Owning several funds does not mean owning several different things. When funds hold the same companies, the exposure to each company adds up across the funds, and the total can be far larger than any single fund's factsheet suggests. The fix is a look-through calculation: multiply each fund's weight in a company by the share of the portfolio held in that fund, then add the results.

Key takeaways

  • In a four-fund illustrative portfolio built from holdings files dated October 14, 2026, the top three companies add up to 27.88% of the portfolio, against 21.53% in the S&P 500 fund alone.
  • Look-through exposure to one company is the sum of (portfolio share of each fund x that fund's weight in the company). Different share classes of one issuer must be added together.
  • The overlap between two funds, the sum of the smaller weight for every shared holding, ranged from 35.8% to 92.6% across the six pairs in the example. A high percentage says funds are similar; it does not say which names carry it.
  • Holdings are published on fund websites (daily for the funds used here) and in SEC Form N-PORT filings, where only the quarter-end month is public and can arrive up to 60 days late. A February 2026 SEC proposal would keep that quarterly pattern.
  • The 25% single-issuer tax test for regulated investment companies is applied to each fund separately, at quarter-ends, and says nothing about the same stock held through several funds.

How overlap builds up

Overlap comes from construction, not carelessness. Four common routes put the same stocks in several funds at once.

A single fund's concentration in its biggest stocks is a topic of its own. What matters here is the extra layer: the same names counted again in funds that look separate.

Look-through exposure, step by step

The formula is a weighted sum across funds:

Exposure to company i = (share of fund 1 x weight of i in fund 1)
                      + (share of fund 2 x weight of i in fund 2)
                      + ... for every fund held

Each portfolio share is the fund's value divided by the total portfolio value. Each weight is the company's percentage of that fund, taken from its holdings file.

Worked example with published holdings

Illustrative example: a hypothetical portfolio is split 40% / 25% / 20% / 15% across four funds. The split is invented for the arithmetic and is not a suggestion. The weights come from the funds' own holdings files, all dated October 14, 2026, from one provider: the State Street SPDR S&P 500 ETF Trust (SPY), the State Street SPDR Portfolio S&P 500 Growth ETF (SPYG), the State Street Technology Select Sector SPDR ETF (XLK) and the State Street SPDR Portfolio S&P 1500 Composite Stock Market ETF (SPTM). These funds are named only to show the mechanism with their own published data; naming them is not a recommendation to buy, hold or avoid any of them, and funds from other providers built on similar indexes would show the same pattern.

For NVIDIA, the largest holding in all four:

0.40 x 8.05%  = 3.220
0.25 x 14.58% = 3.645
0.20 x 15.38% = 3.076
0.15 x 7.46%  = 1.119
Total look-through weight = 11.060, or 11.06% of the portfolio

No fund shows 11%. The highest single-fund weight is 15.38% in the technology fund, but that fund is only 20% of the portfolio. On a $100,000 portfolio, 11.06% is $11,060 in one company.

Company S&P 500 fund (40%) Growth fund (25%) Tech fund (20%) Broad fund (15%) Look-through
NVIDIA 8.05% 14.58% 15.38% 7.46% 11.06%
Apple 7.53% 6.81% 14.38% 6.97% 8.63%
Microsoft 5.96% 10.79% 11.38% 5.52% 8.18%
Alphabet (both share classes) 5.52% 10.00% 0% 5.12% 5.48%
Broadcom 2.48% 4.48% 4.73% 2.29% 3.40%

The look-through column uses the unrounded published weights, so recomputing from the rounded figures shown can differ by 0.01. Alphabet appears in the files as two lines (Class A and Class C); the table adds them, because both are claims on the same company. Matching by ticker alone would count it as two names and understate the exposure.

What the numbers show

Across all four files the portfolio holds about 1,500 different companies (after merging share classes and leaving out cash, money market, futures and other non-stock lines), and every one of them is already in the broad fund. Yet its ten largest look-through positions add up to 48.25% of it. The top three (NVIDIA, Apple, Microsoft) are 27.88%. Held alone, the S&P 500 fund's top three are 21.53% and the broad fund's are 19.95%. The extra six to eight points (27.88 minus 21.53 is 6.35; 27.88 minus 19.95 is 7.93) come from the growth and technology funds layering more weight onto the same names.

A sensitivity check turns this into a magnitude. A 20% decline in just those ten companies, with everything else unchanged, would reduce the portfolio's value by 0.20 x 48.25% = 9.65%. For comparison, the same 20% decline applied to the broad fund's own ten largest holdings (37.35% of that fund) would cost that fund 7.47%. This is arithmetic, not a forecast. It shows how much of the portfolio rides on a few names, which is what the number of funds hides.

The result is also not an error in anyone's portfolio. In a market-capitalization-weighted index, large companies are large because the market values them highly, and an S&P 500 fund simply reflects that. The question is whether the look-through total matches the concentration the investor intended. Daily holdings files change, so the figures above describe one date only.

