A Nasdaq Data Link (formerly Quandl) Alternative for Point-in-Time Fundamentals

Nasdaq Data Link is not a dataset. It is a marketplace — the former Quandl, now Nasdaq-owned — where one account and one Tables API (/api/v3/datatables/{VENDOR}/{TABLE}) front per-product entitlements you buy separately: Sharadar, Zacks, WorldVest Base, Nasdaq's first-party Equities 360, among others. So "is Nasdaq Data Link point-in-time?" has no single answer; it depends on which product you put in the cart.

It is alive and maintained: on 6 August 2026 the public SHARADAR/SF1 metadata endpoint reported a same-morning refresh, status ON TIME.

We build Tradevo Data: point-in-time US annual fundamentals from SEC EDGAR, one JSON endpoint, $49/mo. We are a far smaller thing than that shelf, and this page is written so you can tell when to buy them instead.

The distinction that actually matters

A filing-date field is not point-in-time data.

Many fundamentals products carry a filing-date column; that alone does not make them backtest-safe. If the schema stores one mutable value per fiscal period next to that date, the date describes the original filing while the number beside it was revised later. Query it "as of 2016" and you get a 2016 date attached to a 2026 number — look-ahead bias wearing a timestamp.

Concretely: Procter & Gamble's FY2015 revenue was filed at $76.279B and reads $70.749B today. Point-in-time means the first-reported value survives as its own row.

How each product on the shelf handles it

Sharadar Core US Fundamentals (SF1) documents a genuine point-in-time design: the As Reported dimensions are described as "excludes restatements / point-in-time view with data time-indexed to the date the form 10 regulatory filing was submitted to the SEC," versus Most-Recent Reported, which includes restatements and is indexed to the report period. Its public metadata endpoint keys rows on ticker, dimension, datekey, reportperiod — filing date as well as period — with datekey and lastupdated as filters. Whether every successive vintage of a period is retrievable, we could not confirm without a key; ask them if that matters to you. Two caveats they state themselves: the stamp is the 10-K/10-Q filing date, not the earnings release, which may have preceded it via 8-K; and ARQ is unavailable for foreign companies. History starts Dec 1997, though their page gives two different year counts, and coverage is quoted as 16,000+ companies (6,000 active, 10,000 delisted) in one section and 14,000+ in another.

Nasdaq Equities 360 (first-party) documents most-recent-reported dimensions (MRQ/MRY/MRT). We checked the metadata for its fundamentals tables (FD, FS, BS, IS, CF): each keys on calendardate, symbol, dimension, reportperiod, and offers calendardate, dimension, figi and symbol as its filters. No filing-date or vintage key appears there. Ask Nasdaq directly if you need PIT from E360. Its summary gives history from Jan 1998 while each fundamentals table is documented from 2008.

Zacks Fundamentals Collection A attaches filing_date, "original date of statement filing with the SEC." But the shared ZACKS/FC and ZACKS/FR tables key on m_ticker, per_end_date, per_type — no vintage dimension — which reads to us as one current row per period, not stored as-reported history. The one restatement signal we found is last_changed_date, "Date that the data was last restated"; we found no stated policy on whether prior values are retained. That is inference from the schema, not confirmation. Depth is the headline: Collection A documents history from Jan 1979 and 19,500+ US and Canadian names including 9,000–10,000+ delisted (its page gives both figures), with daily delta files. Other collections list their own start dates — check the one you would buy.

WorldVest Base (WVB) publishes WVB Global Fundamentals Data. A WVB FAQ says the WVB Dossier data product "does not include the original as-reported statements or restatements." That sentence names Dossier, so confirm it for the fundamentals product. Global breadth is the pitch there, not PIT.

Pricing: we quote none. It is login- or sales-gated — read it on each product's own page on data.nasdaq.com.

Side by side

When Nasdaq Data Link is the better choice

Buy from the Nasdaq Data Link shelf — most likely Sharadar SF1 — if any of this is you:

• You need delisted companies. This is the big one, and it is the reason most serious backtesters should stop reading here. Our universe is active-company focused, which means long-horizon backtests built on us carry survivorship bias. Say it out loud before you model anything: the losers are missing. SF1 and Zacks Collection A both quote substantial delisted coverage; we carry none.
• You need history deeper than 12 fiscal years. Ours thins out before FY2014 and does not reach the 2008 crisis, let alone the dot-com cycle. Dec 1997 on SF1 and Jan 1979 on Zacks Collection A do.
• You need quarterly or TTM. We are annual-only, and there is no roadmap promise here — today we do not have it. If your signal updates on 10-Qs, we are simply the wrong shape.
• You need breadth. Many more fields per product, plus precomputed ratios, is weeks of derivation work you skip versus our 7 raw concepts.
• You need production ergonomics at scale. Full-table export, a lastupdated filter for incremental pulls, daily delta files on Zacks Collection A, per-table status metadata and documented refresh windows, client libraries you did not have to write.
• You need one vendor-neutral integration. Swapping or blending fundamentals sources without re-plumbing is a real architectural benefit, and the rest of the marketplace's datasets sit under the same roof and the same key.
• You need a procurement path. A Nasdaq contract clears a compliance review in a way a solo developer's API does not, and that is a legitimate reason to spend more.

Sharadar's documentation also deserves credit most vendors do not earn: a worked restatement example, an explicit warning that ART is not the sum of four ARQ rows, an exceptions section for companies that do not present a classified balance sheet. That candour is worth paying for, and it is why we point at them rather than around them.

Where we fit, and where we are weak

We are the narrow, auditable, self-serve slice: 313,137 point-in-time rows across 5,195 US companies, 18,712 restatements labelled, every row carrying first_filed, original_value, latest_value, a restated flag and a per-row qa_status. One endpoint, $49/mo, 5,000 requests/day; you pass as_of and we do the join server-side. Full-universe filing lag on rows with a verifiable date: mean 66 days, median 60.

Our limits, before you find them yourself: annual only, US only, 12 years, 7 concepts, active-company focus (so: survivorship bias), no prices or OHLCV, no quarterly, no non-US, no ratios, no estimates, no real-time, no news, no options, no delisted history.

Filing-date gating is not a cure-all for anyone, us included. Amazon's FY2020 diluted shares were filed at 510,000,000 and read 10,198,000,000 today after the 2022 20:1 split was applied retroactively; date gating does not catch that. We also shipped a General Mills tagging error ourselves — FY2024 Revenues reads $2.038B against $19.857B on the contract-revenue tag — and now verify every load against SEC's API.

Try it without paying

The free CC0 sample — 40 large caps, 3,280 rows, no signup — is at github.com/christianpichichero-max/pit-fundamentals. Diff it against whatever you use now before trusting a row.

Not investment advice. Competitor coverage, pricing and terms change; verify current terms on Nasdaq Data Link's own product pages before deciding.