Smart Order Router

Not always, but default to limit orders. Use limit orders for: planned entries, larger orders, illiquid instruments, any non-urgent situation. Use market orders for: true emergencies (stop-loss exit), extremely liquid instruments with tight spreads, when you absolutely must be filled now. A practical middle-ground: place a limit order slightly above the ask (for buys) or below the bid (for sells). This usually fills immediately like a market order but protects against sudden price spikes. You get speed with protection.

Compare the spread to the instrument's price as a percentage. Guidelines: Index futures (ES, NQ): a spread under ~0.02% is normal (often about 1 tick). If it is more than ~0.05%, be careful. Index/ETF options ATM (e.g., SPY): a spread under ~1% is acceptable - the most liquid names trade in pennies. OTM options: under ~3-5% is okay; more than ~5% requires patience. Single-stock options vary widely - check the historical average. If the spread is 2-3x normal, wait for it to tighten or use limit orders and be patient. Also compare the spread in dollars to your expected profit - if the spread cost is more than ~10% of your target profit, execution quality matters a lot.

Several reasons: 1) Queue position - other orders at same price were placed before yours; they fill first. Price may touch your level but not enough volume to reach your order in queue. 2) Time priority - at same price, earlier orders fill first. 3) Price moved too fast - touched and bounced before filling. 4) Your quantity was large - some filled, but price moved before complete fill. Solutions: place orders slightly more aggressive (1-2 ticks better than ideal), accept that some orders won't fill, have contingency plan (raise limit after timeout).

Illiquid options can have 5-10%+ bid-ask spreads. Example: option bid $0.45, ask $0.55 (about a 20% spread). If you market-buy at $0.55 and then market-sell at $0.45, you have lost $0.10, or about 18%, immediately! Even if the option is correctly priced at $0.50, you paid a ~10% premium to enter and will lose ~10% to exit. Round-trip cost: ~20%. The option has to move ~20% just to break even. Solution: never use market orders in illiquid options. Use limit orders at the mid ($0.50) and wait. If it doesn't fill, adjust slowly. Be prepared to wait - or not trade at all if the spread is too wide.

Start tracking: 1) For every trade, record: decision price (when you decided), order type used, execution price, slippage (execution - decision). 2) After a month, analyze: average slippage, which instruments are worst, which times are worst, market orders vs limits performance. 3) Identify patterns: maybe NQ has more slippage than ES, or afternoon trades are worse. 4) Take action: adjust timing, be more patient with limits, avoid the worst conditions. 5) Track improvement: is average slippage decreasing? Simple spreadsheet tracking can save you 0.5-1% on trades over time. That compounds significantly.

TWAP when: you have no view on volume patterns, market structure is unpredictable, equal spread of execution is fine, simplicity is valued. VWAP when: you want to match market's trading pattern, you have good volume predictions, benchmark is VWAP (institutional common), you want to hide in normal flow. Practical consideration: VWAP requires volume profile (historical or predicted). If you don't have this data or it's unreliable, TWAP is safer. For retail traders executing 5-20 contracts, the difference may be minimal. Focus on the basics (right order type, right timing) before worrying about TWAP vs VWAP.

Strategies to reduce impact: 1) Slice the order - 20 contracts in 4 slices of 5 contracts has less impact than 20 at once. 2) Time the slices - spread over 30-60 minutes, let order book refill between slices. 3) Use limit orders - avoid sweeping the book. 4) Execute during high volume - your order is smaller relative to market. 5) Consider iceberg orders if available - shows only portion of true size. 6) Be patient - the less urgent you are, the less impact you create. 7) Opposite side liquidity - if bid side is deep, selling has less impact. Read order book before execution. 8) Avoid trading right after large move - book may be thin.

Depends on urgency: Aggressive (want fast fill): place at or slightly across spread. Buying? Place at ask + 1-2 ticks. Ensures fill while protecting against spike. Moderate: place at spread midpoint. Often fills when market moves to you. Lower urgency. Passive: place at bid + 1 tick (for buys). Better price if filled, but may miss if market moves away. Very passive: place below current bid. Catches only dips. May never fill. Practical approach: start passive, set timer (e.g., 5 minutes), if not filled, move toward aggressive. This captures good fills when available and ensures execution when needed. Track fill rate at different aggressiveness levels to optimize.

