Money Market Fund Reports (N-MFP)

Form N-MFP money-market fund reports, one row per fund per month: net assets, shares outstanding, stable share price, seven-day yield, weighted-average maturity and life, and the Rule 2a-7 daily and weekly liquid-asset ratios that decide whether a fund can meet redemptions. Coverage begins 2016-05 with N-MFP1/2/3; the earlier N-MFP generation is parsed and schema-complete but its presence in the daily dissemination feed is unresolved.

edgar_mmf_filings - the dataset name to pass to the Obscura API.

What one row means

One **money-market fund series' monthly portfolio report** - the cover of an SEC Form N-MFP filing: which series, for which month, its net assets, share count, stable share price, seven-day yield, weighted-average maturity and life, and the daily/weekly liquid-asset ratios that Rule 2a-7 actually constrains. This is the fund-level companion to `edgar_mmf_holdings`. Both are parsed from the SAME submission object, and the split is the same one N-PORT uses: one cover row, N position rows. # FOUR GENERATIONS, ONE TABLE - and the generation is read from the DOCUMENT SEC has rewritten this form's XML three times, and **the vocabularies do not merely add fields, they RENAME the same concept**: | | `N-MFP` | `N-MFP1` | `N-MFP2` | `N-MFP3` | |---|---|---|---|---| | first seen | 2010-11 | 2016-05 | 2016-10 | 2024-07 | | root | bare `edgarSubmission`, no `formData` | `formData` | `formData` | `formData` | | positions | `scheduleOfPortfolioSecuritiesList/scheduleOfPortfolioSecurities` | `scheduleOfPortfolioSecuritiesInfo` | same | same | | net assets | `AssetsNet` | `netAssetOfSeries` | same | same | | WAM | `dollarWeightedAveragePortfolioMaturity` | `averagePortfolioMaturity` | same | same | | liquidity | *not reported* | 5 × `fridayWeek{1..5}` | 4 × `fridayDay` + 5 × `fridayWeek` | 21 × `liquidAssetsDetails` | | category | `InvestmentTypeDomain` | `moneyMarketFundCategory` | same | same | **The generation CANNOT be derived from a date.** Measured on SEC's own quarterly indexes, the generations OVERLAP inside a single quarter at both transitions: 2016 QTR2 carries 546 `N-MFP` and 1,039 `N-MFP1`; 2016 QTR4 carries 484 `N-MFP1` and 935 `N-MFP2`; 2024 QTR3 carries 53 `N-MFP2/A` and 965 `N-MFP3`. Amendments make it strictly worse - an `N-MFP2/A` restating a 2023 period was disseminated in 2024 QTR4, four months after `N-MFP3` began. So `edgar_mmf::Generation` is decided by the XML's own shape, and the form type is only cross-checked against it. Dating the era is the defect that left `edgar_13f_holdings` empty before 2013. # `available_date` is NOT the filing date before 2016-05 Original `N-MFP` filings were **embargoed 60 days** under the original Rule 30b1-7, and the embargo is directly observable: an original `N-MFP` appears in NO daily index on its filing date. Probed on seven dissemination days spanning 2011-07-12 → 2016-04-06, `N-MFP` originals number **zero** in every one, while the quarterly full index attributes thousands of them to those same dates. From 2016-05-06 onward `N-MFP1` originals appear same-day (457 rows), and `N-MFP2`/`N-MFP3` continue same-day through 2025. So the pre-2016-05 corpus is a genuine look-ahead trap: its `FILED AS OF DATE` is up to 60 days BEFORE anyone outside SEC could read it. See `docs/availability/edgar_mmf_filings.md`. See `docs/datasets/edgar_mmf_filings.md`.

One row per One SERIES in one REPORT MONTH - one money-market fund series' Form N-MFP cover, keyed by the EDGAR archive path (filename). One submission covers exactly one series (generalInfo/seriesId), so the path is the identity; a registrant running six money funds files six submissions per month. NOT one registrant per month, and NOT one share class: share classes are a nested list inside the submission and are summarised here into total_share_classes rather than exploded, because the class grain carries no portfolio and belongs to a different question..

Point-in-time availability

Every Obscura dataset carries available_date: the calendar day the publisher made the row available, day-of, with no session rounding. It is the one column a backtest filters on, and it means the same thing on every dataset in the catalog.

For edgar_mmf_filings: filing_join: available_date is the EDGAR DISSEMINATION day, taken from the bronze object's own day partition (edgar_docs::available_date), never from the submission's FILED AS OF DATE. WHY: original N-MFP filings were embargoed 60 days under the original Rule 30b1-7 and are absent from every daily index probed between 2011-07-12 and 2016-04-06 (seven days, zero N-MFP originals in each) while the quarterly full index attributes thousands of them to those dates; the embargo lifted with N-MFP1, whose originals appear same-day from 2016-05-06 (457 rows). occurred_on is generalInfo/reportDate - the month end the portfolio is AS OF - and is systematically 3-10 days BEFORE availability in the post-2016 era and up to 60 days before it in the pre-2016 era. Evidence class B (our own 404/absence-vs-presence observations over SEC's daily indexes); see docs/availability/edgar_mmf_filings.md.

Refresh cadence

Obscura refreshes edgar_mmf_filings every 6 hours - the most frequent scheduled job that re-collects or re-exports it. This is Obscura's own pipeline cadence, not the upstream publisher's release schedule; when a row became public is recorded per row in available_date.

Schema - 43 columns

The full public column list for edgar_mmf_filings, with the meaning of every field. The same schema is served unauthenticated at https://api.obscura.trade/v1/catalog/edgar_mmf_filings.

Access edgar_mmf_filings

Two delivery paths, one identifier. Both require an Obscura account and an active subscription; the catalog entry and the schema above are public.

import obscura

client = obscura.Client("obs_live_…")

df = client.query(
    dataset="edgar_mmf_filings",
    symbols=["NVDA", "AAPL"],
    start="2024-01-01",
)

Create a free account Browse all 122 datasets

Frequently asked questions

Related datasets