A Financial Modeling Prep Point-in-Time Alternative, Honestly Compared

If you searched for a "Financial Modeling Prep point in time alternative," you probably already like FMP. It's a broad, affordable API — prices, statements, ratios, ratios-TTM, analyst estimates, insider trades, and a lot more, all behind one key. For most dashboards and screeners, that breadth is exactly right.

But you're here because you hit a specific wall: backtesting. When you replay a strategy over history, you need each historical date to see only what was knowable on that date — the numbers as they were first filed, not today's revised figures. That single requirement, point-in-time (PIT) correctness, is where a broad "everything" API and a narrow "get this one thing exactly right" API part ways.

We build the narrow one. Tradevo Data does one job: point-in-time US equity fundamentals from SEC EDGAR, structured so a backtest can't peek at the future. This page is a fair comparison — including a whole section on when FMP is the better call. Radical honesty is our brand, so we'd rather you buy the right tool than the wrong one from us.

The one question that matters: is it lookahead-safe by default?

FMP absolutely gives you fundamentals, and it exposes real filing metadata. Its "As Reported" financial statements return the original, unadjusted figures straight from the filing, and responses carry filing-date fields (e.g., fillingDate) — that's genuinely the raw material for point-in-time work, and it's more than many cheap APIs offer.

The gap is what happens by default. FMP's mainline financial-statement endpoints return standardized, current values — figures that get revised as companies restate and as the provider updates its normalization. To avoid look-ahead bias with that, you have to reconstruct PIT yourself: pull as-reported data, key everything off filing dates, and diff original versus revised figures by hand. FMP gives you pieces to do that; it does not, as far as we can verify, ship a server-side as_of query or an explicit restatement flag / original-vs-latest pairing that does the reconstruction for you. (Verify this against FMP's current docs and pricing — vendors change endpoints often, and we'd rather you confirm than trust us.)

That's the whole distinction. FMP is broad and gives you the ingredients. Tradevo Data is narrow and hands you the finished, PIT-safe primitive.

What Tradevo Data actually is

One JSON endpoint, built around a single idea — what was public on this date?

• Server-side as_of: GET /v1/fundamentals?ticker=AAPL&as_of=2019-05-01 returns only rows where first_filed <= as_of (same-day inclusive). The lookahead guard lives on our server, not in your notebook.
• Every row carries its provenance: first_filed (the date the value became public), original_value (the first-reported, PIT-safe number), latest_value (the current revision), and a restated flag (set when a same-tag revision, including 10-K/A amendments, moves the number by more than 0.5%). 18,539 restatements are labeled, not silently overwritten.
• QA is visible, not hidden: each row has a qa_status. Suspect rows are flagged, never dropped — you decide what to trust.
• Coverage: 5,214 US companies, 313,562 point-in-time rows, up to 12 fiscal years, across 7 core concepts — Revenue, NetIncome, Assets, StockholdersEquity, OperatingCashFlow, EPSDiluted, DilutedShares. It's annual only today (10-K + 10-K/A), sourced from SEC EDGAR (public domain).
• Sane operations: 5,000 requests/day, key issued instantly after Stripe checkout, cancel anytime. $49/mo. Docs at /docs.

On honesty about lookahead: on a 40-company sample, measured across reliable-filing rows, the mean lag from fiscal period-end to first_filed was 43.4 days (max 61). That's a property of that sample, not a guarantee across all 5,214 tickers — we scope it deliberately rather than dress it up as a dataset-wide SLA.

Side-by-side

We're not listing FMP's tier prices as hard numbers because they move; check their page directly.

When Financial Modeling Prep is the better choice

This is the part we won't soften, because it's usually true:

• You need more than fundamentals. Prices, intraday bars, ratios, analyst estimates, earnings transcripts, insider trades, ETF holdings — FMP covers a huge surface for one subscription. We cover seven fundamental concepts. If you want one key for many jobs, that's FMP, not us.
• You need quarterly data. FMP has quarterly statements today. We're annual-only right now (quarterly is on our roadmap, not shipped). For quarterly rebalancing, we're simply not ready yet.
• You need non-US coverage. We're US-only. FMP reaches international markets.
• You're not backtesting. For a live screener, a research dashboard, or fundamentals you'll eyeball, "current standardized value" is fine — and FMP's breadth wins.
• You want bulk downloads. We don't ship Parquet bulk exports yet; it's one JSON endpoint. Heavy bulk workflows may fit FMP or a warehouse-style vendor better.

We're also not claiming to be the only cheap PIT option. Credible alternatives like Sharadar and Tiingo exist, and they're worth your evaluation too. Our pitch is narrower: an honest, budget tier of research-grade PIT correctness — the kind that used to start at $10k–$50k/yr — for backtesters who want the lookahead guard built in.

Try it before you trust it

We think verification beats marketing, so start free — no signup, no card:

• Free sample (40 large caps, 3,212 rows, full methodology): github.com/christianpichichero-max/pit-fundamentals — inspect the original_value / latest_value / restated columns yourself and decide if the PIT logic holds up.
• Full API, $49/mo, cancel anytime: tradevodata.com — key issued instantly after checkout, docs at /docs.

If FMP's breadth is what your project needs, use FMP with a clear conscience. If your backtest keeps quietly cheating because today's numbers leak into yesterday's decisions, that's the exact problem we built for.

Not investment advice. All figures describe data products, not expected returns. Competitor pricing and features change — please verify current FMP pricing and capabilities yourself before purchasing.