Chapter 15: The overlap audit
Set an S&P 500 fund beside a total US market fund and count the names their top-10 lists share. The answer, taken straight from the funds' own pages (Vanguard data as of April 30, 2026), is 10 of 10. Split $10,000 evenly between the two and just over a third of your money, 36.1%, lands in those ten shared names, because the top ten makes up 38.4% of the first fund and 33.7% of the second. Two tickets, two labels, one bet.
Chapter 12 watched Mara discover this the uncomfortable way, when the eleven funds spread across her five accounts collapsed into four actual bets, and it taught the pairwise version of the check. Chapter 14 then built the clean three-line answer. This chapter sits between those ideas and turns the discovery into a habit: the overlap audit, a two-minute count you run before any fund purchase and once a year afterward, in the owner's hour Chapter 16 schedules. Every number it uses is printed on the fund pages you learned to read in Chapter 4, and this chapter attaches the receipts.
Why every broad fund leads with the same ten names
Most broad funds weight their holdings by company size, so the largest American companies sit at the top of every broad American list, in nearly the same order, at nearly the same weights. That is arithmetic, not laziness. The S&P 500 covers approximately 80% of available US market value (S&P Dow Jones Indices), which makes a total US market fund the S&P 500 plus a long, thin tail of smaller companies. Any fund that buys large US stocks by their market weight must lead with the same giants, whatever name is on the door.
Overlap, in other words, is structural. The audit's job is not to eliminate it, which would mean abandoning broad index funds altogether. The job is to make every duplication a stated choice instead of a stowaway, because unstated overlap quietly turns a portfolio far more concentrated than its owner believes. Chapter 12's one-job-per-fund rule supplies the test.
The receipts, three pairings at a time
The calculator below puts real top-10 lists side by side: holdings and weights from the funds' own pages, Vanguard data as of April 30, 2026 and Invesco Nasdaq-100 data as of June 10, 2026. Shaded rows appear in both lists. One honest counting note before you read it: Alphabet appears twice in the US lists because it issues two share classes, and the audit counts each line as the fund page prints it.
The first pairing is the one this chapter opened with. An S&P 500 fund and a total US market fund share 10 of 10 top names, and a 50/50 split holds 36.1% of the money in just those shared names. The calculator's verdict line calls it the same bet, twice. The tail of smaller companies is real, but anyone holding both should be able to say why, and diversification is not the reason.
The second pairing adds the fund people buy for growth. An S&P 500 fund and a Nasdaq-100 fund share 8 of 10 top names. The only two names on the Nasdaq side that the S&P side's top ten lacks are Micron and AMD, which take the slots Meta and Berkshire occupy on the S&P list, and an S&P 500 fund owns Micron and AMD anyway, just below its top ten. A 50/50 split holds 36.0% in the shared names, essentially the same concentration as the first pairing, with one difference that matters: the Nasdaq-100 fund's top ten is 45.7% of the whole fund, against 38.4% for the S&P 500 fund. Buying it next to an S&P 500 fund is not a second bet. It is the first bet with the dial turned up.
The third pairing shows what passing looks like. A total US fund and a total international fund share 0 of 10 top names, and 0.0% of a 50/50 split sits in shared names. The international fund's entire top ten is just 12.0% of the fund, the lightest concentration of the four lists. These are Chapter 14's first two lines, and they pass the audit by construction: genuinely different baskets, each with a job the other cannot do.
How do you run the overlap audit in two minutes?
Lay two top-10 lists side by side and count the shared names, then glance at the sector weights. That is the whole mechanic; the discipline is running it across everything you own at once, because overlap hides in the gaps between account statements. The card below is the full routine.
- List every fund across every account. The audit unit is the household portfolio, not the statement. Mara's five accounts each looked sensible alone.
- Pull each fund's top-10 holdings. Chapter 4, line 2: every fund page prints the list and the weights.
- Count shared names, pair by pair. Seven or more of ten means one bet wearing two names. Four to six deserves the sector check. Three or fewer is genuine variety.
- Check sector weights for the middle cases. Same sectors, same order, similar sizes: one bet in two costumes.
- Assign each fund a job (Chapter 12). Two funds with the same job is a replacement decision, never an addition.
- Fix it where fixing is free. Exchanges inside tax-sheltered accounts trigger no tax bill (Chapter 2). In taxable accounts, read the section below before selling anything.
Mara's audit, in dollars
Chapter 12 told the story; here is the ledger. Eleven funds, $310,000, four bets.
| The bet | Funds held | Dollars | Share of $310,000 |
|---|---|---|---|
| US giant companies | 6 | $186,000 | 60% |
| Bonds | 2 | $54,250 | 17.5% |
| International stocks | 2 | $46,500 | 15% |
| US small companies | 1 | $23,250 | 7.5% |
| Total | 11 | $310,000 | 100% |
The first row is the one that stopped her. Six funds with six different names, $186,000, 60% of everything she owned, riding a single bet she had bought six separate times across a decade of job changes.
