Calculate your real current spending — not your estimate

Pull three months of actual bank and card data before calculating your FI number — estimates are reliably too low.

Why it works

People consistently underestimate their spending due to availability bias: salient, large purchases are easy to recall while small recurring expenses are invisible. The average person underestimates food, entertainment, and miscellaneous spending substantially. A FI number built on an underestimated baseline produces a retirement plan that runs out of money — not from investment failure but from spending reality that was never honestly captured.

How to do it

1. Download three to six months of bank and all credit card statements.
2. Categorize every transaction, including irregular or "one-time" items (these recur annually).
3. Annualize all irregular spending: car registration, insurance lump sums, travel — divide by 12.
4. Add 10-15% as an irregular-expense buffer; life consistently costs more than the line items suggest.

Evidence

Memory-based spending estimates are reliably lower than actual measured spending, consistent with availability bias and motivated underestimation. The discrepancy is documented across financial behavior research. (observational)

Studies on spending estimation error use samples that may not represent high-income earners with complex spending; the direction of the bias (underestimation) is consistent.

Common mistake

Building the FI number on "what I plan to spend in retirement" rather than current actual spending — retirement spending projections are even less reliable than current spending estimates.