How to Monitor ETF Overlap Over Time
Monitor ETF overlap by saving a holdings-weight baseline, recalculating it only when comparable official holdings snapshots become available, and tracking changes in shared weight, underlying stock concentration, and sector exposure. Always report each fund's as-of date; without a comparable new filing, there is no confirmed overlap drift.
1. Establish a dated holdings-weight baseline
A useful baseline is not just one overlap percentage. It is a frozen, reproducible set of inputs and outputs. Save the portfolio allocation you want monitored, the official holdings snapshot used for every fund, and the calculation result. The baseline should record:
- Portfolio input: each ETF and its share of the submitted portfolio.
- Source identity: issuer or SEC source, source type, and source URL.
- Two different dates: when the source was checked and the holdings as-of date represented by the file.
- Comparable holdings: normalized positive-weight equity positions, their identifiers, and their reported weights.
- Baseline outputs: pairwise weighted overlap, shared holdings, combined underlying-stock weights, top-five concentration, sector weights, and Unclassified weight.
- Method version and limits: matching rules, excluded asset types, and sector-classification coverage.
Monitoring without a brokerage connection uses the last allocation you submitted. It can detect changes inside the funds' published holdings, but it cannot see your trades, contributions, withdrawals, cash, or rebalancing. Update the submitted allocation when your portfolio changes, then establish a new input baseline.
2. Monitor four related, but different, metrics
| Metric | Definition | What a change can show |
|---|---|---|
| Weighted overlap | For each normalized equity position held by both funds, take the smaller reported weight and add those contributions. | Whether two funds have gained or lost shared holdings weight. Report every monitored pair; do not invent one score for a portfolio of three or more funds. |
| Shared holdings | The count of normalized positive-weight equity positions present in both snapshots. | Which names entered or left the intersection. Count is context because a tiny position and a large position each count as one. |
| Underlying-stock concentration | Multiply each stock's fund weight by that fund's submitted portfolio share, add across funds, then inspect individual and top-five weights. | Whether repeated exposure is making one company, or the largest group of companies, a larger or smaller share of the submitted portfolio. |
| Sector exposure | Add combined underlying-stock weights by sector; keep unmapped positions in Unclassified. | Whether the look-through sector mix changed, provided classification coverage is comparable at both endpoints. |
Percentage-point change = new weight − baseline weight. A move from 8.10% to 4.89% is −3.21 percentage points, not “−3.21%.” A change driver is each security's change in contribution to the metric; for pair overlap, that contribution is the smaller of its two fund weights.
3. Check sources regularly; confirm drift only on new comparable data
A monthly holdings check is a practical cadence for a long-term ETF portfolio, but the underlying sources do not share one publication schedule. Some issuer files are offered daily; public SEC filings arrive on a delayed schedule. Checking a source every day can tell you whether a new file exists. It does not create a new holdings snapshot every day.
Source checked is the retrieval time. Holdings as of is the date represented by the issuer file or filing.
A report should compare new endpoints only after every fund required for the metric has a newer official snapshot that meets the comparison rule set at baseline.
Monthly checking is therefore a measurement schedule, not a promise that holdings changed or that a human-written report will arrive every calendar month. If nothing comparable was published, the useful result is a freshness status, not a recycled number described as new drift.
4. Real before/after example: SPYD and XLE
This example uses archived official daily holdings from the same issuer, State Street. At each endpoint, both funds carry the same as-of date, so the comparison is not mixing an old snapshot for one fund with a new snapshot for the other.
| Measure | July 24, 2026 | August 11, 2026 | Change |
|---|---|---|---|
| SPYD and XLE weighted overlap | 8.10% | 4.89% | −3.21 pp |
| Shared holdings | 6 | 4 | −2 |
| Endpoint date alignment | SPYD and XLE both July 24 | SPYD and XLE both August 11 | Comparable |
The main drivers were Phillips 66 (PSX) and APA. Both remained in XLE but left the later SPYD snapshot, removing 1.68 and 0.79 percentage points of overlap contribution. Together they explain 2.47 points of the 3.21-point decline; smaller contribution changes in EOG, ONEOK, Kinder Morgan, and Chevron explain the remainder. This identifies why the measured overlap changed. It does not say whether either fund or the combination is suitable.
Reproduction: match positive-weight equity positions by normalized identifier and calculate Σ min(SPYD weight, XLE weight). Source pages: State Street SPYD and State Street XLE. See the current SPYD vs XLE comparison separately; its displayed value may update after this article's fixed historical endpoints.
5. How mixed data dates create false drift
Suppose Fund A publishes an August 11 snapshot while Fund B's latest comparable snapshot is still July 24. Recalculating the pair combines Fund A's newer weights with Fund B's older weights. The number may move, but the movement contains a date mismatch: it cannot isolate a confirmed change between two comparable endpoints.
Show both as-of dates if a mixed-date current estimate is useful, label the result not comparable, and wait for the missing compatible snapshot before reporting confirmed overlap drift. Never convert “the source was checked” into “the holdings updated.”
