Skip to content

ANALYSIS

The data room.

Every chart, statistic, and line of arithmetic behind this site — drawn from my research files and the pinned public datasets the calculator runs on. Nothing on this page is fetched from the web; every number traces to a register row you can audit.

THE MARKET RECORD

What patient money has actually done

The whole argument rests on one property of broad-market investing: over long holding periods, the American market has never failed a patient investor. One dollar in January 1871 grew to about $1.06 million by January 2026 — about $40,600 after inflation.

Growth of $1 in large U.S. stocks, 1871–2026
LOG SCALE · DIVIDENDS REINVESTED · JANUARY VALUES
DERIVED-CALCR-038
Nominal (gold) and in 1871 purchasing power (teal, CPI-deflated). Computed with the calculator's own verified monthly chain — same formula, same pinned data.
Source: Robert J. Shiller dataset (shillerdata.com), pinned snapshot 2026-08-06 — the same data the calculator runs on. R-038/R-039
Annualized return by holding period, 1871–2026
EVERY ROLLING JAN→JAN WINDOW · NOMINAL TOTAL RETURN
DERIVED-CALCR-039
Single years ranged from −44.26% to +56.66%. Stretch the window and the floor rises: the worst 20-year run still returned +3.07% a year (1929–49), the worst 30-year +5.08%, the worst 40-year +5.45%. In real (inflation-adjusted) terms the worst 20-year window was +0.69% a year (1962–82) — thinner, but still never negative.
Source: Robert J. Shiller dataset (shillerdata.com), pinned snapshot 2026-08-06 — the same data the calculator runs on. R-038/R-039

This is my central exhibit, rebuilt from scratch on licensing-cleared data so you can check it: a one-year bet on stocks is a coin flip; a working lifetime in stocks has never lost. Social Security’s horizon is the longest in American finance — a worker taxed at 20 pays for a check at 70 — yet the trust fund is barred from holding a single share.

THE HUMAN STAKES

Longer lives, a lower floor

Two series from my files frame why the funding model matters. People collect far longer than the program’s designers planned for — and the official poverty floor beneath them is low. Poverty among Americans 65 and older fell from 35.2% in 1959 to 10.2% in 2022 — Social Security’s great achievement, and unfinished business: one retiree in ten is still below the line. Why that floor has never broken is the subject of Legacy System Flaws.

Remaining life expectancy at 65, 1900–2050
PERIOD LIFE EXPECTANCY · BY SEX · SSA ACTUARIAL STUDY 120
VERIFIED-PRIMARYR-040
In 1940 — five years after the program was designed — a 65-year-old man could expect 11.9 more years, a woman 13.4. The study projects 17.5 and 20.1 for 2026. Small gaps in the women's line are missing pages in the source workbook, left unfilled rather than interpolated.
Source: SSA Actuarial Study No. 120 (Bell & Miller), Table 10 — from my research files (unmodified official workbook). R-040
Americans 65+ below the official poverty line, 1959–2022
OFFICIAL POVERTY MEASURE · ALL RACES
VERIFIED-PRIMARYR-042
The plan's benefit floor is set at twice the poverty line precisely because of this chart: the fall from 35.2% is the program working; the plateau near 10% is the part the current funding model has never been able to finish.
Source: U.S. Census Bureau, Table 3 (hstpov3), CPS ASEC — from my research files (unmodified official export). R-042

THE ARITHMETIC

The numbers the plan stands on

The honest baseline first: the retirement trust fund runs low in 2033, after which continuing payroll taxes still cover about 81 cents of every scheduled dollar. The 75-year gap is 3.82% of taxable payroll — fixable today by a 3.65-point payroll-tax hike or a 22.4% across-the-board cut. The plan argues for a third road.

