QQQ and XLK are both described as ways to own large American technology. They are not the same bet, and the difference is not the one most comparisons reach for. It is not the fee, and it is not the number of holdings. It is that XLK concentrates roughly twice as hard as QQQ, and that a portfolio holding both is less spread out than one holding only QQQ.
All of this comes from the funds' own quarterly filings rather than from a fact sheet.
Data as of 2026-06-30 (filing period) · received by the SEC 2026-08-28 · analysis written 2026-09-26. Both funds file again for the September quarter in late November, and these figures will be re-measured then.
What each fund holds
| QQQ | XLK | |
|---|---|---|
| Net assets | $490.1bn | $123.9bn |
| Stocks held | 103 | 77 |
| Behaves like | 34.1 equal holdings | 16.3 equal holdings |
| Ten largest | 45.0% of the fund | 64.5% of the fund |
| Holding count overstates spread by | 3.0× | 4.7× |
The "behaves like" figure is the effective number of holdings: one divided by the sum of the squared weights. A fund of 103 names where one of them is 7.6% does not behave like 103 holdings, and this says by how much. The method, and its limits.
So XLK holds three quarters as many stocks as QQQ but behaves like half as many. Its ten largest positions are 64.5% of the fund — roughly two dollars in three.
The 57% they share
The two funds hold 34 of the same stocks, and those account for 56.8% of the weight.
| Shared holding | Weight in QQQ | Weight in XLK |
|---|---|---|
| NVIDIA | 7.60% | 14.65% |
| Apple | 6.67% | 12.85% |
| Micron Technology | 5.64% | 5.42% |
| Microsoft | 4.35% | 8.38% |
| Advanced Micro Devices | 4.11% | 5.28% |
| Broadcom | 2.81% | 5.41% |
| Intel | 3.04% | 3.68% |
| Applied Materials | 2.49% | 3.20% |
| Lam Research | 2.35% | 3.02% |
| Cisco Systems | 2.01% | 2.59% |
The same companies, weighted roughly twice as heavily in XLK. NVIDIA and Apple alone are 27.5% of XLK against 14.3% of QQQ.
Held together, the two behave like 24.8 equal holdings — fewer than QQQ's 34.1 on its own. Adding XLK to QQQ does not spread the money out. It concentrates it further into the names both already hold.
What only one of them holds
QQQ's exclusive positions are 43.0% of it; XLK's are 15.3% of it. That asymmetry is the real distinction between the two funds.
In QQQ, not in XLK — because the Nasdaq-100 is an exchange listing, not a sector:
Amazon 4.03%, Tesla 3.30%, Alphabet 3.27% + 3.03% (two share classes), Meta Platforms 2.62%.
Four of the largest American companies, none of them classified as technology. QQQ owns them because they list on the Nasdaq. XLK cannot own them because the S&P classification puts them in consumer discretionary, communication services and elsewhere.
In XLK, not in QQQ — because they are technology companies that do not list on the Nasdaq:
IBM 1.47%, Oracle 1.39%, Amphenol 1.21%, Seagate 1.21%, Corning 1.12%.
The thing worth understanding
QQQ is an exchange's list. XLK is a sector definition. Neither is a technology fund in the plain sense, and each is wrong about it in the opposite direction: QQQ holds Amazon, Tesla, Alphabet and Meta while excluding IBM and Oracle; XLK does the reverse.
If you want the mega-cap platform companies, QQQ has them and XLK does not. If you want semiconductors and enterprise hardware weighted heavily, XLK does that — NVIDIA at 14.7% is a position size that will drive the fund on its own.
What neither fund does is give you diversified exposure to technology. QQQ behaves like 34 holdings and XLK like 16, out of a US market where VTI behaves like 67 across 3,539 stocks.
If you hold both, this is what you own
Half in each, looked through to the underlying companies rather than described fund by fund:
| Combined weight | Company | Held by |
|---|---|---|
| 11.1% | NVIDIA | both |
| 9.8% | Apple | both |
| 6.4% | Microsoft | both |
| 5.5% | Micron Technology | both |
| 4.7% | Advanced Micro Devices | both |
| 4.1% | Broadcom | both |
| 3.4% | Intel | both |
| 2.8% | Applied Materials | both |
| 2.7% | Lam Research | both |
| 2.3% | Cisco Systems | both |
Every one of the ten largest positions is in both funds. 70.9% of the money is in the 34 companies both hold — you are not diversifying across two funds, you are buying those 34 companies twice and everything else once.
