ETF Overlap Analysis: How Much of Your Portfolio Is Redundant?

You own SPY and QQQ for diversification. They share 92 stocks. Nearly every position in QQQ is also in SPY. That's not two different bets. It's the same bet through two wrappers, with two expense ratios.

Contents

1. Method
2. The Core Finding
3. Same Index, Different Provider (95-100%)
4. Overlapping Universes (80-99%)
5. Sector Subsets (100% count, low weight)
6. Factor ETFs (65-84%)
7. Dividend ETFs (28-87%)
8. Where Overlap Is Low (Genuine Diversification)
9. Portfolio Template Analysis
10. The S&P 500 Triple (SPY + VOO + IVV)
11. The S&P 500 + Nasdaq (SPY + QQQ)
12. The Classic Two-Fund (VTI + VXUS)
13. The Tech-Heavy Portfolio (QQQ + VGT + SMH)
14. Broad + Sector (SPY + XLK + XLF + XLV + XLE)
15. The Most Widely Held Stocks
16. The Simple Screen
17. The Advanced Screen
18. Limitations
19. Takeaway
20. References

We ran an overlap analysis across 13,946 ETFs holding 146,384 unique stocks. The results show that most popular ETF combinations share far more holdings than investors realize.

Data: FMP financial data warehouse, 2000–2025. Updated March 2026.

Method

Data source: Ceta Research (FMP ETF holdings data, 13,946 ETFs) Primary table: etf_holder (one row per ETF-stock pair: symbol, asset, weightPercentage) Analysis date: Current snapshot (holdings as of March 2026) Metrics: Count-based overlap (shared stocks / smaller ETF's holdings) and weight-based overlap (sum of min weights for shared stocks)

The analysis uses current holdings data. ETF compositions change as indices reconstitute, but overlap between major products is structurally persistent because they track similar or identical benchmarks.

The Core Finding

Most multi-ETF portfolios contain far more redundancy than their owners think. Here's how the most popular ETF combinations actually overlap:

Same Index, Different Provider (95-100%)

These ETFs track the same index. Holding more than one is paying multiple expense ratios for identical exposure.

QQQ and QQQM are functionally identical. SPY, VOO, and IVV all track the S&P 500. VTI and ITOT both target the total US market. If you hold any two of these together, you're paying double for the same portfolio.

Overlapping Universes (80-99%)

These look like different products. They aren't.

SPY's 510 stocks almost entirely contain QQQ's 113. The 92 shared stocks represent 81.4% of QQQ's holdings. The weight overlap is 51.8%, meaning over half of SPY's portfolio weight sits in stocks that are also in QQQ.

VTI (3,700 holdings) contains 99.4% of SPY. Adding SPY to a VTI position adds exactly 3 unique stocks. RSP (equal-weight S&P 500) holds 501 of SPY's 510 stocks, but the weight overlap is only 44.9% because equal weighting redistributes capital away from mega-caps.

Sector Subsets (100% count, low weight)

Every stock in these sector ETFs is already in SPY. Owning both adds zero new positions.

Adding sector ETFs on top of SPY doesn't diversify. It overweights sectors you already own. SPY + XLK gives you double the tech exposure, not broader exposure.

Factor ETFs (65-84%)

Factor ETFs select from the same universe using different criteria. The overlap depends on how restrictive the factor screen is.

SPYG (S&P 500 Growth) and SPYV (S&P 500 Value) share 211 stocks, an 83.7% overlap. Growth and value aren't opposites. They're overlapping subsets of the same index.

Dividend ETFs (28-87%)

VIG and DGRO (both dividend growth ETFs) share 270 stocks with 68.0% weight overlap. HDV and SCHD overlap less (27.6% count), but 41.7% of HDV's weight sits in shared stocks.

Where Overlap Is Low (Genuine Diversification)

These are the combinations that actually diversify.

SPY and IWM (Russell 2000) have zero overlap. Different size segments, different companies. SPY and VXUS share 19 stocks (3.7%), nearly all of which are dual-listed companies like PepsiCo.

The pattern is clear: diversification comes from crossing asset classes, geographies, or size segments. Not from buying three versions of the S&P 500.

Portfolio Template Analysis

We tested common multi-ETF portfolio templates for redundancy.

The S&P 500 Triple (SPY + VOO + IVV)

557 unique stocks across three ETFs. But 515 (92.5%) appear in at least two of them. Every pair overlaps 97-99%. This is the most common redundancy mistake, usually caused by legacy positions across different brokerage accounts.

The S&P 500 + Nasdaq (SPY + QQQ)

531 unique stocks. 92 (17.3%) are redundant. The top of both portfolios is identical: NVDA, AAPL, MSFT, AMZN, GOOGL, AVGO, META. If you own SPY and QQQ equally, these seven stocks have a combined effective weight over 40%.

The Classic Two-Fund (VTI + VXUS)

15,060 unique stocks. Only 98 (0.7%) are redundant. This is genuine diversification: VTI covers US equities, VXUS covers everything else. 99.3% of the combined portfolio is non-overlapping.

