Open a desk on the exchange: build a watchlist, follow a name’s price from 1880 to now, and get the full quote history.
Every year, the Social Security Administration publishes the name of nearly every baby born in America. Treat those names like stocks — with parents as the traders — and you get a 146-year market history: blue chips, hot IPOs, slow collapses, and the occasional meme-stock frenzy. Green marks a name heating up and red one cooling off; pink and blue are reserved for the girls’ and boys’ sides of the market.
In 2025, American parents gave 3.32 million babies a name common enough for the SSA to list — 51.5% boys, 48.5% girls. That’s the float of the name market, and it keeps shrinking: registered births peaked at 4.33 million in 2007 and have fallen almost every year since.
The chart below is the market’s trading volume since 1880. The early decades undercount badly — Social Security didn’t exist until 1935, so people born in the 1880s only appear if they later applied for a card. The Baby Boom is the huge swell from 1946; the 2008 recession starts the long modern slide.
The human sex ratio at birth is typically around 105 boys for every 100 girls, which works out to about:
This has been observed across populations for centuries, although it varies a bit by country and over time. The natural follow-up is: why?
The leading explanation is that it’s an evolutionary compensation for the fact that males have higher mortality throughout life, especially early in life. Historically, boys were more likely than girls to:
By producing a slight surplus of males at birth, the number of men and women who survive to reproductive age ends up closer to equal. This idea traces back to the statistician Ronald Fisher and what’s now called Fisher’s principle. The basic argument is elegant:
That explains why it’s close to 50:50. The persistent 105:100 bias is thought to reflect the extra mortality faced by males before reproduction.
Male fetuses and infants are, on average, biologically more vulnerable. Compared with females, they have higher rates of miscarriage later in pregnancy, stillbirth, premature birth, birth complications, and infant mortality. Part of this appears to be genetic:
In other words, nature starts with a few extra boys because more boys are lost before adulthood.
Not exactly. Typical natural populations fall between about 103 and 107 boys per 100 girls. The ratio can be nudged slightly by parental age, environmental stress, severe famines, and some environmental toxins. After major disasters or prolonged stress, researchers have sometimes observed a small dip in the proportion of male births, though the effect is modest and not universal.
The births curve isn’t random noise — nearly every big move lines up with history. (This is SSN-application data, so figures before ~1935 reflect record-keeping, not real fertility; the shape from mid-century on is solid.)
Each panel is total births over its own window, on its own scale. The axis starts near the low of the window rather than at zero — the only way a 5% move is visible at all — and every gridline is labelled so you can see what it cost. Where a panel carries two dashed lines, the dim one is the cause and the bright one is the print it lands in — rarely the same year, since a war ends nine months before the babies arrive. From 1940 on, a second line carries the total fertility rate on the right-hand axis: births are that rate multiplied by the number of women of childbearing age, so when the two lines disagree, the population is doing the talking rather than the parents.
Three slow forces do most of the work: the economy (births rise in good times, fall in recessions), war (disrupts, then rebounds), and reproductive technology and policy — the Pill (1960) and Roe (1973) drove the 1965–76 collapse more than anything else. These operate over years, not overnight.
The intuitive story — a blackout, storm, or pandemic, then a baby boom nine months later — is basically false. Demographer J. Richard Udry debunked the famous 1965 New York blackout “boom” in 1970: births were dead normal, and the legend traced back to a few anecdotal newspaper interviews. Stress and disaster, if anything, slightly reduce births. COVID produced a bust, not a boom.
Total births = the fertility rate × the number of women of childbearing age. The 1976–90 rise was mostly more Boomer women, not more babies per couple. The cleaner measure — the total fertility rate — has sat below the replacement level of 2.1 since 1971 and hit a record-low 1.62 in 2023. And the post-2008 decline looks less like a dip than a structural shift (delayed childbearing, student debt, housing and childcare costs, collapsing teen births) that acute events only nudge.
The dynasty table ranks spellings, because that is all the SSA publishes. Sophia and Sofia are one name filed under two headings, and neither heading is ever credited with the other’s babies. Add each name’s true respellings back together, recompute the crown year by year, and a second line of succession appears — one that has never once been printed.
Nothing here is concealed. The SSA ranks Sophia fifth in 2025, and all five of its spellings are printed on the same public list every May: together they come to 25,021 babies, nearly double the 13,544 girls officially crowned as Olivia, and past Liam’s 20,818 too — which makes Sophia the most-given name in the country to a baby of either sex. It stays invisible only because nobody adds the column up.
