Taking Social Security at 62 or 70? Let's actually run it.
The usual claiming advice is a rule of thumb. This post builds a specific household and runs the decision through a Monte Carlo engine instead.
Rules of thumb (wait until 70, spend 4 percent, keep three years of cash) cannot see your taxes, your health coverage, or your odds of the money lasting. I spent my career as a quantitative analyst and credentialed actuary building the stochastic models insurance companies use to test whether their promises hold up over decades. I'm keeping my name private for now. The method here is the same: build the household, run it, report the numbers.
The rule of thumb
The standard advice is to wait. Full retirement age is 67 for anyone born in 1960 or later; claiming at 62 takes a permanent cut of about 30 percent, and holding out to 70 earns delayed credits worth about 8 percent a year, landing 24 percent above the full benefit. The Social Security Administration publishes the early-claiming reduction and the delayed-retirement credits.
But a bigger check that starts eight years later means eight years of living off savings instead. Whether that trade works depends on lifespan, taxes, and whether the portfolio can absorb the gap, so I built the household.
The household
A single filer, age 62, retiring now, with $850,000 saved: $600,000 in a traditional 401(k), $150,000 in a taxable brokerage account, and $100,000 in a Roth. She spends $54,000 a year, lives in Florida (no state income tax), and her benefit at full retirement age would be $2,400 a month.
Two runs differ by one decision: claim at 62 for $1,680 a month, or wait until 70 for $2,976 a month, 77 percent more per year and inflation-adjusted either way. Each run is 2,000 simulated lifetimes, each with its own market path and a lifespan drawn from standard mortality tables.
Illustrative projection from the Seraph engine v1.51, August 2026 (2,000 trials, today's dollars), not a forecast of your results or financial advice. One household, one set of assumptions. Refreshed from the original July run (engine v1.38): the engine now models marketplace premium subsidies, which raised the claim-at-62 numbers and removed the age-88 crossover the original showed.
What the numbers say
The two paths land in a dead heat: about 60 percent either way, nothing like what the 77 percent headline suggests. And for most of her retirement, claiming early leaves her better off, because she is not draining savings to bridge the eight-year gap; in this run the wait-to-70 path never catches up, and the odds converge only in the long-life tail.
Median Net Worth
Median liquid net worth by age, today's dollars
The odds tell the same story: claiming early is at least as safe at essentially every age, and by 95 the two are even.
| At age | Claim at 62 | Wait to 70 |
|---|---|---|
| 80 | 91% | 87% |
| 85 | 78% | 76% |
| 95 | 62% | 62% |
Waiting until 70 is not free money. It is spending down savings through the 60s to buy a larger, inflation-protected income, and with her simulated life expectancy around 87, the bigger checks arrive too late to repay the drawdown in the median outcome. Delaying is still a longevity hedge, insurance against outliving your money, but for this household the model prices the insurance at about what it pays: even the tail odds at 95 come out level.
Two things the claiming advice skips
First, retiring at 62 means buying health insurance until Medicare starts at 65, about $15,000 a year here, and who actually pays it depends on the claiming decision. Claiming at 62 covers $20,160 of spending with Social Security, so the portfolio withdrawals shrink and her income for subsidy purposes stays near $46,000 to $55,000, low enough that marketplace premium subsidies cover nearly the whole bill. Waiting to 70 funds everything from the portfolio, which pushes that income above the subsidy cutoff, and she pays full price all three years. Claiming early is worth roughly $15,000 a year in subsidies to this household, and that interaction, not the benefit arithmetic, is why the bigger check never pulls ahead.
Second, the plan is only a 60 percent proposition either way. The claiming decision barely moves it; trimming spending or working two more years would move it far more.
What changed in 2026
Benefits rose 2.8 percent this year through the annual cost-of-living adjustment. The program's trustees project the retirement trust fund can pay full benefits only until 2032, after which, absent action from Congress, scheduled benefits would drop to about 78 percent (2026 Trustees Report summary). That does not change the 62-versus-70 arithmetic directly, but it is a reason not to treat the larger check at 70 as ironclad; we priced that cut for this same household in a follow-up post.
The 2025 tax law also added a temporary $6,000 deduction for filers aged 65 and older, through 2028 and phasing out at higher incomes (IRS guidance). It changes how much of a retiree's income, Social Security included, gets taxed, and it interacts with claiming timing.
Run your own household.
The answer moves with your savings mix, your spending, and your family's longevity. The app runs the same 2,000-lifetime projection on your numbers.
Run your own numbersThe real questions are how long you expect to live, how tight the plan already is, and how much of your spending Social Security needs to cover. The bigger check is real; whether it is worth eight years of drawdown is a question you can answer with numbers.
Sources
- Social Security Administration, "Retirement Age and Benefit Reduction" and "Delayed Retirement Credits."
- Social Security Administration, "Cost-of-Living Adjustment (COLA)" for 2026.
- Social Security and Medicare Boards of Trustees, 2026 Trustees Report summary.
- Internal Revenue Service, One Big Beautiful Bill Act: deductions for working Americans and seniors.
- Projections generated with the Seraph: Retirement Planner engine, v1.51 (August 2026; originally published on v1.38). Figures are illustrative, in today's dollars, for the sample household described above.