Marketstack Alternative for Fundamentals APIs
If you searched this, you probably already have Marketstack working for end-of-day or intraday prices and just hit a wall on the fundamentals side. That's a fair place to end up. Marketstack is genuinely good at what it's built for — broad ticker coverage, a simple REST API, a free tier to prototype against, and docs that don't require a sales call. None of that is in question here.
The issue is narrower and more specific: Marketstack's fundamentals endpoints return current, as-reported-today values. There's no first-filed date, no distinction between the original filing and later restatements, and no way to reconstruct "what the data actually said on March 15, 2019." For price/volume work that's irrelevant. For a fundamentals-driven backtest, it means every historical run is quietly using information the market didn't have yet — lookahead bias baked into the data layer before you've written a single line of strategy code.
Tradevo Data exists specifically to close that gap. It's not a general market-data platform — it's SEC-EDGAR-only, US equities only, annual filings only (10-K / 10-K/A, no quarterlies yet), covering 5,189 companies across 7 core concepts (Revenue, Net Income, Assets, Stockholders' Equity, Operating Cash Flow, Diluted EPS, Diluted Shares) with 312,751 point-in-time rows and up to 12 fiscal years of history. Every row carries a first_filed date, the original_value as first reported, the latest_value after any revisions, and a restated flag — 18,723 rows are flagged as restated across the dataset.
What Marketstack Is Genuinely Good At
- Wide asset coverage: equities, indices, forex, and more, across many exchanges globally.
- A free tier and low-friction paid tiers for people who just want prices in JSON.
- Mature, well-documented REST API most people can integrate in an afternoon.
- Good enough for dashboards, screeners, and anything where "current fundamentals" is actually what you want (e.g., a live valuation snapshot, not a historical backtest).
If your use case is "show me this company's latest P/E on a dashboard," Marketstack's current-value fundamentals are fine. This isn't an argument against Marketstack in general — it's an argument against using any current-value fundamentals feed, from any vendor, inside a historical backtest.
Where Current-Value Fundamentals Break Backtests
The mechanism is simple. A company files a 10-K for FY2019 in March 2020. Two years later it restates net income for FY2019 by 8% in an amendment. A current-value API — Marketstack's fundamentals included — will happily hand you that restated number if you query "FY2019 net income" today, with no indication a restatement happened or when the original figure became public. Backtest a strategy that trades earnings surprises using that number as of a 2020 rebalance date, and you're testing against information that didn't exist until 2022.
Our own measurement on the free sample (40 companies, methodology on GitHub) put the gap between filing date and the next possible query date at a mean of 43.4 days and a max of 61 days on that sample's reliable-filing rows — this figure describes the 40-company sample only, not the full dataset, and that's before counting actual restatements, which are a separate and often larger problem. We don't have a clean way to quantify how much of Marketstack's fundamentals data is post-restatement versus original, because that distinction isn't exposed in their API. That absence is itself the risk.
For the underlying mechanics, see /blog/lookahead-bias-fundamental-backtests and /blog/point-in-time-fundamentals-data.
Side-by-Side
| Marketstack (fundamentals) | Tradevo Data | |
|---|---|---|
| Data source | Aggregated/vendor-normalized | Direct from SEC EDGAR filings |
| Point-in-time fields | Not exposed | first_filed, original_value, latest_value, restated |
| Restatement tracking | None visible | 18,723 rows flagged |
| Frequency | Mixed / current-value | Annual only (10-K/10-K/A); quarterly is roadmap |
| Concepts | Broader statement line items, varies by plan | 7 core concepts, consistently tagged |
| Coverage | Global, many exchanges | US only, 5,189 companies |
| History | Varies | Up to 12 fiscal years |
| Bulk download | Depends on plan — see their pricing page | Included in $29/mo: full CSV + whole-universe snapshot |
| Format | JSON | JSON API; Parquet is roadmap |
| Free tier | Yes — see marketstack.com pricing page | Free sample: 40 companies, 3,280 rows, full methodology, no signup |
| Price | See their pricing page | $29/mo flat, 5,000 req/day |
See the lookahead leak on real filings — 40 companies, 3,280 rows, CC0, no signup. Three minutes in a notebook, before you decide anything.
We're not going to put a number next to Marketstack's plans here — pricing tiers change and we'd rather you check their pricing page directly than trust a stale figure on our site.
When Marketstack (or Building It Yourself) Is Actually the Better Choice
Be honest with yourself about the use case before switching anything:
- You need non-US markets, forex, or crypto. We're US-equity-only, SEC-EDGAR-only. Marketstack's breadth is a real advantage here and we can't match it.
- You want current fundamentals for a live dashboard, not a backtest. If lookahead bias isn't a concern — you're not simulating historical decisions — current-value data is simpler and Marketstack's broader coverage may serve you better.
- You need intraday or tick-level prices. That's not our product at all; Marketstack (or a dedicated market-data vendor) is the right tool.
- You need quarterly fundamentals today. We're annual-only right now; quarterly is on our roadmap but not shipped. If quarterly point-in-time is a hard requirement immediately, look at Sharadar, Tiingo, or QuantConnect's data offerings — all are credible options worth comparing on their own pricing pages.
- You have in-house engineering time and want full control. SEC EDGAR's bulk XBRL frames API is public domain and free. Parsing it yourself gives you complete control over tagging logic and restatement handling — at the cost of weeks of parsing, tag-normalization, and edge-case work (multi-tag concepts, amendment sequencing, share-count adjustments) that we've already done. If your team has the bandwidth and the requirement is narrow, DIY is a legitimate answer.
If none of those apply — you're backtesting US equity strategies on annual fundamentals and need to know what was actually knowable on a given date — that's the specific gap Tradevo Data fills.
Try Before You Subscribe
You don't need to take any of this on faith. The free sample covers 40 companies with 3,280 rows and the full point-in-time methodology, no signup required: github.com/christianpichichero-max/pit-fundamentals. Run your own lookahead check against it before paying anything.
If it holds up, the full dataset — 5,189 companies, 312,751 rows, bulk CSV download and whole-universe snapshots included — is $29/month with an instant API key after Stripe checkout, cancel anytime: https://tradevodata.com/. Docs are at /docs.
Not investment advice. Verify any competitor pricing yourself on their official pages before deciding — we intentionally don't quote it here.