Risk Intelligence01Options Overlay02Technology03Infrastructure04Partnership05Contact06Request a book audit
HedgX/Risk Intelligence

From raw broker data to governed decisions.

How HedgX turns MT5 data into evidence-based routing, exposure and hedging decisions — account by account, and across the book.

02
Risk intelligence
Ten dimensions. Four books. One decision layer.
The core philosophy

Every trader is an economic exposure.

Broker risk cannot be reduced to “winner = A-book, loser = B-book.” That logic is too simple to manage real broker exposure.

  • A winning trader may be temporarily lucky.
  • A losing trader may carry dangerous recovery risk.
  • A profitable account may belong to a copy cluster.
  • A small account may reveal a toxic strategy before it scales.
  • Negative float today may become broker liability tomorrow.
The framework

Each trader reviewed across 10 dimensions

01Expected broker value
02Floating P&L risk
03Tail-risk exposure
04Trading behaviour
05Strategy type
06Market timing
07Liquidity cost
08Account correlation
09Data quality
10Decision confidence
The risk intelligence stack

Five integrated stages — each more decisive than the last.

Each stage transforms broker reality into a more structured, more decisive input for the dealing function.

01
Data

Data ingestion

Reconciled, normalised broker data
  • — MT5 reports
  • — Deals & orders
  • — Positions
  • — Balance operations
  • — Book / group mapping
02
Features

Feature engineering

Engineered signals over transactions
  • — Economic features
  • — Exposure features
  • — Behavioural features
  • — Relationship features
  • — Data-quality features
03
Models

Model layer

Classification & quantitative scoring
  • — Behavioural class
  • — Toxicity scoring
  • — Tail-risk estimate
  • — Cluster detection
  • — Confidence
04
Decisions

Decision layer

What gets actioned, by whom
  • — Routing review
  • — Exposure adjustment
  • — Options hedge trigger
  • — Escalation
  • — Action queue
05
Governance

Governance

Audit, oversight, escalation
  • — Reason codes
  • — Confidence gates
  • — Audit trail
  • — Reconciliation
  • — Reporting
Model outputs

Every account, scored across nine dimensions.

Each decision is traceable to a specific score, reason code or confidence level.

01Classification

A · B · C · D
A
B
C
D

02Confidence

0.00 – 1.00
0.82

03Expected broker value

indexed Δ
▲ positive

04Floating-risk score

0 – 100
23

05Toxicity score

0 – 100
15

06Cluster status

isolated · linked · master
  linked (2 accounts)

07Tail-risk estimate

indexed
elevated

08Liquidity-cost impact

indexed Δ
▼ low

09Reason codes

audit string
TX-04 · CL-02
Institutional risk intelligence

Nine questions HedgX answers, in real time.

Grouped into five decision domains — from data, through analysis, to action.

Routing
01
Which clients should be internalised?
02
Which clients should be externally hedged?
Surveillance
03
Which clients require closer monitoring?
04
Which strategies create hidden risk?
Exposure
05
Which accounts are linked through copy?
06
Where is the real floating exposure?
Valuation
07
What is the expected broker value?
08
What is the worst plausible outcome?
Decision
09
What action should the dealing team take now?
Dynamic book classification

Beyond A and B. Four books.

A more advanced framework than simple A-book / B-book labels. Each book has explicit indicators and explicit handling.

A-Book
Hedge externally / do not fully internalise
Indicators
  • Evidence of a repeatable edge
  • Adverse-selection signals
  • Exposure the book should not carry
B-Book
Standard internalisation
Indicators
  • Random retail flow
  • No adverse-selection signal
  • Manageable exposure
C-Book
High-confidence negative expectancy
Indicators
  • Persistent negative expectancy
  • No toxicity signals
  • Managed with controls
D-Book
High-confidence monitored exposure
Indicators
  • Verified, repeatable signal
  • Specialist exposure handling
  • Active monitoring and sizing
Broker P&L intelligence

Closed client P&L is not enough.

Broker profitability is built from many components. HedgX attributes each one — including floating-P&L impact and what-if cost — to surface the true broker-side outcome.

100Client closed P&L+18Floating Δ+22Comm · swaps+35Spread · markup−14Rebates−28LP cost · slip−6Book-switch127Broker P&L
Illustrative values · indexed to client closed P&L = 100
Floating exposure

A first-class risk input — not a footnote.

Closed P&L tells you what already happened. Floating exposure tells you what is about to happen — and where the broker actually carries risk.

Two paradoxes

Closed P&L can mislead

01

Profitable on closed P&L — yet carrying a major unrealised liability.

02

Apparently exposed — yet negative float may resolve into broker profit.

Floating review inputs

Floating exposure changes the real risk

01Current floating P&L
02Floating P&L change
03Margin level
04Stop-out distance
05Open-position structure
06Basket recovery behaviour
07Grid & martingale exposure
08Deposit behaviour
09Symbol volatility
10Correlated exposure across accounts
Trader behaviour analysis

Same P&L. Different risk.

Two accounts can show identical P&L with completely different exposure profiles. HedgX measures how — not just how much.

01

Normal retail flow

Risk fingerprint
EV
Float
Cluster
Behaviour
Toxicity
02

High-risk recovery system

Risk fingerprint
EV
Float
Cluster
Behaviour
Toxicity
03

Copy following

Risk fingerprint
EV
Float
Cluster
Behaviour
Toxicity
04

Event-window trader

Risk fingerprint
EV
Float
Cluster
Behaviour
Toxicity
05

Execution-sensitive flow

Risk fingerprint
EV
Float
Cluster
Behaviour
Toxicity
Principle

We do not only ask whether a trader won or lost — we ask what the behaviour means for the broker.

