What a trade tells you. And what it doesn't.
An executive buys shares. What can you learn from that? Discover how to read their trades, connect the evidence and tell a fact from an assumption.
Evidence first. Interpretation clearly bounded.
Select a field to see what it tells you and what it leaves unanswered.
In this purchase example, quantity multiplied by price per share gives the amount. It does not tell you how much of the person's wealth it represents.
Find your answer
Look up a chapter or explore the guide at your own pace.
What's behind a trade?When does it become public?How can the data be compared?What can I infer?Which markets can I research?How do we build these figures?Build with the APIHow figures are calculated and what is excludedHow rankedMethodology and coverageRead methodology notesWhat's behind a trade?
Read the fields of a disclosure and separate reported facts from unknown motives.
An insider is an executive, a person with managerial responsibilities or someone closely associated with them. Their reportable transactions become public; this alone does not imply unlawful use of inside information.
Read a trade, field by fieldFrom the source to your research
Start with the disclosure published by the authority, including its original labels and notes.
This is a schematic sequence for learning, not a live view of processing or a measure of how long it takes.
Explore the methodology in depth
Evidence before interpretation
Insider disclosures identify who acted, the issuer, role, instrument, date, and authority. OpenInsider treats them as evidence, not stories.
The API and public layer share the same canonical resources. Filings bridge normalized fields and authority documents.
- Source record
- CNMV or AMF publishes a disclosure page, data record, or document.
- Canonical entities
- Issuer and party identities are resolved without discarding published names or source roles.
- Research object
- Transactions become queryable while retaining filing and authority provenance.
When does it become public?
Distinguish the trade date from notification and publication.
The trade takes place
This is the transaction date. It does not mean the information is already public.
The person notifies
Within 3 business days of the tradeUnder MAR, the person subject to the reporting obligation notifies both the issuer and the competent authority promptly, no later than this deadline.
The issuer publishes
Within 2 business days of receiptThe issuer makes the information public within this deadline after receiving the notification, not after the trade date.
Reporting depends on the applicable thresholds and jurisdiction. This simplified MAR timeline is not legal advice, and disclosures are not a real-time feed.
The baseline kicks in at EUR 20,000 yearly. Authorities may set EUR 10,000 or EUR 50,000 instead. Notifications are due within 3 business days; issuers publish within 2.
Explore the methodology in depth
A European reporting baseline, delivered nationally
Article 19 of MAR sets the EU disclosure baseline for managers and their associates. National authorities handle their own interfaces and thresholds.
The EU template can show individual transactions and an aggregate for same-nature, same-day, same-venue trades. Purchases and sales are never aggregated together.
Read the current Market Abuse Regulation and its implementing disclosure template.
How can the data be compared?
See how different source labels become comparable fields without losing the original.
As written in the source
In the normalized model
BUYPurchaseFictional example · In this example, the source labels 'Adquisición' and 'Acquisition' map to BUY. The original wording is preserved so you can check the classification against the document.
Source currency is preserved. Public aggregates only include values validated as exact EUR.
View the model's six fields
Normalization without erasing the source
Names, roles, currencies, and dates vary across filings. The canonical model makes fields comparable while keeping original evidence.
| Layer | Preserved evidence | Normalized research field |
|---|---|---|
| Person or entity | Published name, aliases, and party type | Canonical insider identity |
| Role | Original national-language label | Comparable relationship classification |
| Transaction | Reported nature, instrument, venue, quantity, and price | Typed values and consistent direction |
| Filing | Authority ID, status, document, and amendment relation | Stable filing resource with version provenance |
| Monetary amount | Source-reported value and currency | Validated exact-EUR aggregate or unavailable |
| Date | Transaction, notification, and publication dates | Separate timestamps with explicit meaning |
What can I infer?
Compare purchases, sales and other activity without treating them as recommendations.
Net and total tell different stories
Choose an example to see what changes when a disclosure contains purchases, sales or other activity.
