How Much Do You Need to Save for Retirement?

The two most quoted retirement rules — save 25 times your annual spending, and withdraw 4% per year — are useful starting points and dangerous endpoints. They were derived from US historical data, assume no taxes, ignore government benefits, and use a fixed withdrawal rate that doesn’t match how retirees actually spend. For Canadians and Americans with real-world tax situations, government benefits, and variable spending patterns, the actual number is almost always different — and often more achievable than the headlines suggest.

The 25x Rule and Why It Falls Short

The 25x rule comes from a simple inversion of the 4% withdrawal rate: if you can safely withdraw 4% of your portfolio each year, you need 25 times your annual spending to be financially independent.

The logic is clean but the application is flawed for most people:

It ignores taxes entirely. If you have $1.5M in an RRSP or Traditional IRA, you don’t have $1.5M to spend — you have $1.5M minus whatever you owe in income tax on every dollar you withdraw. Depending on your province or state, that could be 25–45% of each dollar. Calculating your retirement number based on gross withdrawals from registered accounts systematically understates what you need.

It ignores government benefits. CPP and OAS in Canada or Social Security in the US meaningfully reduce how much you need to pull from your portfolio. A Canadian retiree receiving $1,500/month in combined CPP and OAS ($18,000/year) needs $18,000 less in annual portfolio withdrawals. At a 4% withdrawal rate, that’s the equivalent of $450,000 in savings you don’t actually need.

It uses a fixed spending assumption. Research on actual retiree spending shows a “smile” pattern: spending is higher in early active retirement (travel, activities), decreases in the mid-70s as activity slows, then can spike late in life for healthcare. A fixed dollar amount per year is a poor fit for this profile.

The 4% Rule: Where It Came From

The 4% rule originates from William Bengen’s 1994 research, later reinforced by the 1998 “Trinity Study” (Cooley, Hubbard, and Walz), which analyzed historical US stock and bond returns from 1926–1995 and found that a 60/40 portfolio could sustain a 4% withdrawal rate for 30 years with high probability.

Limitations worth knowing:

• 30-year horizon only. Retire at 55 and you may need a 40–45 year portfolio. The 4% rule’s success rate drops materially over longer horizons.
• US historical returns. US equities have outperformed global averages over this period. International portfolios may have lower expected returns.
• Sequence of returns risk. A bear market in the first 5 years of retirement can permanently impair a portfolio on the 4% withdrawal schedule, even if long-run returns recover.
• Inflation sensitivity. The study assumes CPI-linked spending increases. Healthcare inflation often exceeds CPI, meaning real spending can grow faster than assumed.

A 3.5% withdrawal rate is often cited as more conservative for early retirees or those with 40+ year horizons. Some researchers suggest 3.3% for maximum safety across international portfolios.

The Income Replacement Approach

A more intuitive framework: target replacing 70–80% of your pre-retirement income. The logic is that retirement eliminates several major expenses:

• Mortgage payments (typically paid off by retirement)
• Saving for retirement (obviously)
• Work-related costs (commuting, professional clothing, work lunches)
• Payroll taxes in Canada and US

These adjustments can reduce required income by 20–30% relative to working years. A household spending $90,000/year while working may live comfortably on $65,000–$70,000 in retirement.

But this framework also needs adjustment for your specific situation. If you plan extensive travel in early retirement, the number may be higher. If you have paid-off property and low fixed costs, it may be lower.

Why Location Changes Everything

The “retirement number” varies enormously by province and state because taxes interact differently with portfolio income:

• An Ontario retiree drawing $70,000/year from an RRIF faces approximately $14,000–$16,000 in combined federal/provincial tax.
• The same withdrawal in Alberta faces roughly $11,000–$13,000 — a $2,000–$3,000/year difference that compounds over decades.
• In Texas or Florida (no state income tax), a US retiree withdrawing $70,000 from a Traditional IRA owes only federal tax — roughly $8,500–$10,000. In California, add another $5,000–$7,000.

Location also affects CPP/OAS and Social Security amounts indirectly (QPP vs CPP differs), healthcare costs in the US, and housing costs. A retirement target calculated without province/state specifics can be off by 15–25%.

A Realistic Example

Sarah earns $80,000/year in Ontario and has no defined benefit pension. She wants to retire at 65.

Simple 25x calculation: $80,000 × 0.75 = $60,000 target spending. $60,000 × 25 = $1,500,000.

More precise calculation:

• CPP at 65: $1,100/month ($13,200/year)
• OAS at 65: $750/month ($9,000/year)
• Portfolio withdrawal needed: $60,000 - $22,200 = $37,800/year
• Pre-tax withdrawal required (accounting for Ontario tax on RRIF income): approximately $44,000–$46,000/year
• At a 4% withdrawal rate: $44,000 / 0.04 = $1,100,000

The actual portfolio needed is about $1.1M — $400,000 less than the naive 25x calculation because government benefits cover more than $22,000 of the spending need. Sarah needs significantly less than the headlines suggest.

The exact number changes if she delays CPP and OAS to 70 (lower portfolio withdrawals needed after 70, but higher gap spending from 65 to 70), has a spouse, or holds different account types with different tax treatment.

The Gap Between 25x and Reality

Most Canadians retire on less than 25 times their spending — and many do so comfortably. Government benefits, paid-off housing, and lower expenses in retirement are real factors that the 25x rule ignores. The Americans who rely on Social Security have a similar dynamic: the benefit offsets a meaningful portion of required portfolio withdrawals.

The more useful exercise is to build a year-by-year income projection that includes your specific benefits, your account types, and your tax situation — and to run it under multiple scenarios (early retirement, bear market in year one, long lifespan) to understand where the real risk lies.

How Cinderfi Helps

Cinderfi builds a year-by-year retirement projection that accounts for CPP/OAS or Social Security, RRIF minimums, provincial and state taxes, and the tax treatment of each account type. Enter your savings, income, and province or state — the planner shows your projected annual shortfall or surplus, the portfolio balance over time, and the probability of success under Monte Carlo simulation. The retirement number Cinderfi calculates is specific to your tax situation, not a rule of thumb.

Model this in your own plan — try Cinderfi free.