Strategy fingerprinting

Routing follows strategy — not just P&L.

The same closed P&L can come from very different patterns. Each demands a different routing, exposure and monitoring decision — set per account, by confidence and exposure, under the broker's mandate.

01Profiled

Scalping

High-frequency micro moves
02Profiled

News trading

Macro / event windows
03Profiled

Grid systems

Algorithmic ladder logic
04Profiled

Martingale recovery

Double-down loss recovery
05Profiled

Swing trading

Multi-session direction
06Profiled

Copy following

Mirror-master replication
07Profiled

Gambling behaviour

Random sizing, no edge
08Profiled

Chronic negative-EV

Persistent losing patterns
09Profiled

Execution-sensitive

Latency or spread arbitrage
Toxic flow detection

Catch dangerous flow before it scales.

Some flow creates disproportionate broker risk — sensitive to price movement, execution conditions or volatile windows. HedgX surfaces it early, and HX Guard scores it inside the server.

01Adverse selection
Flow that is systematically right about the next move
02Execution-edge capture
Profitability that depends on the broker's execution conditions
03Event concentration
Results concentrated in scheduled or unscheduled volatility
04Regime dependence
Strong P&L only in specific market conditions
05Coordination
Behaviour that is coordinated across accounts or identities
06Automation
Unapproved automated trading in retail groups
07Structural edge
Persistent lifetime profitability with a repeatable signature
The goal
Identify dangerous flow before it becomes a material broker loss.
Copy-trading & cluster risk

Ten small accounts. One large exposure.

HedgX analyses trade similarity to surface coordinated risk hidden across accounts. Cluster behaviour is treated as a single economic exposure.

Cluster view · illustrative
01020304050607080910M
Linked accounts
10
Confidence
High
Treated as
One exposure
Combined float
Elevated
Detection signals
  • Coordinated trade behaviour across accounts
  • Leader / follower relationships
  • Linked-account probability
  • Cluster-level P&L
  • Combined floating exposure
  • Group / IB concentration
Portfolio-level risk

Acceptable per account. Dangerous in aggregate.

Risk managed at book level — not just account level. HedgX evaluates concentration across five portfolio dimensions.

Net exposure

By symbol

Gross exposure

By book

A-Book
B-Book
C-Book
D-Book
Concentration

By account

#101166
#100968
#101479
#101521
#100822
Group / IB

Campaign exposure

IB-North
IB-Asia
IB-LatAm
Affiliate
Direct
Correlation

Open positions

EUR/GBP
XAU/XAG
DJ30/SP500
BTC/ETH
What-if risk analysis

What if the decision had been different?

A strong dealing function studies alternative outcomes — not just actual results. HedgX measures missed P&L, loss avoided and forward-policy delta.

ACCOUNT FLAGGED01DECISION POINT02OUTCOME WINDOW03REVIEW04ALTERNATIVE ROUTEACTUAL ROUTE
What-if output
Missed broker P&L
+$48k
Loss avoided
+$31k
Forward-policy Δ
+12pp
Decision discipline, measured.
Illustrative · routes shown for demonstration only
Prediction-market intelligence

What the market expects — against what the book holds.

HedgX tracks prediction-market pricing on upcoming economic, corporate and geopolitical events and maps each outcome to live A-book and B-book exposure — a forward view of sentiment before the event reaches price.

Event coverageScheduled macro dataCentral-bank decisionsCorporate & micro eventsGeopolitical developmentsElections & policy
EventCategoryImplied probability24h ΔSymbols affectedBook exposureSignal
FOMC decision — rates heldMacro
72%
+6ppUSD majors · US indicesB-book net long US500Monitor
US CPI prints above consensusMacro
41%
+9ppXAUUSD · EURUSDB-book net long XAUUSDReview
ECB delivers rate cutMacro
64%
−3ppEURUSD · GER40A-book balancedNormal
Mega-cap earnings beatMicro
58%
+4ppNAS100 · US equitiesB-book net short NAS100Review
Trade-tariff escalationGeopolitical
28%
+11ppXAUUSD · USDCNH · indicesB-book concentrated in goldEscalate
Regional ceasefire agreedGeopolitical
35%
−5ppCrude oil · XAUUSDB-book net long USOILMonitor
Illustrative · events, probabilities and exposures shown for demonstration only
01

Aggregate

Prediction-market pricing across macro, micro and geopolitical events.
02

Map

Each event linked to the symbols and asset classes it moves.
03

Overlay

A-book and B-book exposure, clusters and floating P&L per symbol.
04

Flag

Books positioned against a likely outcome, or odds repricing sharply.
05

Act

Routing review, leverage windows, options cover, broker notice.
Why it matters

Sentiment on news, macro and geopolitical events becomes a measurable input to routing, leverage windows and hedging — not an afterthought once the move has happened.

Liquidity & hedge-cost awareness

Two mistakes. Both expensive.

Hedging is not a binary good. The correct decision depends on risk-adjusted economics — measured across the full execution stack.

Mistake 1

Over-hedging flow that should be internalised.

Structurally losing flow is sent out — the broker pays spread and commission on profit that would have stayed internal.
Mistake 2

Under-hedging flow that creates asymmetric risk.

Adverse-selection or event-driven flow is internalised — the broker carries a position the book cannot absorb.
HedgX evaluates
01
LP commission
02
Spread quality
03
Slippage
04
Rejects
05
Execution speed
06
Bridge performance
07
Hedge cost
08
Collateral
09
Flow quality
A third route

Internalise the flow — and buy the tail. Options overlay →

Next chapter

Options overlay.

Cap the downside. Keep the book.