Illustrative example · fictional data- Directional net
- €60,000
- Total reported value
- €165,000
Net = BUY - SELL; OTHER is excluded. Total = BUY + SELL + OTHER.
Combined activity, distinct meanings
This example combines the purchase, sale and free compensation share grant. Net subtracts the sale from the purchase; total includes all their reported values.
Total describes the value reported across all categories. Net describes only the balance of purchases and sales; neither reveals intent.These are hypothetical exact-EUR values, with no currency conversion. The comparison shows reported amounts, not a price chart or a forecast. OTHER is non-directional, and its value is not necessarily cash paid.
Before drawing a conclusion
- Who acted?
Check the person or entity and their relationship to the company. Similar names do not prove a shared identity.
- What actually happened?
Read the type of transaction, instrument, quantity and value. Separate a purchase from a grant or another reported operation.
- Which date are you reading?
Distinguish the transaction, notification and publication dates. Public information may describe an earlier event.
- What context is missing?
Look for related filings and explanatory notes. Missing evidence is not zero activity, and a trade alone does not reveal a motive.
- What does the original say?
Open the source document and check its notes and any amendments before relying on a summary.
Explore the methodology in depth
Explain signals; do not sell certainty
Missing evidence is shown as unavailable, not zero. Source-native labels remain visible when normalization cannot safely resolve them.
Charts and rankings are research aids that expose supporting values, sample size, and source links.
Other reported operations — typically free share grants from compensation plans, scrip dividends, pledges or loans — increase or encumber holdings without a proactive market decision. Their amounts are reference valuations (often the closing price, sometimes zero), not cash paid, and they never feed buy/sell signals.
OpenInsider provides public information and context. It does not provide investment or legal advice.
Which markets can I research?
Check disclosure, listing and price availability separately.
Spain · CNMV
France · AMF
Spain's CNMV and France's AMF are live. France's history starts January 1, 2023.
Regulatory disclosures
Public records of reportable transactions, organized by authority and jurisdiction.
Exchange listings
The exchanges and instruments for which listing information is available, checked independently of disclosures.
Price series
Historical market prices, with their own availability and limits. A disclosure does not guarantee a price series.
Disclosure, listing, and price coverage are independent. Italy and Germany remain planned.
Review market coverageExplore the methodology in depth
Europe-wide contract, availability by market and data type
- Live disclosure jurisdictions
- Spain · France
- Live disclosure authorities
- CNMV · AMF
- Planned jurisdictions
- Italy · Germany
Spain + France disclosures live
Regulators are authoritative. OpenInsider is a normalized research layer, not advice.
How do we build these figures?
Check counting units, amendments, sources and operational limits before reusing data.
Before using a figure
- Check what is counted
One filing may contain several transactions. People, issuers and analytical episodes are different units; do not interchange their counts.
- Read currency and nature
Public monetary aggregates use exact EUR only, without conversion. Other operations may represent a grant or transfer rather than cash paid.
- Do not confuse absence with zero
An unavailable amount, ambiguous identity or missing price series does not prove zero activity. Illustrative figures do not fill those gaps.
- Distinguish coverage from access
The accessible window depends on access mode and plan. An unseen record may reflect that scope, coverage or an error, not that a transaction never happened.
What each summary counts
Issuer figures summarize six months of effective filings. Local ledger filters, such as hiding zero-value rows, do not recalculate those aggregates. On a person’s profile, the count and net use the loaded sample of up to 100 accessible filings: they do not necessarily represent the full archive.
Within a filing, transaction counts refer to rows. Net = buys − sells; total = buys + sells + other operations. Some amounts are derived from price × quantity when no total is disclosed: a calculated amount is not an explicitly reported source figure. Economic-nature labels provide context distinct from the original direction.
How amendments are handled
Superseded filings remain evidence and link to successors when that relationship is identified. Issuer aggregates and the person sample’s net exclude known superseded versions to avoid counting them again. An evidence ledger may show both.
