One common investing mistake more than triples the risk of running out of money in retirement, according to a new analysis of 154 years of market data. The mistake is selling stocks after a crash and buying back later.

The analysis, published at firenum.com, replayed every 30-year retirement since 1871 using the Shiller S&P 500 dataset. A retiree who stayed invested through downturns saw a 98.4% success rate. A retiree who carried a "behavior gap" of 1.5% a year, the cost of selling low and buying back high, succeeded only 79.3% of the time. The failure rate rose from 1.6% to 20.7%.

For early retirees the stakes climb. Over a 40-year retirement, the typical length for someone who leaves work in their mid-40s, staying invested succeeded 86% of the time. Panicking dropped that to 46%, worse than a coin flip.

Line chart of how often a 60/40 retirement lasted, every monthly start since 1871, 4% withdrawn and raised 3% a year: staying invested lasted 98.4% of the time over 30 years and 86.2% over 40; with an assumed 1.5% a year behavior gap, 79.3% over 30 years and 46.4% over 40, falling to 28.5% over 50.
A 1.5%/yr behavior gap cuts a 40-year retirement's odds from 86% to 46%. Over 30 years: 98.4% against 79.3%. Free to use under CC BY 4.0: credit firenum.com with a link.

When the mistake changed the outcome, it drained the portfolio a median of 2.2 years sooner over 30 years and 7.7 years sooner over 40.

Bar chart of the median years a 60/40 retirement ran out sooner with an assumed 1.5% a year behavior gap, among starts since 1871 where the gap shortened it: 1.5 years over 20 years, 2.1 over 25, 2.2 over 30, 5.8 over 35, 7.7 over 40, 9.8 over 45 and 11.8 over 50.
When the behavior gap bites, the cost grows with the retirement: 2.2 years lost over 30 years, 7.7 over 40, 11.8 over 50. Free to use under CC BY 4.0: credit firenum.com with a link.

The 1.5% is an assumption, not a measurement of any one investor. It is the size Vanguard's research on "advisor's alpha" puts on behavioral coaching, the work of keeping clients invested during a crash: about 1.5% a year, roughly half of an advisor's estimated value. Dalbar's annual study of real investor returns shows how much the gap swings: the average equity investor trailed the S&P 500 by 8.48 points in 2024 and by 0.72 points in 2025 (Dalbar 2026 QAIB release, 16-04-2026). The firenum analysis turns a 1.5% gap into years of lost retirement over FIRE-length horizons.

The study also names the catch most coverage omits. A 1% assets-under-management fee on a $1M portfolio costs $10,000 a year and rises as the portfolio grows, consuming most of the 1.5% an advisor adds. A flat-fee or hourly advisor charges the same at any portfolio size, so the value stays with the client.

We make planning software, so people expect us to say you don't need an advisor. The numbers say otherwise. Staying invested through a crash is worth more than any feature we could build, and the fairest way to pay for that help is a flat fee.

the team behind firenum.com

This release was distributed on the wire.

The takeaway for do-it-yourself planners is a written plan they trust enough to hold during a downturn. A plan stress-tested against a 40% crash in advance is easier to keep when markets fall. For those who want a human in their corner, the analysis points to fee-only advisors who charge flat or hourly rates. The full study and methodology are at firenum.com/blog/is-a-financial-advisor-worth-it.

About firenum.com

firenum.com is a free retirement planning tool with Monte Carlo simulation, historical backtesting and stress testing. No account needed. The planner runs in your browser and your plan is saved there; a free account keeps it on any device.

Media contact

Writing about this research? Reach us at contact@firenum.com for data, methodology details, or quotes. Sources: Vanguard Advisor's Alpha (Kinniry et al.), Dalbar QAIB, Shiller S&P 500 dataset (1871-present), Bengen (1994).

Stress-test your own plan against a 40% crash.