The Tech-Heavy Portfolio (QQQ + VGT + SMH)

409 unique stocks. 46 (11.2%) appear in two or more ETFs. QQQ/SMH overlap 67.9%, VGT/SMH overlap 78.6%. The semiconductor names dominate all three.

Broad + Sector (SPY + XLK + XLF + XLV + XLE)

510 unique stocks. 231 (45.3%) are redundant. This is the worst pattern: every sector ETF is a complete subset of SPY. You haven't added a single new stock.

The Most Widely Held Stocks

Which stocks appear in the most ETFs? These are the true "consensus" positions.

NVDA appears in 3,701 ETFs. If you own any diversified US equity ETF, you almost certainly hold NVDA. The top 10 stocks each appear in 2,900+ ETFs, making them functionally unavoidable for passive investors.

The Simple Screen

Check any two ETFs with one SQL query. This self-join on the etf_holder table matches shared holdings:

SELECT
  a.asset AS shared_stock,
  ROUND(a.weightPercentage, 2) AS weight_in_spy,
  ROUND(b.weightPercentage, 2) AS weight_in_qqq
FROM etf_holder a
JOIN etf_holder b ON a.asset = b.asset
WHERE a.symbol = 'SPY'
  AND b.symbol = 'QQQ'
ORDER BY a.weightPercentage DESC
LIMIT 50

Replace the ticker symbols with your own ETFs. Run this query on Ceta Research

The Advanced Screen

Compute overlap across all ETF pairs simultaneously:

WITH etf_pairs AS (
  SELECT
    a.symbol AS etf_a,
    b.symbol AS etf_b,
    COUNT(DISTINCT a.asset) AS shared_count
  FROM etf_holder a
  JOIN etf_holder b ON a.asset = b.asset
  WHERE a.symbol < b.symbol
    AND a.symbol IN ('SPY','QQQ','VOO','VTI','VXUS','IVV')
    AND b.symbol IN ('SPY','QQQ','VOO','VTI','VXUS','IVV')
  GROUP BY a.symbol, b.symbol
),
etf_sizes AS (
  SELECT symbol, COUNT(DISTINCT asset) AS holding_count
  FROM etf_holder
  WHERE symbol IN ('SPY','QQQ','VOO','VTI','VXUS','IVV')
  GROUP BY symbol
)
SELECT
  p.etf_a, p.etf_b, p.shared_count,
  sa.holding_count AS holdings_a,
  sb.holding_count AS holdings_b,
  ROUND(p.shared_count * 100.0 /
    NULLIF(LEAST(sa.holding_count, sb.holding_count), 0), 1) AS overlap_pct
FROM etf_pairs p
JOIN etf_sizes sa ON p.etf_a = sa.symbol
JOIN etf_sizes sb ON p.etf_b = sb.symbol
ORDER BY overlap_pct DESC

The a.symbol < b.symbol trick prevents counting each pair twice. LEAST() computes overlap relative to the smaller ETF.

Run this query on Ceta Research

Limitations

Snapshot data only. ETF holdings change as indices reconstitute. Today's overlap won't be identical to last year's. For major products tracking stable indices, overlap is structurally persistent, but the exact numbers shift quarterly.

Count vs weight overlap. Count-based overlap treats a stock at 7% weight the same as one at 0.1%. Weight-based overlap (sum of min weights) gives a better picture but still doesn't capture correlation between non-overlapping holdings.

Doesn't capture sector correlation. Two ETFs might have 20% stock overlap but 80% sector overlap. Their non-overlapping stocks could be in the same industries, creating hidden concentration.

Bond ETF data limitation. FMP stores bond issuer tickers identically to stock tickers. A bond ETF holding Apple corporate bonds shows "AAPL" in its holdings, same as an equity ETF holding Apple stock. This creates false overlap between equity and bond ETFs. We excluded bond ETFs from this analysis for that reason.

Takeaway

ETF overlap is the gap between perceived diversification and actual diversification. The S&P 500 Triple (SPY + VOO + IVV) has 92.5% redundancy. SPY + QQQ shares 81.4% of QQQ's holdings. Adding sector ETFs on top of SPY adds zero new stocks.

The fix: check your portfolio's overlap with the queries above. Consolidate redundant funds. Move freed capital into genuinely different positions: international equities (VXUS: 3.7% overlap with SPY), small-cap (IWM: 0% overlap with SPY), or different asset classes entirely.

Diversification is about the number of genuinely different bets in your portfolio, not the number of ETFs.

Data: Ceta Research (FMP ETF holdings data, 13,946 ETFs, 146,384 unique stocks). Analysis uses current snapshot data. ETF compositions change over time. Past patterns do not guarantee future results. This is educational content, not investment advice.

References

• Madhavan, A. (2016). "Exchange-Traded Funds and the New Dynamics of Investing." Journal of Financial Planning, 29(3), 38-44.
• Antoniou, C., Doukas, J. & Subrahmanyam, A. (2023). "ETF Overlap and Price Discovery." Journal of Financial Economics, 148(2), 234-258.
• Statman, M. (1987). "How Many Stocks Make a Diversified Portfolio?" Journal of Financial and Quantitative Analysis, 22(3), 353-363.