The tape as published, beside the same ten with respellings merged. The arrow is how far a name moves once its own spellings stop competing with it; the ×N badge counts the spellings folded in — hover for the split. Click any name to chart it.
Spelling variants are the clean case: Sophia and Sofia are unarguably the same name. Nicknames are murkier — should Theo fold into Theodore?
But is Theo a nickname or its own name now? Thousands of parents deliberately put Theo — not Theodore — on the birth certificate, and the SSA is faithfully recording that choice. Folding nicknames imposes an opinion about what counts as “the same name”; merging spellings only corrects for orthography. So this desk merges spellings and leaves nicknames as the fuzzier story the data can’t settle.
Over 30 years the leaderboard has completely turned over. Liam — a niche pick in 1995 — gained more market share than any name in the country. Michael lost the most of any name on record: 2.2% of all boys in 1995, under 0.5% today.
A falling name is still a big name — 8,094 boys were named Michael in 2025, still a top-25 boys’ name. But in this market, yesterday’s index fund is tomorrow’s “dad name.”
A name crossing the SSA’s 5-baby floor for the first time is an IPO; a name that drops below it and falls off the tape is delisted. In 2025 the market minted 5,902 new listings — 1,238 of them names never recorded in 146 years — and delisted 6,579. Out of ~31,000 listed names, roughly a fifth turns over at the edges every year.
Below the big board is the pink-sheets market: names given to fewer than 200 babies a year, barely above the SSA’s 5-baby listing floor. Most stay obscure forever. A few go vertical — Wrenley traded at 13 babies in 2014 and 1,858 in 2025.
The momentum screen hunts for the next Wrenley: names still tiny in 2025 that have risen every single year since 2022, at least 2.5× overall. The board below it is proof the trade can pay. The fine print under that is proof it usually doesn’t.
One row on that board is not like the others. The name at the top of it moved ×204 — the largest penny trade in the data — and it is the only name on this page that got there with no reason attached.
Most names drift. A few get shoved — a song hits, a show premieres, a scandal leads the newscast — and the birth records answer within months. These are the market’s sharpest single-event moves.
Real events, real prices. Click any card to open the full chart in the terminal.
One famous person can move this market overnight. Each chart marks the event — a breakout, a premiere, a death — with a dashed line; what follows is the trade.
Liam and Olivia dominate coast to coast, but not everywhere — and every state has local specialties that trade far above their national price. Utah loves Mckay at 61× the national rate; in Hawaii, Mahina trades at 259×.
Tap a state for its leaders and its most distinctive names.
Its top names and local specialties will appear here. Map colors show each state’s #1 name.
Some names map the country’s geography. Parents reach for what’s outside the window: Ocean clusters on the coasts and vanishes across the landlocked Midwest, Canyon belongs to the desert Southwest, Forest to the rainy Pacific Northwest. Each mini-map shades every state by how much more common the name is there than nationally — bright teal for the strongholds, dim teal where it’s barely used, neutral grey where it isn’t used at all.
The SSA records every baby as M or F — there is no unisex or non-binary category in this data — and counts each name separately per sex, so we measure instead of label: a name is shared in a given year when at least 20% of its babies are the minority sex. By that bar, almost no American babies had a shared name for over a century — the share hovered near 2% from 1880 to 1990. It has tripled since, peaking at 6.4% in 2022 before easing to 5.8% in 2025, when 47 names with 500 or more babies managed even a 30–70 split. The biggest are on the board below, where Charlie trades at a perfect 50/50; click any name to pull the full quote.
Browse mid-century SSA data and you’ll find a few hundred girl Johns, boy Jessicas, and boy Marys in almost every year. Most are not naming trends—they’re data artifacts.
Until the late 1980s, Social Security numbers were issued from handwritten applications keyed by clerks, and about 0.3% of records had the sex field entered incorrectly. Applied to 90,000+ annual boy Johns, that error rate alone produces a few hundred phantom girl Johns. The signature is unmistakable: recorded girl Johns rise and fall almost perfectly in proportion to boy Johns, at roughly 1:300.
That relationship largely disappears after Enumeration at Birth, introduced beginning in 1987, when Social Security numbers began being assigned directly from hospital birth certificates. Recorded girl Johns fall from 233 in 1985 to just 6 in 2025.
The line never reaches zero, though, because a small number of these records are genuine. Family surnames become given names, traditions like Johnnie persist, and occasional examples—such as actress Michael Learned—have always existed. Enumeration at Birth removed the clerical artifacts; it revealed the underlying baseline.