Measuring overlap between two funds

Look-through exposure follows one company across funds. Pairwise overlap compares two funds as wholes. A common measure adds, for every security either fund holds, the smaller of its two weights:

Overlap(A, B) = sum over all securities i of min(weight of i in A, weight of i in B)

If the funds are identical the result is 100%; if they share no holdings it is 0%. Securities held by only one fund contribute zero. Share classes of one issuer are merged first.

Fund pair Overlap Why
S&P 500 fund and broad fund 92.6% The broad index contains every S&P 500 stock
S&P 500 fund and growth fund 66.6% The growth index is drawn from the S&P 500
Growth fund and broad fund 62.6% Same nesting, one step removed
Growth fund and tech fund 53.2% Large growth names are mostly technology names
S&P 500 fund and tech fund 38.7% The tech fund is a slice of the S&P 500
Tech fund and broad fund 35.8% Technology is a smaller share of the broader index

Computed from the October 14, 2026 holdings files, with share classes of the same company merged (Alphabet, Fox and News Corp in these files) and cash, money market and futures lines left out, since they are not shared companies. Two cautions apply. First, a 92.6% overlap between the S&P 500 fund and the broad fund is a design feature, not a defect, because the broad index is mostly the S&P 500 in a larger wrapper; the overlap tells the investor that holding both buys little new. Second, the percentage does not show where the overlap sits. Two funds can overlap 40% because they share a few giant names or because they share hundreds of small ones, and only the look-through table distinguishes the two.

Sector and factor overlap is a separate layer. Funds with no shared holdings can still move together if they lean on the same sector, so comparing sector weights across funds adds a second view. Sector labels are classifications chosen by the index or data provider, and different providers can draw the lines differently.

Where to find holdings data

Two public sources cover most U.S.-registered funds.

  • The fund's own website. Many ETF providers post the full holdings as a downloadable spreadsheet. The four funds above publish a daily file; the technology fund's page carries a "Download All Holdings: Daily" link and the file states its as-of date. Other providers may publish less often, so the as-of date matters.
  • SEC Form N-PORT on EDGAR. Under 17 CFR 270.30b1-9, registered management investment companies and ETFs organized as unit investment trusts file a monthly report of holdings on Form N-PORT. Money market funds and small business investment companies are excluded. The reports for each fiscal quarter are due within 60 days after quarter-end.

Not every monthly filing is public. According to the SEC's 2024 announcement of amendments, only the report for the third month of each fiscal quarter is made public under current practice. Those 2024 amendments, which would require monthly filing within 30 days and make every month public 60 days after it ends, are not in force: the eCFR text of the rule notes they were delayed to November 17, 2027, and May 18, 2028, the later date for smaller fund groups.

The 2024 changes may never take effect in that form. On February 18, 2026, the SEC proposed amendments to Form N-PORT (Release IC-35962) that would give funds up to 45 days after month-end to file each monthly report and would keep public disclosure quarterly: only the third month of each fiscal quarter, released with a 60-day delay. The proposal was published in the Federal Register on February 23, 2026, and as of October 15, 2026 no final rule had been published. On the same day, in a separate final rule (Release IC-35963), the SEC extended the compliance dates for Names Rule items on Form N-PORT to the same November 2027 and May 2028 dates. Check the SEC's site for later action before relying on any of these dates.

The practical result is that an N-PORT figure can be several months old by the time it is readable. Pairing a daily ETF file with a quarter-old mutual fund filing mixes dates, which can shift rankings when holdings change in between. Record the as-of date of every file used.

An open-end fund's prospectus, filed on Form N-1A, and its statement of additional information describe the fund's investment strategy and policies, including any policy to concentrate in one industry, which the form defines as investing more than 25% of net assets in an industry or group of industries. Unit investment trusts, such as the S&P 500 fund in the example, do not use Form N-1A. These documents show the rules a fund follows; the holdings file shows what it owns on a given day.

What fund rules do and do not cover

Funds that want to be taxed as regulated investment companies (RICs) must meet a diversification test at the close of each quarter of the tax year. Under 26 U.S.C. 851(b)(3), no more than 25% of a fund's total assets may be invested in the securities of any one issuer, and at least 50% of its assets must meet a separate test that limits any one issuer to 5% of total assets and 10% of the issuer's voting securities. Government securities and the securities of other RICs are excluded from the 25% limit, and cash items count toward the 50% test. The 25% limit also applies to groups of issuers the fund controls in related businesses and to qualified publicly traded partnerships.

Two conditions make the test looser than it sounds. Under section 851(d), a fund that met the test at a quarter-end does not fail later just because market values move; only a discrepancy caused by a new purchase counts, and even that can be cured within 30 days after the quarter closes. So a stock can drift above 25% of a fund between purchases without breaking the rule.