The audit itself took one evening and a kitchen table. Mara printed eleven top-10 lists and started counting: the six US funds shared seven to nine of their ten names in every pairing, exactly the same-bet territory the checklist flags. The fixes were mostly free. Five of those six funds became the cheap total-market index fund that won Chapter 4's checkup, exchanged inside tax-sheltered accounts where swaps trigger no tax bill (Chapter 2). The small-cap fund went too, once she confirmed the total-market fund already owns those companies in its tail at their market weight. She kept her current 401(k)'s S&P 500 fund, because that menu offers nothing broader, and she can name the overlap out loud: it is the calculator's first pairing, 10 names of 10 shared with her index fund, a known twin rather than a stowaway. Eleven tickets became four lines with four jobs, and the $186,000 that had been one accidental bet became one deliberate one.
Hugo's two minutes
The steakhouse adviser from Chapters 10 and 11 followed up with Hugo this month, pitching a Nasdaq-100 fund as "growth diversification" for the S&P 500 index fund that holds most of Hugo's $1.4 million. Hugo ran the audit before replying. Shared top-10 names: 8 of 10. The two additions, Micron and AMD, simply replace Meta and Berkshire near the top of a list he already owns, and his index fund holds both of them below its top ten anyway. The Nasdaq fund's top ten came to 45.7% of the fund against his fund's 38.4%, so the pitch amounted to the same bet with the volume raised, at several times the expense ratio. His reply ran one sentence: thanks, but I already own this. Two minutes of counting, and the seller's-pitch test never even needed to leave its holster.
What should you do with overlap you already own?
Fix it free where you can, and slowly where you cannot. Inside a 401(k) or IRA, exchanging an overlapping fund for the one that should hold the job triggers no tax bill, which is why Mara's cleanup cost her nothing (Chapter 2 explained that plumbing). A taxable account is different: selling a fund you have held for years can trigger capital gains tax, and the audit never orders a fire sale. You have three gentler tools. Keep the overlapping fund and stop feeding it. Point every new dollar at the job that is actually uncovered, so the portfolio grows toward its target instead of trading toward it. And if you do trim, prefer shares with small gains or with losses. The sell-versus-hold arithmetic, tax lots, and rates belong to the Finvest Tax Playbook. The audit's goal is a portfolio where every duplication is a choice you can state in one sentence; zero shared holdings is neither possible nor the point.
Before buying any fund, set its top-10 list beside what you already own and count the shared names: seven or more of ten is a replacement decision, never an addition. Rerun the count across all accounts once a year, and require every fund to hold a job no other fund already holds.
Where people go wrong
Auditing one account at a time. Each of Mara's five statements looked diversified on its own; the $186,000 single bet only appeared when the eleven lists hit one table. Overlap lives in the gaps between accounts, so the audit covers everything or it catches nothing.
Trusting the names. Growth, blue chip, dividend, equity income: four labels, one basket. The Names Rule (Chapter 4) forces a name to point at its holdings, but it cannot force two different names to point at different holdings. Only the lists settle it.
Counting funds as diversification. Chapter 1 flagged this myth on day one. Eleven funds can be four bets, and Chapter 14's three funds hold more securities than Mara's eleven did. Diversification is a property of the holdings, never of the number of tickets.
Treating all overlap as an emergency. A known twin kept for a 401(k) menu reason, or a taxable fund kept to avoid volunteering for a tax bill, is stated overlap, and stated overlap is fine. The dangerous kind is the kind discovered by accident, usually during a crash, when all the different names fall in unison.
Key takeaways
- Overlap is structural: broad funds weight by company size, and the S&P 500 covers approximately 80% of available US market value (S&P DJI), so every broad US fund leads with the same giants. The audit makes duplication stated, never accidental.
- The receipts, from the funds' own pages (Vanguard data as of April 30, 2026; Invesco data as of June 10, 2026): S&P 500 + total US share 10 of 10 top names, with 36.1% of a 50/50 split in those names; S&P 500 + Nasdaq-100 share 8 of 10, with 36.0%, and only Micron and AMD replace Meta and Berkshire; total US + international share 0 of 10, with 0.0%.
- Concentration differs more than contents: the four top-10 lists are 38.4%, 33.7%, 45.7%, and 12.0% of their funds, so a Nasdaq-100 fund next to an S&P 500 fund is the same bet with the dial turned up.
- The two-minute audit: every fund, every account, top-10 lists side by side; seven or more shared names of ten means replacement, not addition; one job per fund (Chapter 12).
- Mara's ledger: $186,000 + $46,500 + $54,250 + $23,250 = $310,000, with 60% in one bet bought six times. The cleanup was free inside tax-sheltered accounts; in taxable accounts, redirect new money first and take the sell-versus-hold math to the Tax Playbook.
Sources: S&P DJI S&P 500 factsheet · Vanguard fund pages (holdings as of April 30, 2026) · Invesco QQQ fund page (holdings as of June 10, 2026) · Investor.gov on ETFs · Finvest Tax Playbook