Standard deviation of returns scaled to a yearly horizon. A volatility of 18% means returns typically swing ±18% per year around the mean. Higher volatility = larger swings up and down.
σ_annual = σ_daily × √252
InsidersTradesSigma
Plain-English definitions of every technical term used on this site. CAGR, Sharpe, MAR, Form 4 and the rest, no jargon. Hover any dotted-underlined term anywhere on the site to see its card.
Standard deviation of returns scaled to a yearly horizon. A volatility of 18% means returns typically swing ±18% per year around the mean. Higher volatility = larger swings up and down.
σ_annual = σ_daily × √252
Replaying a strategy on historical data to estimate how it would have performed. Vulnerable to overfitting: a strategy that looks great in backtest often fails live.
Percentage of months (or rolling 12-month windows) where the strategy outperformed its reference index (CAC 40 for EU, S&P 500 for US). Above 55% is meaningful, above 70% is rare.
Compound Annual Growth Rate. The constant yearly return that, if applied every year, would grow your starting capital to the final value over the period measured. It smooths out volatility into a single annualised number.
Multiple insiders at the same company buying within a short window (±30 days). Cluster signals historically outperform isolated insider trades because they suggest shared private information.
If CEO + CFO + 2 directors all buy in the same month, isCluster = true · si CEO + CFO + 2 administrateurs achètent dans le même mois, isCluster = true.
Synonym of Signal Score. The 0-100 number that weighs transaction size, role, clustering, sector context, and momentum into a single comparable conviction score.
Subscore (typically 35 pts of the composite) capturing the strength of insider purchasing activity for a given company: who bought, how much, how recently, and whether others followed.
Non-executive director on the board. Limited operational view but legal duty of care. Multiple board members buying together is a meaningful cluster signal.
Chief Executive Officer. Top operational executive. CEO insider buys historically carry the strongest signal: they have the most context and the most reputational risk.
Chief Financial Officer. CFO insider trades are particularly informative for earnings-related signals: they see the books before everyone else.
Chair of the Board of Directors. Strategic oversight role, often combined with significant equity stake. Founder-chairman buys are tracked separately as a high-conviction subset.
Stock's sensitivity to the broad market. Beta = 1 moves with the market; >1 = more volatile (e.g. tech ~1.3); <1 = less volatile (e.g. utilities ~0.6). Negative = inversely correlated.
Annual dividend per share divided by current share price. A 3% yield means 3 € per year for every 100 € invested at current price.
Yield = Dividende annuel / Cours
Earnings Per Share. Net profit divided by number of shares outstanding. Trailing EPS = past 12 months; forward EPS = next 12 months forecast.
P/E based on analyst forecasts of next 12-month earnings instead of trailing. Lower forward P/E than trailing P/E means analysts expect earnings growth.
Secret token used to authenticate REST/MCP API calls. Generated from your account settings. Treat it like a password: never commit to git, never paste in chat.
Free promotional tier (until end of public beta) granting unlimited API calls and signal access. Will revert to standard tiered pricing once stable.
Passwordless sign-in: we email you a one-time URL that logs you in directly. Lower friction than passwords and harder to phish.
Maximum API requests allowed per minute/hour/day per key. Free tier: 60/min. Pro tier: 600/min. Beta unlimited: no cap. Exceeding triggers HTTP 429 responses.
Term missing? Email us. Full methodology on the Methodology page.
CAGR = (V_final / V_initial)^(1 / années) − 1
A Sigma CAGR of 32.7% means 100 € invested would grow to roughly 178 € after 2 years on average · 100 € placés deviennent ~178 € après 2 ans en moyenne.
Sources: CFA Institute · Investopedia
Annualised return divided by maximum drawdown. Measures how much pain (worst peak-to-trough loss) you have to tolerate to capture the average return. Above 1.0 is solid.
Calmar = CAGR / |Max Drawdown|
Any decline from a previous portfolio high. Different from Max DD: drawdown is the live, ongoing dip from the latest peak; Max DD is the worst it ever got historically.
Price drift between the day an insider actually traded (transactionDate) and the day the regulator published the filing (pubDate). A large positive leak means the market already moved before you could act on the signal.
If the stock rose 4% between trade and publication, leak = +4% · si le cours a monté de 4% entre trade et publication, leak = +4%.
The worst peak-to-trough drop the strategy ever suffered during the backtest. A -25% Max DD means at the worst point the portfolio was 25% below its previous high.
Max DD = -25% over 2 years means you would have endured a 25% paper loss before recovery · vous auriez supporté 25% de perte latente avant rebond.
