Optimism, Restraint, and the Cost of Compounding

Two sentences, one track record

"I'll definitely be earning more by then" and "I can definitely stop whenever I want" are two different sentences with the same actuarial track record: worse than the person saying them expects. This chapter maps the three patterns that specifically distort borrowing decisions — the wiring that makes debt feel smaller, safer, and more temporary than it is.

Optimism bias: other people's spreadsheets

People systematically underrate the odds that they specifically will hit a job loss, a medical bill, a divorce, a rate shock. Bad outcomes are real, acknowledged, statistically appreciated — as things that happen to other people's spreadsheets. The consequence for borrowing is precise: debt gets sized against best-case future income rather than a realistic range of it. The mortgage that works if nothing goes wrong; the car payment that fits if the overtime continues; the "I'll pay it off when the bonus lands." Volume 1's When Preparation Isn't Enough documents just how common the "other people" events actually are — job loss, medical crisis, and divorce are base-rate ordinary, not lightning strikes — which makes optimism bias, in borrowing, a systematic mispricing of one's own future. The defense isn't pessimism; it's sizing debt against the range — including the version of next year where something ordinary goes wrong (Emergency Buffer Sizing's room for error, applied at origination).

Restraint bias: the loaded option

People overestimate their own future self-control — reliably, and by a lot. Financially, restraint bias is why keeping a high credit limit "just in case — I won't actually use it" feels safe and behaves like a loaded option: the limit doesn't know you intended restraint. It's why the "just to look" trip to the store or the marketplace app so often ends in acquisition — the future self who was supposed to say no turns out to be the same self who wanted to go looking. The research pattern is humbling: confidence in one's restraint increases exposure to temptation (the confident keep the cards in the wallet, walk the aisles, keep the apps installed) — so the bias doesn't just fail to protect; it escorts you to the exact places it fails. The defense, once again, is structural rather than heroic: don't negotiate with future temptation — remove it from the route. Self-set guardrails, friction added on purpose, limits right-sized to actual use (the same systems-beat-willpower conclusion as Loss Aversion, Present Bias, Mental Accounting, Anchoring, now applied to credit exposure).

Exponential growth bias: the misunderstanding that compounds

The most expensive of the three, because it's not a one-time misjudgment — it compounds for as long as the misunderstanding lasts. Linear intuition says small payments make small progress: pay something every month, and the balance must be shrinking. But compound interest doesn't grow linearly — it grows on itself — and a minimum payment can lose to its own interest charge for years before the balance visibly moves.

The canonical arithmetic: a $5,000 balance at 24% APR, paid at a 2% minimum, can take over a decade to clear and cost more in interest than the original amount borrowed. The federally mandated payoff-disclosure box on every card statement (Volume 1, Statements Decoded) exists specifically because this number surprises almost everyone who encounters it — Congress concluded, correctly, that nobody's intuition computes it. That's exponential growth bias, named by statute: a disclosure law whose entire purpose is compensating for a known defect in human arithmetic.

Where Plenee fits

A planned Total Cost of Ownership / true-payoff feature would take that disclosure-box math — currently a one-time number printed monthly — and make it live and personalized: updating as your balance and payments change, not printed once and gone quiet (see pending_features.md; this is the feature aimed squarely at exponential growth bias). For optimism bias, Plenee's job is showing the realistic range, not the best case: a projected cash position that includes what a bad month actually looks like, not just an average one. And for restraint bias — the user's own guardrails, reflected back without judgment: Plenee never sets your limit; it makes the one you set visible.