Methodology

1. Data sources

All data on Disclosyr comes directly from public regulatory databases. We do not purchase or license trade data. Sources include: SEC EDGAR (Form 4, 13F), Senate eFD, House Clerk FD, AMF France, BaFin Germany, AFM Netherlands, FSMA Belgium, and FCA UK.

Each filing is ingested in raw form, stored with its original payload, and normalised into a unified schema. The original source URL is always shown on every trade card.

2. Trade date vs disclosure date

Every trade on Disclosyr shows two dates: the date the trade actually occurred (trade date) and the date it was reported to the regulator (disclosure date). The difference is the disclosure delay, shown in orange on every card.

Regulatory deadlines vary: US Form 4 must be filed within 2 business days; STOCK Act PTRs within 45 days; EU MAR filings within 3 business days. In practice, delays of 20–45 days are common.

3. Return calculation

We calculate return from the disclosure date, not the trade date. This is a conservative and honest choice: by the time a disclosure is public, prices have often already moved.

For closed positions (a buy followed by a matched sell), we compute the realised return between the two disclosure dates.

Important:our returns will understate the insider's actual returns by the disclosure delay. A trade disclosed 30 days late means the insider had 30 extra days at the actual trade price. We display this delay explicitly on every card. Readers should factor it in.

4. Leaderboard aggregation

The leaderboard ranks insiders by total return across all their disclosed trades in a given period (1M, 3M, 6M, 1Y, 3Y, 5Y, all-time). We compute:

• Total return: volume-weighted average return across all disclosed positions in the period
• Win rate: percentage of disclosed trades that have a positive return at measurement date
• Benchmark return: SPY for US-domiciled insiders, STOXX 600 for EU insiders, over the same period
• Return vs benchmark: insider total return minus benchmark return

Leaderboard data is recomputed nightly. Performance figures may lag by up to 24 hours.

5. Price data

We use end-of-day closing prices from Financial Modeling Prep (FMP). Prices are stored at ingestion time and not retroactively adjusted. Splits and dividends are not yet factored into return calculations. This is a known limitation and will be addressed in a future release.

6. Limitations and disclaimers

• Returns are calculated from disclosure dates, not trade dates (conservative bias)
• 13F holdings are reported quarterly, so inferred trades carry additional uncertainty
• Transaction amounts may be ranges (as reported) rather than exact figures
• EU MAR data may be incomplete for smaller markets or older filings
• No split or dividend adjustment is currently applied
• Past performance does not indicate future results