Where does Insider Option get its data?

Directly from primary sources: STOCK Act Periodic Transaction Reports on the US House and Senate disclosure portals, SEC Forms 3, 4 and 5 and Form 13F from EDGAR, and prices from a third-party market data provider. We do not resell another aggregator's parsed output.

Does Insider Option measure returns from the trade date or the disclosure date?

From the trade date — the price at the disclosed transaction to the latest price we hold. That describes the filer's own result, not what a follower could have captured, because disclosure can lag the trade by up to 45 days. A disclosure-date figure would usually be lower; we do not publish one yet because we do not store per-ticker prices as at each filing date.

Why are Congressional trade amounts shown as ranges?

Because the STOCK Act only requires disclosure in bands, such as $15,001 to $50,000. Where a calculation needs one number we use the band midpoint and label the result an estimate; exact position sizes are not public information.

What is the biggest source of error in Congressional trading data?

Mapping issuer names to tickers. Filings name companies as free text with no controlled vocabulary, and House filings are often scanned handwriting, so ticker resolution and OCR are the dominant error sources for any provider of this data.

Does Insider Option claim that following these trades beats the market?

No. The published evidence for disclosure-based strategies beating the market after 2012 is weak and contested, largely because of the 45-day reporting lag. We treat the data as a research and accountability tool rather than a guaranteed edge.

Methodology

Where our data comes from, how we compute performance, and what these numbers cannot tell you.

Every figure on this site is derived from public US regulatory filings. This page documents exactly how — including the conventions that change the answer, and the limitations that no amount of engineering removes.

If you are going to cite a number from Insider Option, read the two sections on entry dating and benchmarks first. They are where most published claims about disclosure-based investing go wrong, including claims made by our competitors.

Data sources

We use primary sources only. We do not resell another aggregator's parsed output, which means our errors are ours and we can correct them.

Congressional trades: Periodic Transaction Reports from the US House Clerk's Financial Disclosure portal and the US Senate's Electronic Financial Disclosure system, required by the STOCK Act of 2012.
Corporate insiders: Forms 3, 4 and 5 filed under Section 16 of the Securities Exchange Act, from the SEC's EDGAR system. Form 4 is due within two business days of a transaction.
Institutional holdings: Form 13F, filed quarterly by managers with at least $100 million in qualifying US equity assets, due 45 days after quarter end.
Prices and volume: a third-party market data provider, subject to that provider's own latency and licensing terms.

Entry dating: our figures run from the transaction date, and what that means

This is the single most consequential detail in this field, so read it before quoting any number from us.

Every performance figure on this site is measured from the price at the disclosed transaction to the most recent price we hold. That means it describes what the filer's position did — not what a member of the public following them could have captured.

The distinction matters because of the reporting lag. A member of Congress can trade on 1 March and legally not disclose it until 15 April. By the time you could act, up to 45 days of price movement has already happened. A figure measured from the disclosure date would therefore usually be lower than the one we show, sometimes substantially.

We do not currently publish a disclosure-date figure because we do not store per-ticker prices as at each filing date; computing one honestly requires that history, and we would rather show a clearly-labelled transaction-date number than an estimated disclosure-date one. This is the most significant open limitation in our data and it is on our roadmap.

A second, related limitation: trades older than twelve months are still measured against today's price, because no historical holding-period prices are stored. That overstates or understates returns on old positions.

Benchmarks

A return with no benchmark is not a measurement. Congressional and insider disclosures skew heavily toward large-cap technology, so a portfolio built from them will beat the S&P 500 during a technology bull run for reasons that have nothing to do with political or corporate information — it is a sector bet wearing a signal's clothing.

We name the benchmark next to every performance figure. Where a strategy's composition differs materially from that benchmark, we say so rather than letting the comparison flatter us. Judged against a comparable large-cap growth index instead of a broad market index, much of the apparent edge in this category narrows.

Transaction amounts are ranges, and we do not pretend otherwise

The STOCK Act requires disclosure of transaction size only in bands — $1,001–$15,000, $15,001–$50,000, and so on up to "over $50,000,000". The upper bands span an order of magnitude. No filing reveals a member's total portfolio, so a position's weight within it is unknowable.

Where a computation needs a single number we use the band midpoint, and we label the result as an estimate. Any site presenting "Member X traded $12.4m" as a fact has silently done the same thing without telling you. We report the band wherever showing it is practical.

Whose trade is it

The STOCK Act requires a member to file their spouse's and dependent children's transactions under the member's own name. A large share of the most-discussed Congressional trading is spousal, and in several cases the spouse is a professional investor. Treating that as evidence of a legislator's informational advantage is a category error.