$1,456/mo
what today's average earner gets claiming at 62
Source: SSA PIA formula + bend points + AWI (ssa.gov/oact/cola/piaformula.html et al., Wayback-pinned) · R-012
$2,508/mo
the plan's floor — twice the Census poverty line (my target)
Source: 2 × R-014 (Census threshold) · R-013
9.36% / 7.09%
long-run market average since 1871, nominal / after inflation — used only as a labeled assumption
Source: Derived from Shiller pinned vintage, Jan 1871–Jan 2026 · R-018
20.3 / 23.1 yrs
average remaining life expectancy at 62, men / women
Source: ssa.gov/oact/STATS/table4c6.html — 2023 period life table (Wayback 20260723073511) · R-020

Worked examples you can re-run

These are the calculator’s golden test cases — hand-worked, locked in the test suite, and re-derived on every data refresh: $1 invested in January 1960 grew to $470.42 by January 2022 ($49.03 in 1960 dollars). $250 at a 1990 birth grew to $10,359.02 by 2026. $50 a month from 1980 through 2019 — $24,000 in — became $361,333.17.

Run your own birth year in the calculator →

Data values: the poverty floor over time

U.S. Census weighted-average annual thresholds, single person 65+, nominal dollars — from the my files (R-041). The plan’s floor doubles the current line.

YearThreshold, single 65+Threshold, single under 65CPI-U (1982–84=100)
1959$1,397$1,50329.2
1970$1,861$2,01038.8
1980$3,949$4,29082.4
1990$6,268$6,800130.7
2000$8,259$8,959172.2
2010$10,458$11,344218.1
2021$12,996$14,097271.0

THE WALL OF EVIDENCE

The charts I kept on my wall

Long before this site, I kept the market’s long record where I could see it every day — framed on my wall. These are my own copies: the Ibbotson SBBI chart from my MetLife Resources years, and the framed 2015 Morningstar Andex chart that hangs behind me in my portrait on the About Landon B. McDonald page.

Ibbotson Associates 'Stocks, Bonds, Bills, and Inflation 1925–1993' wealth-index chart from my MetLife years: $1 invested at year-end 1925 grows to $2,757.15 in small-company stocks and $800.08 in large-company stocks by 1993, against $28.03 for long-term government bonds, $11.73 for Treasury bills, and $8.13 of inflation, with historical events annotated along the curves.
Ibbotson Associates, Stocks, Bonds, Bills, and Inflation 1925–1993 — the wealth-index chart from my MetLife Resources years: $1 at year-end 1925 grew to $2,757.15 in small-company stocks and $800.08 in large-company stocks by 1993.© Ibbotson Associates — from my collection; reprint permission pending
The 2015 Morningstar Andex chart in its gold frame on my wall — nearly nine decades of U.S. market history on one poster: growth of $1 across asset classes on a log scale, with recessions, interest rates, and inflation tracked beneath.
The framed 2015 Morningstar Andex chart on my wall — the “picture worth 1000 words” he shows visitors first.© Morningstar — from my collection; reprint permission pending
It amazes me how people are often more willing to act based on little or no data than to use data that is a challenge to assemble.
Robert J. Shiller, American economist, academic, and best-selling authorfrom my quote collection
Though past performance provides no guarantees, “Reversion to the mean is still the most powerful force in finance.”
Steve Galbraith, Maverick Capital / Columbia Business SchoolCFA Institute Magazine, Jan–Feb 2007, p. 5 — from my quote collection

THE DATA LIBRARY

What sits behind this site

This site is the public face of a research library I assembled over two decades — about 22,800 cataloged files. The library stays offline; what publishes is only what survives the claims register: primary sources stated as fact, my own projections shown with the working behind them, and nothing in between.

37,912
daily Dow Jones closing values, 1885–2023 (MeasuringWorth series)
Source: research/phase0/MANIFEST.csv (count incl. header −1) · R-043
97 yrs
of official SBBI annual return data, 1926–2022, verified against published figures
Source: research/phase0/MANIFEST.csv (count incl. header −1) · R-043
194
sheets of SSA life tables — the full Actuarial Study 120 workbook, 1900–2100
Source: research/phase0/MANIFEST.csv (count incl. header −1) · R-043
63 yrs
of Census poverty thresholds and rates, 1959 onward
Source: research/phase0/MANIFEST.csv (count incl. header −1) · R-043

How to audit this page

Every number above carries a register ID (the R-xxx tag on each chart and stat). The register records the source file, the verification status, and the derivation for each one. Market figures are computed from the same pinned Shiller snapshot the calculator uses, with the formula disclosed on the methodology page. Government series (SSA life tables, Census poverty data) are unmodified official exports from my files.