MSR Decode Discuss a matter

Capabilities / Market Integrity & Compliance

Is a flagged anomaly real, or a data artifact?

For platforms, risk teams and compliance functions with a surveillance flag, an unusual-activity question, or a market-data quality concern. MSR Decode independently reconstructs the underlying activity to test whether a flag reflects real trading behavior or an artifact of the pipeline that produced it.

Published example — trade-endpoint case
Default view

Maker-side fills silently omitted → apparent phantom positions

Complete view

Fills re-fetched without the default → every flagged redemption reconciled

150 of 150 reconciled

An API default, not a trading anomaly. Read the case →

Who this is for

Platforms and marketplaces with listing or integrity duties; risk and compliance teams validating a surveillance flag; teams evaluating whether displayed market data faithfully describes underlying activity.

Typical questions

  • Is an existing surveillance flag or anomaly real, or a data artifact from the pipeline that produced it?
  • Did unusual trading precede a public announcement or resolution event?
  • Is a market's displayed volume generated by independent directional participation, or by repeated complementary trading between the same counterparties at minimal net cost?
  • How did a contract's settlement design affect participant incentives?
  • What would a predefined, recurring review of a specific universe of markets or accounts actually catch?

Evidence MSR can examine

  • Flagged trades, the control logic that flagged them, and an independent reconstruction of the same activity.
  • On-chain settlement events and contract terms, tested against the market's own published rules.
  • Order-level fills and timing, tested against public information and liquidity conditions.
  • Raw venue API responses, cross-checked against the underlying chain.

Analytical methods

The same disclosed, ten-step standard used throughout MSR's public research; see the full methodology. Independent reconstruction tests whether the flagging pipeline itself is the source of an anomaly — our own pipeline is the null hypothesis in every engagement, public or commissioned. A recurring-review engagement is scoped as a defined, predefined-cadence project against a stated universe, not continuous real-time monitoring.

Potential work product

Scope determines which of these ship for a given engagement.

  • Executive finding — a clear answer to the bounded question.
  • Timeline — trading activity against relevant external events.
  • Source schedule — records, URLs, APIs, datasets and provenance.
  • Technical workpaper — methods, assumptions, limitations and reproduction material.
  • Documented review trail — for a recurring-review engagement, a dated record of what was checked and when.

Scope boundaries

MSR performs defined forensic reviews and recurring human-led analysis on a bounded question or universe. It is not offered as an enterprise trade-surveillance software platform, and MSR cannot certify a control's overall effectiveness from one sample. A negative or insufficient-evidence finding is a valid outcome; see the independence policy.

Discuss a matter

MSR reviews the question, available records, scope and appropriate work product before accepting an engagement.