Two names are a fifth of it: NVIDIA and Apple together are 20.9%. The pair behaves like 24.8 equal holdings, against 25.2 if you naively averaged the two funds' own figures — so the overlap costs you almost nothing extra beyond what the funds already were, which is the point. The concentration was there before you combined them.
Run your own weights through it — the calculator reads through every fund's filing the same way.
XLK was twice as concentrated two years ago, and something changed in 2024
The current figure is one filing. The archive has twenty-eight of them, and XLK's is not a smooth line:
| Filing period | Stocks | Behaves like | Top ten | Largest position |
|---|---|---|---|---|
| 2019-09-30 | 69 | 12.0 | 64.6% | 19.6% |
| 2021-09-30 | 76 | 9.7 | 67.5% | 22.0% |
| 2023-09-30 | 66 | 8.6 | 69.6% | 23.2% |
| 2024-03-31 | 67 | 9.4 | 67.2% | 24.1% |
| 2024-06-30 | 68 | 9.9 | 65.5% | 22.2% |
| 2024-09-30 | 72 | 14.8 | 61.9% | 15.0% |
| 2025-09-30 | 71 | 15.4 | 61.6% | 14.9% |
| 2026-06-30 | 77 | 16.3 | 64.5% | 14.7% |
Between two adjacent filings the largest position falls from 22.2% to 15.0% and the fund's effective holdings rise by half, from 9.9 to 14.8. The stock count moved by four. Nothing about the companies changed that quarter in a way that could do this; what changed is how much of the fund any one of them is allowed to be.
The filings record the change and do not explain it. N-PORT reports positions and weights, not index rules, so this site can show you the break and its date and will not tell you the rule behind it. What can be said from the data alone: before mid-2024 XLK behaved like roughly nine equal holdings, since then roughly fifteen, and that is a different fund from the one a 2023 comparison would have described.
It matters for the comparison above. XLK is less concentrated than it was — and still less spread out than QQQ by half.
Seven years, and the gap that concentration did not explain
Monthly total returns as each fund filed them in Item B.5, compounded over the whole series on record:
| Whole series | Months | Last 12 months | Best month | Worst month | |
|---|---|---|---|---|---|
| XLK | +417.3% | 84 | +51.2% | +19.9% | −12.0% |
| IVV (S&P 500, for scale) | +183.4% | 84 | +22.3% | +12.8% | −12.4% |
XLK returned more than twice the broad index over seven years, and its worst filed month is better than the S&P 500's. That is not what a fund behaving like nine to sixteen holdings is supposed to look like, and it is the honest counterweight to everything above: concentration is a description of a portfolio, not a prediction about it. The same nine-to-sixteen holdings that make XLK fragile in principle are what produced the return in this particular seven years.
QQQ's own filed series under the share class the SEC's table names runs only nine months, too short to compound here, so no figure is published for it rather than one stitched from a different source. How these are read, and the funds where our two readings disagree.
Neither fund lends much
A detail the fact sheets do not carry. From Item B.4.a, which names every firm borrowing a fund's stock:
- XLK: 0.21% of the fund on loan, to 3 named borrowers, 61.1% of it with Wells Fargo Bank.
- IVV, for comparison: 0.17% on loan across 21 borrowers.
Both are trivial next to a small-cap fund like IWM, which lends 16% of itself. Large, liquid megacaps are cheap to borrow elsewhere, so there is little demand to lend them — which is itself a fact about what these funds hold. The whole lending dataset.
What this does and does not tell you
- It is one date. Both figures are 30 June 2026 holdings. Index funds change with their indices, and XLK's concentration in particular moves with NVIDIA's price.
- Overlap is by weight, not by count. Two funds can share few names and much money, which is exactly what happens here.
- It is not advice. Nothing here says which fund to hold, or whether to hold either.
- The full position lists behind every figure: QQQ and XLK as filed, free under CC BY 4.0.