We do not infer amendments from similar names or figures alone. Extraction, classification or identity corrections require verification. Updating a profile does not itself establish that all derived analyses have already been recalculated.
Dates, updates and gaps
Transaction, notification, official publication and access dates differ. The latest observed publication does not establish source completeness or the update time of the entire website. We do not provide real-time data.
France’s validated public history starts on 1 January 2023. Disclosure, listing and price availability are checked separately. Source delays, unreadable documents and records awaiting validation may cause gaps. Availability claims stay within the public contract.
Check market coverageSources and reuse limits
CNMV and AMF supply regulatory filings. Original documents remain authoritative. Context links, logos and market data are secondary resources, not regulatory evidence or independently verified identities.
Public access does not grant an unlimited license or remove the rights of people mentioned. Reuse depends on source and data type. The service does not assign filing permissions to prices or enrichment.
Methodology change history
This history records documented revisions of this guide. It is not a deployment log, a private incident log or a record of every filing correction, which should be checked on the relevant records.
- Explanation of counts, amendments and reuse
Adds reading rules, person-sample limits, sources, date distinctions and correction and citation links. This explanatory revision does not change calculations, amounts, coverage or permissions and requires no recalculation. This entry is not evidence of a production deployment.
How figures are calculated and what is excluded
Company figures
Issuer figures summarize six months of effective filings: superseded filings are not counted again. The ledger preserves evidence and may have a different scope; hiding zero-value rows does not change these aggregates.
Public monetary aggregates use exact EUR, without currency conversion. Net = buys − sells; total includes other operations, which do not necessarily represent cash paid. Unavailable data is not zero.
Person and organisation figures
The count and net use the loaded sample of up to 100 accessible filings, not the full archive or the ledger page. Net excludes filings superseded by an amendment. Companies count distinct disclosed relationships.
Public monetary aggregates use exact EUR, without currency conversion. Net = buys − sells; total includes other operations, which do not necessarily represent cash paid. Unavailable data is not zero.
Filing figures
Amounts refer to this filing; a superseded version remains evidence but must not be added again to its successor. The transaction count counts rows, not people or filings.
Public monetary aggregates use exact EUR, without currency conversion. Net = buys − sells; total includes other operations, which do not necessarily represent cash paid. Unavailable data is not zero.
How ranked
Net equals eligible purchases minus eligible sales, divided by stored EUR market capitalization, multiplied by 100. The five highest positive ratios and five lowest negative ratios lead each list. Equal ratios are ordered by absolute net amount, then company ID. Party counts and the share of traded amount provide context; they do not change the ranking.
Only reviewed ordinary equity acquisitions and disposals from CNMV and AMF count, by execution date in the last three rolling calendar months (UTC). Superseded, cancelled and unlinked amendment filings, grants, gifts, subscriptions, derivatives and funds are excluded. An issuer is excluded if a comparable filing has incomplete or non-EUR amounts, or no verified current-ISIN EUR market cap observed within 120 days. This is eligible comparable flow, not complete coverage of all activity.
An activity proxy, not a forecast or probability. Selling does not imply a bearish outlook. Identified parties may be people or organizations; counts do not establish independent decision-makers. Market cap is a stored observation, not a live quote.
Methodology and coverage
Signed sample-adjusted excess = median cohort excess × n / (n + 5), where n is the number of evaluated episodes. This is not a probability. At least six episodes are required; single-issuer histories are eligible. Trades are grouped into 30-day episode chains, but outcome windows can overlap: grouping does not guarantee statistically independent observations. Peers belong to the same jurisdiction and exclude the insider being evaluated. Returns with an absolute magnitude of 1,000% or more are excluded pending review, not replaced with zero. Historical outcomes are not forecasts or a trading strategy.
Read methodology notes
Loading insights
Take a question to the evidence
Explore a company, read a disclosure and check what supports your interpretation. This guide is a reference you can return to, not investment or legal advice.