Live or held-out data the strategy was NOT trained on. OOS performance is the only honest measure: if a strategy works only in-sample, it has memorised noise.
Risk-adjusted return: how much annualised return you earn per unit of annualised volatility. Above 1.0 is good, above 2.0 is excellent, below 0 means the strategy lost money on average.
Sharpe = (R_strategy − R_riskfree) / σ_annualised
Sharpe 1.32 means the strategy earned 1.32 units of return per unit of risk · la stratégie gagne 1.32 unité de rendement par unité de risque.
Sources: William F. Sharpe (1966)
Number of trades matching the strategy filter over the full backtest period. Larger samples make every other metric more reliable. Below 100 signals, treat the numbers with caution.
Like the Sharpe ratio but only penalises downside volatility (losses), not upside swings. More relevant for asymmetric strategies that aim for large wins.
Sortino = (R_strategy − R_target) / σ_downside
Performance measured 90 calendar days after the filing's publication date (pubDate). The default holding period for our retail-friendly signals, balancing signal decay vs transaction-cost drag.
Backtesting technique: train on a rolling past window, test on the next out-of-sample window, slide forward, repeat. Closer to how a strategy would have run live with monthly recalibration.
Percentage of trades or holding periods that ended profitable. A 62% win rate means roughly 6 trades out of 10 closed in the green. High win-rate ≠ high return: a few big losers can wreck a high-win-rate strategy.
Boolean flag set to true when a declaration is part of a cluster (≥2 distinct insiders at the same company buying within 30 days). Surfaced as a badge on declaration cards.
An actionable signal published by the platform. Type is signal-buy (insider accumulated, likely undervalued) or signal-sell (insider unloaded heavily). Includes target horizon and conviction score.
Current generation of the signal scoring model (May 2026). Recalibrated quarterly with walk-forward validation. v3 added cluster detection and sector-relative z-scoring vs v2.
Our flagship rules-based strategy: buy when score ≥ 70 and cluster confirmed, hold T+90, equal-weight up to 20 positions. Tracked live out-of-sample since 2026-05-16.
Composite 0-100 score combining transaction size, insider role (CEO/CFO weighted higher), clustering with other insiders, sector context, and price momentum. Higher = stronger conviction signal.
The full set of companies eligible for ranking. Defaults to all companies with at least one DIRIGEANTS filing across our 28 tracked markets. Can be restricted to a market, sector, or size class.
A person with regulatory obligation to disclose trades in their own company's shares: corporate officers, directors, ≥10% shareholders, and PDMRs. Trading on non-public information remains illegal; the disclosed trades themselves are legal.
Market Abuse Regulation (EU 596/2014). The EU/EEA framework that requires PDMRs to disclose own-account trades within 3 business days. Enforced by AMF (FR), BaFin (DE), CNMV (ES), Consob (IT), AFM (NL), FSMA (BE), etc.
Sources: Regulation (EU) 596/2014
Canadian National Instrument 55-104. Requires Canadian insiders to file trade reports to SEDI (System for Electronic Disclosure by Insiders) within 5 calendar days.
Sources: CSA · SEDI
Corporate officer below the C-suite (e.g. SVP, General Counsel, COO). Covered by Section 16 / MAR disclosure rules. Officer buys are weighted slightly less than CEO/CFO buys.
Person Discharging Managerial Responsibilities. EU/UK regulatory category covering executives, board members, and senior managers who must disclose personal trades under MAR Art.19 within 3 business days.
Sources: MAR Art.19 · ESMA
Date the regulator officially published the filing. The only date a retail investor could legally have known about the trade. All performance numbers use pubDate as t=0.
US filing every officer, director, and ≥10% shareholder must submit to the SEC within 2 business days of trading their company's stock. Source of all US insider data on this site.
Sources: SEC EDGAR · Section 16
Section 16 of the US Securities Exchange Act of 1934 — the legal basis for SEC Form 3/4/5 insider reporting obligations. Targets officers, directors, and beneficial owners of ≥10%.
Date the insider actually executed the trade. Always earlier than pubDate (lag varies by regulator: SEC ~2 days, AMF ~4 days, SIX SER often ~7 days). The difference creates the leak.
Total value of all outstanding shares (price × share count). Used to classify companies as micro (<$300M), small ($300M-$2B), mid ($2-10B), large ($10-200B), mega (>$200B).
Price-to-Earnings ratio. Stock price divided by trailing 12-month earnings per share. Higher = more expensive relative to current profits. Industry-dependent; tech often trades at 25-40, banks at 6-12.
P/E = Prix / BPA