How Much You Need for Retirement

Step 1: Estimate retirement spending

The most common planning error is using "replace 80% of your current income" or some similar income-based rule. Spending, not income, drives retirement needs.

Start from your current spending and adjust for retirement changes:

Likely to decrease:

• Payroll taxes (FICA gone; only ordinary income tax on withdrawals)
• Retirement contributions (no longer saving)
• Commuting and work-related expenses (clothes, lunches, transportation)
• Housing in some cases (downsizing, paying off mortgage, moving to lower-cost area)

Likely to increase:

• Healthcare — especially before Medicare at 65
• Travel and leisure, especially in early retirement
• Hobbies and activities that replace work
• Long-term care, typically only in later retirement

Unchanged or mixed:

• Food, utilities, groceries
• Property tax, insurance
• Entertainment and dining out

A commonly cited heuristic: retirees spend 70–85% of their pre-retirement take-home pay. But this varies dramatically. Some retire to lower spending; others (especially in early retirement) find they spend more on travel and activities than they did while working. Use your own numbers.

Step 2: Account for guaranteed income

Sources of guaranteed retirement income:

Social Security. The largest guaranteed income for most retirees. Benefits are based on earning history and claiming age. Use ssa.gov to get an estimate for your exact situation. Benefits are inflation-adjusted annually.

Pensions. Fewer and fewer Americans have defined-benefit pensions, but where they exist, they reduce the portfolio need substantially. Note whether the pension has a COLA; non-adjusted pensions lose real value to inflation.

Annuities. Commercial annuities can convert a lump sum to guaranteed income. SPIAs (single-premium immediate annuities) are the simplest; more complex variable and indexed annuities exist but often have higher fees.

Rental income. Net of expenses, property management, and vacancy assumptions.

Part-time work. Many retirees work part-time in early retirement, which reduces the portfolio requirement during those years.

Step 3: Calculate the portfolio target

Target portfolio = (annual spending − annual guaranteed income) × withdrawal multiplier

The withdrawal multiplier depends on your assumptions:

• 4% rule: multiplier = 25
• 3.5%: multiplier = 28.6
• 3%: multiplier = 33.3

More conservative multipliers for:

• Very early retirees (40s or 50s) with long horizons
• Times of high stock valuations (expected future returns lower)
• Risk-averse planners

Less conservative multipliers (smaller) for:

• Later retirees (70+)
• Large guaranteed income buffers (Social Security + pension covering most expenses)
• Willingness to adjust spending in bad markets

Worked example

A couple planning to retire at 65:

• Estimated annual spending: $90,000 (in today's dollars)
• Expected Social Security (combined): $50,000/year
• Pension: $10,000/year (no COLA)
• Net portfolio need (year 1): $90,000 − $60,000 = $30,000
• Using 4% rule (25×): $30,000 × 25 = $750,000

Because the pension doesn't have a COLA, they need to plan for its real value declining. In practice this means adjusting the target up slightly.

Step 4: Account for lumpy expenses

The "25× spending" rule assumes constant annual spending. Real retirement has bumps:

• New car every 7–10 years
• Home maintenance and major repairs
• Healthcare deductibles and premiums
• Travel in early years
• Long-term care in later years

Some planners build a separate "sinking fund" for these, or add 5–10% to the portfolio target as a buffer.

How much you've saved: the percentage

Fidelity has popularized age-based savings benchmarks:

These are rough and depend on spending assumptions, but they're a reasonable gut check. If you're at age 50 with 6× your salary saved, you're roughly on track.

How Horizons uses this

Horizons calculates your retirement readiness score by comparing your projected portfolio at retirement against your projected spending need minus projected guaranteed income. The Monte Carlo engine runs this comparison across hundreds of scenarios to show not just whether the plan works on average but how robust it is across a range of market outcomes.