Cost Curves

Market impact cost typically follows a concave power-law function of trade size. The cost increases with order size, but at a decreasing rate.

• Formula: Cost ≈ σ * (Q/V)^α, where σ is volatility, Q is order size, V is average daily volume, and α is the exponent (typically 0.5–0.8).
• Square-Root Model: The most empirically validated form where α = 0.5, implying that doubling the order size increases impact by roughly 41%.
• Linear vs. Concave: Concavity reflects the market's ability to absorb larger orders by attracting new liquidity providers over time.

A cost curve decomposes total market impact into two distinct components with different decay characteristics.

• Temporary Impact: The transitory price pressure caused by order book imbalance. This component reverts after the trade is complete as liquidity replenishes.
• Permanent Impact: The irreversible price change reflecting the information content of the trade. The market interprets aggressive buying as a potential positive signal.
• Decay Function: Temporary impact decays exponentially over a horizon of minutes to hours, while permanent impact persists indefinitely.

Transaction costs scale linearly with asset volatility (σ). A cost curve must be recalibrated dynamically as market conditions change.

• Intraday Patterns: Costs are higher during the opening and closing auctions when volatility spikes.
• Regime Sensitivity: A curve calibrated in a low-volatility regime will underestimate costs during a VIX spike.
• Normalization: Costs are often quoted in basis points of volatility (e.g., 10 bps per daily σ) to allow cross-asset comparison.

The cost curve is parameterized by the participation rate—the fraction of market volume the algorithm consumes. Higher urgency shifts the curve upward.

• Low Urgency (1–3% POV): Minimal impact, but high opportunity cost and delay risk.
• Medium Urgency (5–10% POV): The efficient frontier sweet spot balancing impact against timing risk.
• High Urgency (20%+ POV): Steep impact costs; used only when alpha decay is extremely rapid.
• The curve is actually a surface when plotted against both size and participation rate.

A single cost curve is insufficient for fragmented markets. Curves must be adjusted for venue-specific liquidity and order book depth.

• Order Book Depth: The cumulative volume available within N basis points of the mid-price directly determines the temporary impact for small orders.
• Resilience: The speed at which the order book refills after a liquidity-taking event. Low-resilience assets have steeper curves.
• Spread Cost Integration: The effective half-spread is added as a fixed cost component, independent of order size, creating a non-zero intercept on the curve.

Cost curves are estimated from historical tick data using proprietary transaction cost analysis (TCA) databases.

• Proprietary Data: Broker-dealers build curves from their own executed order flow, capturing actual fill prices against arrival price benchmarks.
• ACD Models: Autoregressive Conditional Duration models capture the irregular spacing of trades and volume events.
• Vendor Models: Third-party providers like Bloomberg and ITG offer pre-trade cost estimates using anonymized, aggregated industry data.
• Challenge: Selection bias—executed orders may differ systematically from cancelled orders, skewing the curve.

The gold standard benchmark for measuring total trading cost, capturing the difference between the decision price and final execution price.

• Components: Explicit costs (commissions, fees) + implicit costs (market impact, delay)
• Formula: (Paper Return) - (Actual Portfolio Return)
• Use case: Post-trade analysis for institutional portfolio managers to evaluate broker performance
• Key insight: Often reveals that implicit costs dwarf explicit costs for large orders

The adverse price movement directly attributable to your own order consuming liquidity and signaling information to the market.

• Temporary impact: Reversible price pressure from order book imbalance
• Permanent impact: Irreversible price change reflecting information leakage
• Square-root law: Impact scales roughly with the square root of order size relative to volume
• Mitigation: Slicing orders using TWAP, VWAP, or liquidity-seeking algorithms

Two foundational execution benchmarks that slice parent orders differently to minimize market footprint.

• VWAP (Volume Weighted Average Price): Concentrates executions during high-volume periods; matches the market's natural rhythm
• TWAP (Time Weighted Average Price): Slices orders evenly across time regardless of volume; simpler but less adaptive
• Selection logic: Use VWAP when volume patterns are predictable; TWAP when they are not
• Limitation: Both are easily gamed by predatory algorithms if schedule is deterministic

The cost of trading against a counterparty with superior information, resulting in permanent unfavorable price movement immediately after execution.

• Mechanism: Informed traders pick off resting limit orders before price discovery occurs
• Measurement: Post-trade drift — if price consistently moves against you after fills, you're being adversely selected
• PIN (Probability of Informed Trading): Microstructure model estimating the fraction of informed order flow
• Defense: Use midpoint peg orders in dark pools and avoid posting large visible limit orders in toxic markets

A systematic method for randomly allocating parent orders across a pre-approved set of broker algorithms and dynamically re-weighting based on measured performance.

• Purpose: Eliminates trader bias and prevents brokers from reverse-engineering your execution logic
• Feedback loop: Post-trade TCA scores feed back into allocation weights
• Metrics tracked: Implementation shortfall, VWAP slippage, fill rate, latency
• Governance: Essential for demonstrating best execution compliance to regulators

Automated system that scans fragmented liquidity across lit exchanges, dark pools, and systematic internalizers to achieve optimal execution.

• Decision logic: Balances price, fill probability, latency, and venue fees
• Regulation: Required under MiFID II and Reg NMS for best execution
• Maker-taker dynamics: Routes may prefer venues offering rebates for adding liquidity
• Challenge: Latency arbitrage — slower routers get picked off by HFT firms monitoring order flow