S&P 500 additions and removals forecast
The companies most likely to join or leave the S&P 500 over the next 90 days, based on today's data.
Candidate lists are available on Plus and above
Unlock the companies most likely to join and leave this index, their probabilities, and the evidence behind each estimate.
Current index constituents remain public for everyone.
Upgrade to PlusHow this forecast works Methodology, limitations and common questions
Methodology
- Model version
- sp500-turnover-rank-90d-v2
- Turnover source
- S&P DJI, What Happened to the Index Effect?
- Exit-risk calibration
- Equibles, observed S&P 500 removal ranks (December 2019–June 2026)
- 92 resolved removals were ranked against the prior complete SPY snapshot: 26 in the bottom 5, 22 in positions 6–10, 12 in positions 11–20, 8 in positions 21–30, 16 in positions 31–200, and 8 in positions 201–500. Positions 31–200 are pooled to remove non-monotonic corporate-action noise; a symmetric 5% within-band tilt orders otherwise equal cohort weights without changing each cohort's total.
How the 90-day probabilities work
The join model starts with the S&P 500's observed average of about 21 additions a year from 2011 through June 2021. In S&P DJI's “What Happened to the Index Effect?” study, 125 of 226 observed additions came from the S&P 400 or S&P 600 and 101 came from outside the Composite 1500. The model reserves those historical shares, then distributes each group's 90-day event rate by the square root of float-adjusted market value. Square root weighting lets size matter without allowing one very large company to absorb the whole estimate.
The exit model keeps the same observed turnover rate, then ranks current members by total company value. Its risk curve is calibrated from 92 resolved removals observed between complete SPY snapshots from December 2019 through June 2026: 48 came from the bottom 10 positions by size and 68 from the bottom 30. The curve therefore concentrates the predictable part of exit risk among the smallest members while retaining a baseline for larger members and missing values. No entry-rule outcome or entry cutoff raises an exit estimate, because an entry rule is not an automatic deletion rule.
Company value, liquidity, earnings and public float come from settled market data and company SEC filings. Every percentage is stored with the daily snapshot, so a later result can be compared with what the model actually published that day.
What each eligibility outcome means
A rule reports one of four outcomes, and the difference between the last two is the whole point. Pass means the figure was read and it clears the bar. Fail means the figure was read and it does not. Not evaluated means a figure the rule needs could not be read, which is a fact about our data and not about the company, so no verdict is reached at all. Not applied means the rule exists in the methodology and we have no way to test it, so it is named on the page and left out of the verdict rather than quietly counted as passed.
How to read the ranking
A larger percentage means a higher estimated chance within the next 90 days. It does not mean the event is scheduled or certain. Join estimates are limited to companies that clear every entry rule today; exit estimates cover every current member, including members that still clear them all.
When the S&P 500 actually rebalances
There is no fixed reconstitution date. Share counts and float are refreshed quarterly, and additions and deletions happen as they are needed through the year, most often because a sitting member is acquired, taken private or split up, which opens the seat a new company is chosen for.
Scoring runs once per Eastern trading day, from settled closing prices and the filings on file that morning. Every model revision is immutable, so a new model may append a higher revision for that trading day without erasing what the earlier model said. Readers use the highest revision, while every published answer remains available for grading.
Limitations
- The S&P committee retains discretion, so a percentage is an estimate rather than a guarantee.
- Each estimate is conditional on today's eligible set and size inputs. New filings and market moves can change it before the 90-day horizon ends.
- Unannounced mergers, acquisitions, restructurings and take-privates are not knowable from the current screen and can change any member's exit probability abruptly.
- The historical turnover anchor covers 2011 through June 2021 and may not match the next 90 days.
- The removal-rank calibration contains 92 resolved observations from December 2019 through June 2026. A different period can produce a different size-risk curve.
Frequently asked questions
What does a S&P 500 probability mean?
It is the model's estimated chance that the named company joins or exits within the next 90 days. It is a forecast with a fixed horizon, not a claim that the committee has made or scheduled a decision.
How are S&P 500 join probabilities calculated?
The join model starts with the S&P 500's observed average of about 21 additions a year from 2011 through June 2021. In S&P DJI's “What Happened to the Index Effect?” study, 125 of 226 observed additions came from the S&P 400 or S&P 600 and 101 came from outside the Composite 1500. The model reserves those historical shares, then distributes each group's 90-day event rate by the square root of float-adjusted market value. Square root weighting lets size matter without allowing one very large company to absorb the whole estimate.
How are S&P 500 exit probabilities calculated?
The exit model keeps the same observed turnover rate, then ranks current members by total company value. Its risk curve is calibrated from 92 resolved removals observed between complete SPY snapshots from December 2019 through June 2026: 48 came from the bottom 10 positions by size and 68 from the bottom 30. The curve therefore concentrates the predictable part of exit risk among the smallest members while retaining a baseline for larger members and missing values. No entry-rule outcome or entry cutoff raises an exit estimate, because an entry rule is not an automatic deletion rule.
Why can a large S&P 500 member have a very low exit probability?
Entry-rule failures do not automatically remove a current member and do not raise this model's exit estimate. The historical exit-event rate is distributed by relative member size, so smaller members rank above larger members without treating the entry cutoff as a deletion rule.
When does the S&P 500 rebalance?
There is no fixed reconstitution date. Share counts and float are refreshed quarterly, and additions and deletions happen as they are needed through the year, most often because a sitting member is acquired, taken private or split up, which opens the seat a new company is chosen for.
What data is the S&P 500 probability forecast built from?
Company value, liquidity, earnings and public float come from settled market data and company SEC filings. Every percentage is stored with the daily snapshot, so a later result can be compared with what the model actually published that day.
These lists are fund holdings, not official index data. Each list is a full-replication tracking fund's own equity holdings, as the fund disclosed them on the date shown, taken either from its SEC Form N-PORT filing or from the holdings file it publishes on its own site. A fund tracks its index closely but is not identical to it, and weights are the fund's, not the index provider's.
Equibles is not affiliated with, endorsed by or sponsored by S&P Dow Jones Indices, FTSE Russell or Nasdaq. Index names are used only to identify which list is being described and remain the trademarks of their respective owners.