The test applies to each fund on its own balance sheet. In the example, NVIDIA's 15.38% in the technology fund is below 25%, and the investor's 11.06% look-through weight was never measured by any test, because the rule does not add holdings across funds. Cross-fund concentration is the investor's to calculate. A sector fund also concentrates in one industry by design and discloses that policy, which is why it overlaps so heavily with broad funds that hold the same companies.

Frequently asked questions

Is overlap always a problem?

No. In the example, the broad fund and the S&P 500 fund overlap by 92.6%, and that is the intended result of how the broad index is built. Overlap becomes a concern when the combined exposure to a few companies is larger than the investor meant to hold, which the look-through table shows.

Do two share classes of one company count as one holding?

For concentration purposes, yes. Alphabet's Class A and Class C shares appear as separate lines in the holdings files, but both are ownership of the same company. In the S&P 500 fund file they are 3.07% and 2.46%, 5.52% combined. Merging them raises the look-through weight; ignoring the merge understates it.

How current are the holdings in an N-PORT filing?

Filings for a quarter are due within 60 days of the quarter's end, and the public version covers only the quarter-end month. A filing read near its release date can therefore describe holdings that are about two months old, and up to five months old just before the next one appears. The February 2026 SEC proposal would keep this quarterly public schedule; it had not been adopted as of October 15, 2026. Many ETF providers publish daily files, which are far more current but change each day.

Does the 25% limit stop a fund from being concentrated?

It limits the share of one issuer in a fund that relies on the tax rules for regulated investment companies, and it is tested at each quarter-end. It is not a cap on a stock's weight at all times: market movements alone do not cause a failure, it excludes government securities and other RICs from the 25% limit, and it ignores what an investor owns elsewhere.

How much did rounding matter in the example?

For the look-through figures, very little. Recomputing NVIDIA's look-through weight from the two-decimal weights shown gives 11.06%, the same as the unrounded 11.0608%. For pairwise overlap, rounding can matter a lot: with every weight rounded to a whole percentage, the S&P 500 fund and the broad fund would appear to overlap by 51% instead of 92.6%, because hundreds of holdings below 0.5% round to zero. Use the full-precision holdings file, not a rounded factsheet table.

A short overlap check to run

  1. List each fund and its share of the portfolio, summing to 100%. Use the Rebalancing Calculator if the shares have drifted from target.
  2. Download each fund's full holdings file and note the as-of date; use files from the same day where possible.
  3. Merge share classes of the same issuer, then match on issuer rather than on name or ticker.
  4. For each issuer, compute the weighted sum, rank the results and add up the top ten.
  5. Compute the overlap percentage for each pair, and compare sector weights as a second view.
  6. Test one plain scenario: multiply the top-ten total by a hypothetical decline and read the result as a share of the portfolio.
  7. Repeat when the holdings files change materially or a fund is added or removed.

Sources

Written by

AdvancedFinanceNet Editorial Team

Published Aug 23, 2026

Portfolio

How Tax Location Changes After-Tax Returns

The same two funds can end up $15,641 apart after 20 years, depending only on which account holds each one. A US worked example compares bonds in a traditional IRA with stocks in a taxable account, and the reverse. It states every assumption: 24% ordinary rate, 15% long-term rate, annual dividend tax, a sale at the end and IRA withdrawals taxed at 24%. The article then shows where the ranking can reverse, as with a Roth IRA or tax-exempt bonds, and how wash-sale and holding-period rules limit the shortcut. It closes with a checklist built around Form 1099-DIV.

Read the full guide
Portfolio

What Drawdown Means and Why It Matters

A 20% drop needs a 25% gain to recover, and in four S&P 500 drawdowns recovery outlasted the fall. Drawdown is the percentage decline from a portfolio's highest value to a later low, and it is measured again after every new high. This guide shows the formula, a hypothetical 24-month example worked month by month, and the recovery arithmetic behind the number. It then compares four S&P 500 drawdowns computed from daily closes, with the days each one took to fall and to recover. It also explains why max drawdown depends on the data frequency, on whether dividends are included and on the length of the record, and ends with a checklist for reading a drawdown figure on a fund page or dashboard.

Read the full guide
Portfolio

How Fees Compound Against a Portfolio

A 1% yearly fee leaves a lump sum about 26% smaller after 30 years, whatever the market returns. That figure comes from one line of arithmetic, and the same line explains why a fund expense ratio, an advisory fee and a platform charge multiply rather than add, why regular savers lose a smaller share than lump-sum investors, and why the extra return needed to justify a higher fee is larger than the fee gap itself. The worked tables use a stated 6% gross return and a 4% real return as assumptions, not forecasts, and end with the sustainable withdrawal a fee removes in retirement plus a short checklist for running the numbers on an actual account.

Read the full guide

Educational content only. Nothing on this site is personal financial, investment, tax or legal advice.