Skip to content

Lesson 3.4 - Account Correlation

Scenario

One Account May Be Noise. Ten Similar Accounts May Be a Pattern.

One account trades XAUUSD quickly during news. That may be normal. Ten accounts trading the same symbol, entering within seconds, using similar size patterns and closing at similar intervals deserve a structured cluster review.

Learning Objectives

  • Understand behavioral clustering.
  • Identify useful correlation dimensions.
  • Separate legitimate similarity from coordinated patterns.
  • Build evidence for review without making unsupported claims.

What Is Behavioral Clustering?

Behavioral clustering groups accounts or trades that show similar characteristics across time, symbol, execution style and outcome. It is an analytical method, not a conclusion of wrongdoing.

Useful Correlation Features

Feature Example
Time correlation Orders enter within a narrow window
Symbol correlation Same symbols repeatedly
Direction correlation Similar buy / sell bias
Size correlation Similar lots or progression
Holding-time correlation Similar open-close interval
Event correlation Same news or thin-liquidity windows
Execution correlation Similar slippage or post-trade movement
Relationship signal UID, device or other approved relationship evidence

Cluster Analysis Flow

Order / Account Data→Extract Features
Features→Compare Timing / Symbol / Size
Comparison→Check UID / Relationship Signals
Evidence→Review with Shift Leader / Risk

Similar Does Not Always Mean Coordinated

Accounts may look similar because many clients trade the same event, use common EAs or react to the same market condition. A cluster review should look for repeatability, relationship evidence and execution-context signals before escalating.

Case Study

Synchronized News Flow

Multiple accounts enter XAUUSD in the same direction within seconds of high-impact news. Most close quickly. A subset shares similar UID relationship signals and favorable post-trade movement.

Reporting Template

Use a short, factual summary:

Time window:
Symbol:
Accounts / UID scope:
Shared features:
Execution context:
Evidence saved:
Current assessment:
Next action / Shift Leader:

Best Practice

  • Review repeated patterns, not one trade.
  • Use approved relationship signals only.
  • Keep personal names out of training records.
  • Escalate evidence, not accusations.

Practical Training

Wrong handling:

"These accounts traded together once, so they are coordinated."

Correct handling:

"The accounts show repeated timing, symbol, holding-time and UID relationship signals. Evidence should be reviewed by Shift Leader / Risk before action."

Summary

Behavioral clustering helps the Dealing Room identify repeatable account patterns. It supports risk review only when used carefully and with multiple evidence points.

Completion Criteria

  • Can explain the key risk or operational objective of this lesson
  • Can identify the required systems, data, or evidence to review
  • Can describe the correct escalation or handling process
  • Has completed Shift Leader / follow-up review or practical confirmation