The instinct with an RRSP is to defer tax for as long as possible — leave it untouched, let it grow, and worry about withdrawals later. That instinct can be expensive. Deferral is not free: the longer you wait, the bigger the account grows, and the bigger the eventual tax bill on the way out.

You must convert your RRSP to a RRIF by the end of the year you turn 71, and, starting the year you turn 72, a rising minimum percentage must be withdrawn and taxed as income every year — whether you need the money or not. Those RRIF minimum withdrawals land on top of CPP and OAS, which by then are usually flowing too. Three income streams stacking at once can push a retiree into a higher tax bracket than they ever paid while working.

It also triggers the OAS clawback. Once individual net income passes roughly $93,500 (2025), OAS is reduced by a 15% recovery tax — so a large forced RRIF withdrawal can quietly claw back a chunk of a benefit you would otherwise keep.

The final bill comes at death. Whatever remains in the RRSP or RRIF is treated as income on the final tax return and taxed all at once — often at the highest marginal rate — unless it rolls over to a surviving spouse. A large deferred RRSP can hand a big share of the estate to the tax authorities instead of to your heirs.

The calculator does not guess. It runs a full year-by-year simulation of your retirement using real Canadian tax rules, then searches for the schedule that maximizes what is left after tax. Under the hood it:

• Simulates each year from now to life expectancy, applying federal and provincial income tax brackets for all provinces and territories, the OAS clawback, and mandatory RRIF minimums.
• Searches across many annual RRSP withdrawal amounts — and, if you let it, across CPP and OAS start ages from 60 to 70 — to see which combination performs best.
• Maximizes the net estate: the after-tax value left over at the end, after the final return's terminal tax on any remaining registered assets.
• Never proposes a plan that runs you out of money — schedules that leave you short are discarded, not recommended.
• Validates the winner against Monte Carlo market scenarios, so you can see whether a plan that looks great on average still holds up when markets misbehave.

The output is a comparison, not a prediction. It shows your current plan beside the suggested meltdown so you can see the direction and size of the difference, then decide for yourself.

Exactly as accurate as its inputs — which are a mix of things you know and things nobody knows. Your ages, current account balances, and today's tax rules are known precisely, and the arithmetic built on them — brackets, OAS clawback, RRIF minimums — is exact. Future investment returns are the opposite: they vary widely from year to year and will almost certainly not follow the smooth path any simulation assumes. Spending, tax rules, and government benefits drift over the decades too.

That is why the results are best read as a comparison between strategies, not a forecast of your actual balance decades from now. The direction of the recommendation — melting down beats deferring, and by roughly this much — is far more durable than any single dollar figure attached to it.

A plan is a snapshot, so keep it fresh: revisit your retirement plan every year or two — update balances, spending, and start-age decisions as real returns and real life diverge from the projection, and re-run the optimizer on the new numbers. The advice that matters is always the one computed from your current situation, not the one from three years ago.