The 20% shared-name threshold used in this analysis sits comfortably above both the historical error rate and the small population of genuinely cross-sex names.
The biggest change in this market is concentration. In 1950, a third of American boys had a top-10 name; in 2025 it was 8%, and the girls’ market is even more dispersed.
Economists would call it de-concentration. In 1950 the girls’ market behaved like it had roughly 100 names in circulation; by 2025 it behaved like 643. Parents stopped buying the index and started stock-picking — which is exactly why the spikes and crashes above have gotten sharper: with no giant incumbents, a small name can moon in a season.
Counting names tells you almost nothing. The SSA recorded 17,297 distinct girls’ names in 2025, but most of them went to a handful of babies each. What you want is a measure that weighs every name by how much of the market it actually holds.
One already exists, and it comes from antitrust. The Herfindahl–Hirschman Index is what the Justice Department and the FTC use to judge whether a merger would leave an industry too concentrated. The recipe is short: take every player’s market share, square it, and add up the squares.
Squaring is the whole trick. It makes large players count for disproportionately more than small ones — a name holding 10% of the market contributes a hundred times what a name holding 1% does. A long tail of rare names barely registers, which is exactly the behaviour you want.
Raw HHI reads backwards — higher means less competition — and its units mean nothing on their own. So flip it over: 1 ÷ HHI. The reciprocal is a number you can actually picture. Two markets, four names each:
That is how every figure on this chart should be read: the market behaves as if this many equally popular names were competing. It is why 17,297 recorded girls’ names in 2025 come out as 643.
Antitrust states shares in whole percentages rather than fractions, so HHI runs from 0 to 10,000 “points” instead of 0 to 1. The 2023 Merger Guidelines call anything above 1,800 points highly concentrated — the equivalent of about five or six equal competitors.
By that yardstick, baby names are absurdly competitive. The 2025 girls’ market scores about 16 points, boys about 21. Even 1880, the most concentrated year anywhere in this data, only reaches 286 for boys and 133 for girls. No year in 146 years comes within an order of magnitude of what a regulator would frown at — this market’s “firms” are simply far too numerous.
Ecologists arrived at the identical formula independently. Sum the squared shares of every species and invert it and you get the effective number of species in a habitat — the inverse Simpson index. A forest with one dominant tree and a scatter of rarities is concentrated in precisely the sense a market is. Naming fashion, market share, and biodiversity turn out to be the same measurement problem.
6,000+ names are listed here — every name that reached 150 babies in a single year since 1880, plus the penny desk’s momentum picks and the landscape names. Type one and pull its full price history.
Every figure on this page comes from one public dataset, and that dataset has sharp edges. Here is what it counts, what it quietly drops, and which stretches of the record you should not read as history.
Every figure on this page comes from the Social Security Administration’s public baby-names files — national counts from 1880, state counts from 1910, republished each May. This page uses the 2025 release, in which a birth means an SSN card application filed under that year of birth. ssa.gov/oact/babynames
The SSA withholds any name given to fewer than 5 babies of a sex in a year. That is the 5-baby floor §3 trades against: cross it and a name lists, fall below it and it drops off the tape. A name that never crosses it is invisible here — which is why the 2024 “named” total (3.33M) runs under actual births (3.61M). The gap is thousands of names trading below the floor.
Listings & delistingsThe SSA ranks spellings, not names: Sophia and Sofia file separately, as do Aiden and Ayden. So every ranking on this page — and everywhere else — measures spelling consensus as much as popularity, and the name at #1 is not the country’s most-given name. “America’s real #1 name,” under the dynasty table in §1, merges the clusters by hand and re-ranks.
Social Security began in 1935, so earlier birth years only include people who lived long enough to get a card. Those volumes are undercounts, and the sex mix reflects who enrolled rather than who was born: 1880s cohorts skew male (numbers came through employment), 1890–1930 cohorts skew heavily female (longer-lived survivors enrolling via 1960s benefit programs). Only from ~1940 on does the data show true births — about 105 boys per 100 girls.
Raw counts mislead across eras — there were twice as many babies in 1957 as in 1935 — so every trend uses a name’s rate per million births of that sex. In 2025, 13,544 girls were named Olivia out of 1,607,267 named girls: 8,427 per million, or about 0.84%. Long windows compare 3-year averages (1994–96 vs. 2023–25) so a single odd year can’t fake a trend. Volatility is the standard deviation of year-over-year log changes, 1995–2025, for names listed ≥27 of 31 years averaging ≥30 per million.