Default change flags
A monthly holdings check uses defaults to identify material movements. When a change crosses those rules, human review explains the measured change and its drivers. OverlapCheck starts with the following documented house rules; different thresholds can be agreed in writing during setup.
| Metric | Default change flag |
|---|---|
| Pairwise weighted overlap | ≥ 3.0 percentage-point change, or a move across the under 30%, 30% to under 70%, or 70%+ house ranges |
| Underlying-stock concentration | ≥ 1.0 percentage-point change in one stock or the top-five total, or one stock crossing the 5.0% house level |
| Sector exposure | ≥ 2.0 percentage-point change, only when sector-classification coverage is comparable |
| Shared-holdings count | Shown for context; count alone does not trigger a change flag |
These are OverlapCheck house thresholds, not industry standards, personalized risk limits, or investment advice. Crossing one prompts human inspection of the dated holdings data; it is not a recommendation to trade.
Data freshness states a report should show
| Status | Meaning |
|---|---|
| Comparable update | Every fund required for the metric has a newer successful official snapshot, and its dates meet the comparison rule documented at setup. Confirmed drift can be calculated. |
| No new comparable data | At least one required fund has no newer comparable official snapshot. The source may have been checked, but no overlap drift is confirmed for the period. |
| Mixed-date / not comparable | The available snapshots do not represent compatible dates or reporting periods. A current estimate may be shown with the dates, but it is not labeled confirmed drift. |
| Fallback snapshot | A source check failed and a previously stored snapshot was reused. It is labeled stale or fallback, never presented as a new update. |
No new comparable holdings data was available. The source was checked, but no overlap drift is confirmed for this period.
6. Sample human review of a material change
This is the anonymous report format populated with the public SPYD/XLE example above, not a customer report. No customer identity, account, brokerage credential, or account number is used. The 50/50 allocation is illustrative and is held constant across both endpoints.
Material-change review
Example portfolio A
50% SPYD / 50% XLE · State Street official daily holdings
| Metric | Baseline | Current | Change | Result |
|---|---|---|---|---|
| SPYD and XLE weighted overlap | 8.10% | 4.89% | −3.21 pp | Flagged |
| Shared holdings | 6 | 4 | −2 | Context only |
| Largest stock: XOM | 10.39% | 10.37% | −0.03 pp | No change flag |
| Top-five stock concentration | 27.91% | 26.91% | −1.00 pp | Flagged |
| Energy sector exposure | 54.35% | 52.35% | −2.00 pp | Flagged |
What changed
Pairwise overlap crossed the default change threshold. PSX and APA leaving SPYD's shared set explain 2.47 of the 3.21 percentage points. In the fixed 50/50 look-through, top-five stock concentration fell 1.00 point and Energy exposure fell 2.00 points. XOM remained the largest underlying stock at 10.37%, but its own weight changed by only −0.03 point.
Interpretation boundary
These flags describe dated public holdings and the submitted allocation. They are not trade recommendations and do not measure performance, volatility, fees, taxes, or suitability.
7. Method, limitations, sources, and author
OverlapCheck calculates pairwise overlap from normalized positive-weight equity positions. Cash, derivatives, and other non-equity assets are not included in the pair metric. Portfolio look-through multiplies each fund position by the fund's submitted allocation; sector totals are then built from classified underlying positions. For the US-fund example above, the current sector map uses constituent membership across State Street's eleven Select Sector SPDR ETFs as the GICS classification reference; positions outside that map remain Unclassified. Identifier matching can leave economically related share classes or listings separate when a reliable match cannot be established.
Issuer schedules, delayed filings, source failures, mapping changes, and mixed as-of dates can postpone a comparable update. Sector drift is not confirmed when a classification-coverage change could explain the result. Monitoring describes published holdings; it does not observe brokerage activity, track performance or tax lots, or provide real-time alerts or investment advice. Read the full calculation methodology and known limitations and the ETF overlap definition.
- Official State Street SPYD holdings page
- Official State Street XLE holdings page
- Official SEC Form N-PORT instructions
- About OverlapCheck and editorial standards
Author and maintainer: OverlapCheck. We use an organization byline because no individual author profile is published. To report a source or arithmetic problem, send the affected URL to info@overlapcheck.com.
FAQ
How often should I check ETF overlap?
Check sources regularly, but recalculate drift only when comparable new holdings snapshots are available. OverlapCheck runs 12 monthly holdings checks during an annual service term; some issuers publish more often, while SEC N-PORT data arrives on a delayed filing schedule.
Does ETF overlap change every day?
Not necessarily. A source can be checked every day without publishing a new holdings snapshot. The source-check time and the holdings as-of date are different fields.
Can different as-of dates create false overlap drift?
Yes. If only one fund updates, a before-and-after difference may reflect the date mismatch rather than a confirmed change in the pair. Show both dates and wait for a comparable endpoint before calling the difference drift.
Does ETF overlap monitoring track my trades or contributions?
No. There is no brokerage connection. Reports use your last submitted fund allocation, so you need to send an update after trades, contributions, withdrawals, or rebalancing change that allocation.
Is high ETF overlap bad?
Not inherently. Overlap is descriptive and does not measure correlation, volatility, fees, factor or geographic exposure, suitability, or overall diversification.
What happens when no comparable update is available?
The monthly check records that no new comparable holdings data was available and does not claim confirmed overlap drift. Human review is added only when a material change is detected.