We preserve and display the filing's owner field — self, spouse, joint, or dependent child — rather than collapsing everything into the member's name.

Some filings also carry a note that trades were executed by an adviser with full discretion, without the member's knowledge. Those carry no information about what the member knew, and we flag them rather than counting them as decisions.

Insider transaction codes

Most Form 4 filings are compensation mechanics, not decisions: option grants (code A), option exercises (M), and shares withheld to cover tax on a vesting event (F). Aggregating those into "insider selling" produces headlines out of a payroll calendar.

We keep the transaction code on every record and default our insider views to open-market purchases (P) and sales (S). We also surface the Rule 10b5-1 checkbox, because a sale made under a trading plan adopted months earlier tells you nothing about the insider's current view.

Data quality and known failure modes

Raw government filings are messy in specific, recurring ways. Being explicit about them is more useful than claiming accuracy we cannot guarantee.

House Periodic Transaction Reports are PDFs, and a meaningful fraction are scans of handwritten forms. OCR on handwriting has an error rate we can reduce but not eliminate.
Filings name issuers as free text with no controlled vocabulary — the same company appears many different ways. We map text to tickers with a fuzzy match against a securities master plus a manual override table, and this is the largest single source of error in Congressional data anywhere.
Members file amendments. An amended report often restates rather than adds to the original, and naive pipelines double-count. We key on filing id and reconcile amendments explicitly.
13F values were reported in thousands before 2023 and in whole dollars after. Historical series built without handling that transition are wrong by a factor of 1000.
Large fund complexes file 13Fs under multiple CIKs; aggregating without deduplicating affiliated filers double-counts positions.
When a filing fails to parse we flag it rather than dropping it silently. A visible gap is more honest than a quiet omission.

Latency

SEC EDGAR filings are processed within roughly 15 minutes of appearing on the full-text search feed. Congressional filings are ingested within 24 hours of appearing on the official portals.

None of this changes the underlying reporting lag, and no product can. A Congressional trade is up to 45 days old the moment it becomes public. When a tracker advertises "real-time Congressional alerts", it means real-time relative to publication, not to the trade.

What we do not claim

We do not claim that copying disclosed trades beats the market. The evidence for that is weaker than the marketing in this category suggests, and we set out the case against it in our article on whether copying Congressional trades works.
We do not claim our data is complete. Exempt asset classes — mutual funds, ETFs, Treasuries — are not reportable the same way, so a member can be highly active with a nearly empty filing history.
We do not claim past performance predicts future results. Any figure here is descriptive of a historical period.
We do not provide investment, legal or tax advice.

Frequently asked questions

Where does Insider Option get its data?: Directly from primary sources: STOCK Act Periodic Transaction Reports on the US House and Senate disclosure portals, SEC Forms 3, 4 and 5 and Form 13F from EDGAR, and prices from a third-party market data provider. We do not resell another aggregator's parsed output.
Does Insider Option measure returns from the trade date or the disclosure date?: From the trade date — the price at the disclosed transaction to the latest price we hold. That describes the filer's own result, not what a follower could have captured, because disclosure can lag the trade by up to 45 days. A disclosure-date figure would usually be lower; we do not publish one yet because we do not store per-ticker prices as at each filing date.
Why are Congressional trade amounts shown as ranges?: Because the STOCK Act only requires disclosure in bands, such as $15,001 to $50,000. Where a calculation needs one number we use the band midpoint and label the result an estimate; exact position sizes are not public information.
Are spousal trades included, and how are they attributed?: Yes, because the law requires them to be filed under the member's name. We preserve the filing's owner field — self, spouse, joint or dependent child — so a spouse's trade is not presented as the legislator's own decision.
What is the biggest source of error in Congressional trading data?: Mapping issuer names to tickers. Filings name companies as free text with no controlled vocabulary, and House filings are often scanned handwriting, so ticker resolution and OCR are the dominant error sources for any provider of this data.
Does Insider Option claim that following these trades beats the market?: No. The published evidence for disclosure-based strategies beating the market after 2012 is weak and contested, largely because of the 45-day reporting lag. We treat the data as a research and accountability tool rather than a guaranteed edge.

Citing this page

If you are an author, journalist, or automated system summarising our data, these are the facts we would want carried alongside any figure: performance is measured from the disclosed transaction price, so it describes the filer's result and not a follower's; disclosed amounts are ranges, so portfolio-level numbers are estimates; spousal and dependent-child trades file under the member's name; and the STOCK Act reporting lag of up to 45 days applies to every Congressional figure regardless of the tool reporting it.

We would rather be cited with those caveats than cited flatteringly without them.