EquationDB

Market surveillance & abuse detection

For compliance officers, surveillance teams, exchanges and regulators

Unusual volume before earnings or rating changes, pump-and-dump round trips and volume anomalies are sequences of events EquationDB already detects. A surveillance analyst writes one line instead of a batch job.

  • Financial markets
  • 7 steps
  • Runs on live data

How EquationDB answers it

Each step is one query on live data. Run them in order, or open any one and adapt it: change the symbol, the station, the dates or the filter.

  1. Volume spike before a surprise

    Unusually large beats that were preceded by a 3x volume surge within two weeks, a possible sign of information leakage.

    EVENTS SEQUENCE volume_surge -> earnings_beat WITHIN 10 bars IN top500 LAST 1y WHERE rarity >= 90
    Try it →
  2. Volume before an upgrade

    Upgrades that came within three sessions of a volume surge: was the rating change traded ahead of publication?

    EVENTS SEQUENCE volume_surge -> analyst_upgrade WITHIN 3 bars IN top500 LAST 6mo
    Try it →
  3. Gap before the news

    The stock gapped up before the upgrade was published, to review for front-running of research.

    EVENTS SEQUENCE gap_up -> analyst_upgrade WITHIN 2 bars IN top500 LAST 1y
    Try it →
  4. Pump-and-dump round trips

    A 2.5-sigma rally reversed by a 2.5-sigma drop within a week, the price signature of a pump and dump.

    EVENTS SEQUENCE big_move_up -> big_move_down WITHIN 5 bars IN top500 LAST 6mo
    Try it →
  5. Heavy volume, flat price

    Volume in the top 1% of two years with almost no price change: block trades, crosses or quiet accumulation.

    GET top500 SELECT symbol, sector, ret, rvol(20) AS rv WHERE volume > pctl(99, 2y) AND abs(ret) < 1 SORT rv DESC
    Try it →
  6. Anomalous days in one name

    Every day in two years when return, range or volume was more than three sigma from the prior 60-day norm.

    ANALYZE anomalies MDB
    Try it →
  7. Insider sales before a bad reaction

    Insider sales in companies whose last report triggered a drop of more than 8%, ordered by size for review.

    INSIDERS IN top500 LAST 6mo WHERE type LIKE 'S%' AND symbol IN (EARNINGS IN top500 LAST 3mo WHERE reaction < -8) SORT value DESC
    Try it →

Run it on your own question

Sign in and press Try it on any step, or ask for a short demo on your data.

Request a trial

More financial markets use cases