Definition
Composite liquidity refers to the practice of aggregating price quotes from multiple liquidity providers (LPs) into a single, unified price stream. The broker's aggregation engine collects bids and asks from each connected LP, then constructs a composite book that shows the best available bid and ask across all sources. The client sees a single spread that represents the tightest combination of available prices.
This model is foundational to NDD / A-Book execution. Rather than relying on a single LP's pricing, composite liquidity creates competition among providers, which generally results in tighter spreads and deeper effective market depth. The quality of the composite depends entirely on the number, diversity, and caliber of connected LPs, as well as the sophistication of the aggregation algorithm.
What It Is / What It Is Not
What Composite Liquidity IS
- Aggregation of prices from multiple LPs into a single stream
- The best bid and best ask may come from different LPs
- Creates competition among providers, tightening effective spreads
- Requires an aggregation engine (bridge, PMS, or proprietary system)
- Core enabling technology for NDD / A-Book execution models
- Depth of book can be constructed from multiple LP levels
What Composite Liquidity IS NOT
- Not the same as ECN order book matching -- composite is pre-trade aggregation
- Not a guarantee of execution at the displayed price (last look may apply)
- Not automatically better than direct LP -- depends on LP pool quality
- Not immune to stale quotes during volatile markets
- Not a routing model -- it is a pricing construction method
- Not a regulatory term; it describes a technical architecture
How Aggregation Works
The aggregation process follows a consistent pattern regardless of the specific technology stack. Understanding these mechanics helps evaluate whether a broker's composite feed is genuinely competitive.
| Stage | Process | Key Consideration |
|---|---|---|
| 1. Collection | Each LP streams bid/ask quotes continuously | Quote freshness depends on LP connection latency |
| 2. Normalization | Quotes converted to common format, lot size, precision | Different LPs may quote different notional tiers |
| 3. Filtering | Stale quotes, outliers, and inactive feeds removed | Aggressive filtering improves quality but reduces depth |
| 4. Ranking | Best bid and best ask selected from remaining quotes | May include depth levels (top 3, top 5, full book) |
| 5. Markup | Broker may apply spread markup before publishing | Markup model affects the apparent competitiveness |
The final composite price is what the client sees on-screen. When the client trades, the order is routed to the LP whose quote was best at that moment, or through a SOR engine if multiple venues are available. The LP may still reject or requote the order (last look), which is why the displayed price is not a guaranteed execution price.
Where It Appears in the Execution Stack
Benefits & Trade-offs
| Factor | Detail | |
|---|---|---|
| Spread tightness | Competition among LPs generally produces tighter effective spreads | |
| Market depth | Multiple LP levels create deeper apparent liquidity | |
| Price diversity | Different LP types (bank, non-bank) provide varied pricing characteristics | |
| Stale quote risk | If one LP lags, its stale price may appear as best bid/ask briefly | |
| Complexity | Aggregation requires sophisticated technology and ongoing LP management | |
| Execution uncertainty | Displayed price is indicative; last look and latency affect actual fill |
Common Marketing Claims vs Reality
| Claim | Reality |
|---|---|
| "Aggregated from 20+ tier-1 banks" | Verify how many LPs are actually active and streaming, versus connected but dormant. Quality matters more than count. |
| "Tightest spreads in the industry" | Tight quoted spreads are meaningless if fill rates are poor or slippage is high. Ask for effective spread data, not indicative. |
| "Deep institutional liquidity" | Depth at top-of-book matters most. Many brokers show theoretical depth that evaporates during volatile conditions. |
| "Raw spreads from 0.0 pips" | Zero-spread snapshots exist but are not typical. Look at average spreads across sessions and instruments, not minimums. |
What to look for in an Execution Policy
- Does the execution policy describe how many and what types of LPs are aggregated?
- Is the aggregation method documented (best bid/ask, VWAP, weighted)?
- Does the broker disclose whether markup is applied on top of the composite?
- Are stale quote filters and their thresholds described?
- Does the policy distinguish between indicative and executable prices?
- Are execution statistics (fill rate, slippage on composite) published?
See a Public Routing Disclosure Example
NDD.broker publishes detailed order routing and execution policy documentation, including LP composition, priority logic, and conflict mitigation. This serves as a reference implementation of the concepts described above.
Educational content only. This is not financial advice. Always consult qualified professionals before making trading decisions.