tax brackets are treated as inflation-indexed rather than double-counted. It's seeded and deterministic, so optimizer comparisons are never sampling noise. The tax model is ordinary-income only, with no LTCG/NIIT stack. That simplification is documented right on the reference panel. Where a number can't be precise, the app says so instead of implying false confidence.

The Screener shows two distinct rankings. The Classic score blends risk-adjusted return with confidence, fee, drawdown, consistency and concentration terms. The default Peer score compares downside-risk-adjusted return within each equity category and shrinks thin histories toward the category average; it deliberately does not subtract expenses again because adjusted prices are already net of fees. The separate Optimizer ranks substitution candidates on the train/eval metric blend you configure.

They apply the model-weight version actually in force on each rebalance date, hold cash slices flat instead of redistributing them, and state the true common data window. Presets the data can't cover are disabled, not silently truncated. Extending a backtest past a model's first version date discloses the extrapolation on the banner, and every statistic follows the window the chart shows. You can also overlay any ticker (a target-date fund, for example) or another model as a benchmark, with excess CAGR, beta, tracking error, information ratio and up/down capture.

Held-to-maturity Treasuries, brokered CDs and bonds live in a Fixed Income view priced by accrual toward par, with no fake $0 prices or false dips on your value chart. You get the ladder by maturity date, FDIC issuer exposure, a forward income calendar, and a Builder that models a target ladder off the live Treasury par-yield curve. A model's ladder slice counts every held rung in rebalancing automatically.

The shipping app uses Tiingo, a properly licensed market-data provider. You connect your own Tiingo API key at setup, and Tiingo's free tier is enough for a personal portfolio. The app also uses official tsp.gov data for TSP funds. "Update prices" grabs only the missing recent days; "Rebuild" re-downloads from scratch. An optional Yahoo/yfinance source exists as an off-by-default, personal-use opt-in, and you can pin a source per ticker.