Explore variable returns during FIRE withdrawals

1. Declare your assumptions

Enter initial capital, first-year annual spending, currency, fixed annual inflation and withdrawal horizon. Assumptions are prefilled and editable, rather than recommendations. Start this more expensive simulation explicitly after reviewing them. Choose withdrawals at the start or end of each year. This is your model without imported market prices, historical calibration, taxes or fees. Changing currency does not perform an FX conversion.

2. Define the return distribution

Supply three personal annual returns and positive integer weights. A weight represents relative sampling frequency, not a historically measured frequency. Years are drawn independently from the same distribution; the model contains no trends, dependence between adjacent years or inflation-return relationship. Your declared assumptions therefore determine the outcome.

3. Run a reproducible simulation

Choose a path count and seed, then run explicitly. The same method, all the same inputs and the same seed reproduce the paths. More paths do not validate the assumptions. Input changes require another run; an older response must not replace a newer draft.

4. Read both the range and a specific path

The annual range summarizes every modeled path, including paths with a previous funding shortfall. Quantiles across different years need not form one actual path. Inspect a selected path to check sampled return, available capital, planned and funded spending and the first unfunded amount.

This simulation permits a partial withdrawal of available money. After the first shortfall, capital remains zero; debt and later income are not modeled. A fully funded final withdrawal with an exactly zero balance is not a shortfall. The separate original withdrawal model uses a different rule: an unfunded withdrawal is not executed.

5. Compare spending sensitivity

Your spending-deviation percentage produces lower, base and higher spending variants. They share the same sampled returns, so differences come only from declared spending. Read each shortfall count with its full path denominator. The fraction applies only within this model; it is neither a validated retirement-success probability nor a recommended safe withdrawal rate.