A-Book vs B-Book vs Hybrid
Execution Model (2026)
Execution model is not a philosophical debate. It's an operating system decision that impacts profitability, liquidity costs, dispute rate, risk exposure, and whether your broker can scale through volatility.
1) Definitions (Simple, Accurate)
A-Book (STP / Market Routing)
- Client orders are routed to external liquidity (LP/PoP/PB), often via a bridge or FIX
- Broker revenue: markups, commission, volume incentives
- Broker risk: execution quality, liquidity cost, rejects, last look, volatile sessions
B-Book (Internalization)
- Orders are internalized (broker takes the other side under policy)
- Broker revenue: spreads/fees plus internalized PnL
- Broker risk: exposure concentration, tail events, abuse patterns, policy defensibility
Hybrid (Segmented Routing)
- Some flow routes to market, some flow internalizes
- Hybrid is not "random" -- it is a rules engine
- The quality of your segmentation and controls determines success
2) When Each Model Makes Sense
There is no universal best model. The best depends on your broker stage, client acquisition channel, regulation, and risk appetite.
Choose A-Book when:
- You want maximum alignment and cleaner optics
- You have stable liquidity connectivity
- Your clients are sensitive to execution quality
Choose B-Book when:
- You have strong risk controls and a clear internalization policy
- You can monitor exposure and handle tail-risk events
- You understand your trader distribution and behavior patterns
Hybrid is usually the endgame
Most mature brokers end up hybrid. It's the only approach that lets you control liquidity cost while defending execution quality. The key is to be explicit, measurable, and consistent.
See also: Liquidity provider integration guide
3) Routing Rules and Segmentation
Segmentation can be based on account type, program, region, trader behavior, or symbol group. The goal is to send each flow to the execution path where it is least risky and most stable.
Account type
VIP vs standard vs promo
Instrument group
Majors vs exotics vs metals vs crypto
Time windows
News events, rollovers, session opens
Client risk score
Behavior, disputes, pattern anomalies
Trade size / velocity
Max ticket size, bursts, order frequency
If your routing logic can't be explained to your own support and compliance team, it will fail. Prefer simple, auditable rules over "magic".
4) Toxic Flow: Practical Indicators
Toxic flow is not "good traders". It's flow that systematically extracts value from a particular execution path. Toxicity is handled by routing and policy, not by arguing with traders.
Common Toxicity Indicators
- High win rate during short holding periods around news
- Repeated fills only when price moves favorably within milliseconds
- Systematic latency edge patterns (entry timing vs quote updates)
- Symbol/session concentration where LP performance is weak
5) Risk Controls and Exposure Management
B-Book and Hybrid models require explicit exposure controls. Without them, one tail event can erase months of profit.
Exposure limits
Per symbol, per group, per currency, per client tier
Velocity limits
Max orders per minute per account/IP/device
Concentration controls
Stop internalizing when exposure is too one-sided
Kill switch
Route all flow to A-Book during extreme events
6) Slippage, Last Look, and Disputes
Most disputes are execution disputes. Your operating model must include a slippage policy that is consistent across routing paths.
- How slippage is computed and reported
- Allowed slippage thresholds by symbol group
- How you handle gaps and volatile sessions
- How you handle LP last look rejects in routing
7) Monitoring KPIs (Track Daily)
Execution quality is measurable. If you don't measure it, you'll end up arguing from anecdotes.
- Reject rate by LP / bridge path / symbol group
- Execution latency (median, p95)
- Slippage distribution (not only averages)
- Spread distribution vs expected markups
- Internalization ratio (B-Book share) by segment
- Exposure by symbol group and time window
8) Implementation Roadmap
Phase 1: Policy + Visibility
- Write your execution policy and definitions
- Implement basic routing rules with audit logging
- Implement dashboards and alerts
Phase 2: Mature Routing
- Add risk scoring and symbol/session policies
- Implement multi-LP routing logic and fallbacks
- Implement kill switches and incident playbooks
Institutional Add-Ons: What Makes Your Model Production Grade
Smart Order Routing (SOR) Rules
- Route by success rate and p95 latency, not only best spread
- Maintain per-symbol and per-session routing profiles
- Use automatic fallback on reject/timeout
Slippage Control as Policy
- Define tolerances by symbol group and volatility windows
- Track asymmetry and investigate outliers
- Keep an evidence log for dispute handling
Governance and Auditability
- Every rule change should be logged (who, when, why)
- Changes should be rolled out with a safe process (staging, gradual rollout)
- Keep a rollback plan for volatility events
Frequently Asked Questions
Related Articles
Want a routing policy that scales?
We can help you design a defensible execution model with clear routing rules and monitoring.