Futures execution: generally more liquid, tighter spreads. Can be more aggressive with timing. Market orders acceptable for liquid contracts in urgent situations. Focus on timing (morning/closing better than lunch). Options execution: liquidity varies enormously. ATM weeklies liquid, far OTM monthly very illiquid. Always use limits, never market orders in illiquid strikes. Spread often your biggest cost. Be very patient with illiquid options - place limit at mid and wait. Consider liquidity when choosing strikes (sometimes better to use slightly less optimal strike if much more liquid). Multi-leg options: sequence matters, watch leg risk.

Partial fill scenarios and responses: Scenario 1: limit order partially fills, price moves away. Options: a) Accept partial, don't chase, b) Adjust limit to get more, c) Cancel remainder and reassess. Scenario 2: slicing order, early slices fill, later slice doesn't. Response: pause, reassess market, either be patient or adjust later slices' price. Scenario 3: multi-leg order, only some legs fill. Response: critical! Must complete hedge leg or close filled leg to avoid naked exposure. Best practices: Set alerts for partial fills. Have predefined rules (e.g., if < 50% fills in X time, do Y). For multi-leg, have contingency ready before executing. Accept that some orders won't complete as planned.

Framework for custom algo: 1) Define objective function: minimize IS, match VWAP, maximize fill rate, etc. 2) Model market: estimate volume curve (historical or predicted), estimate impact function (empirically), estimate spread patterns. 3) Optimize schedule: given model, find optimal trade schedule. For simple cases, analytical solutions exist. For complex, use numerical optimization. 4) Design adaptation rules: how to adjust when reality deviates from model. Set thresholds and actions. 5) Implement pricing: for each scheduled slice, set limit price logic (passive to aggressive based on fill progress). 6) Backtest: run on historical data, measure against benchmark. 7) Paper trade: test in live market without real orders. 8) Deploy gradually: start with small percentage of orders, scale up as validated. 9) Monitor and iterate: continuous improvement based on actual performance.

Order flow analysis for execution: 1) Imbalance indicators: compute bid-ask imbalance over rolling window. Strong buy imbalance suggests price will rise - execute buys more urgently. 2) Trade direction: estimate whether recent trades were buyer or seller initiated (trade at ask = buy, trade at bid = sell). Net buying = bullish pressure. 3) Large orders: detect large orders from jumps in cumulative volume or large prints. Large buy orders suggest institutional buying - may want to frontrun (carefully). 4) Refresh rate: how quickly does depth replenish after trade? Fast refresh = can be more aggressive. 5) Integration: create composite signal from these factors, use to adjust execution aggressiveness in real-time. Challenges: data intensive, requires real-time processing, signal may be noisy. Start with simple imbalance metric.

Portfolio rebalancing execution strategy: 1) Net positions first: if rebalancing requires buying A and selling B, check if any natural netting (e.g., A and B are correlated - can use pairs approach). 2) Prioritize by urgency: rank trades by risk. Risky positions or time-sensitive trades execute first. 3) Consider cross-impact: selling one stock may impact correlated stock. Coordinate execution of correlated trades. 4) Opportunistic vs scheduled: balance between capturing favorable prices when available vs systematic scheduled execution. 5) Track overall portfolio beta/exposure during execution to manage transition risk. 6) Completion time: have target completion time, be more aggressive as deadline approaches. 7) Cost attribution: track execution cost by trade to identify problematic securities. For retail, simplify: prioritize risk reduction first, use TWAP for larger legs, limit orders for all.

Execution algorithm evaluation framework: 1) Benchmark selection: IS vs arrival, VWAP, TWAP depending on algo's goal. 2) Statistical significance: with enough orders, calculate average performance with confidence intervals. 3) Decompose cost: separate IS into delay, impact, timing, spread components. Which component dominates? 4) Conditional analysis: how does algo perform in different market conditions (volatile vs calm, trending vs ranging, high vs low volume)? 5) Fill rate: does algo complete orders? What's the partial fill rate? 6) Risk metrics: what's the variance of IS? Maximum IS? Worst case scenarios. 7) A/B testing: if comparing two algos, randomly assign orders to each, compare with statistical tests. 8) Cost per unit: normalize by order size. 9) Trend analysis: is performance improving or degrading over time? Markets adapt, algos may need tuning.

Multi-asset execution complexity: 1) Correlation awareness: assets often correlated. Execution of one affects others. Model cross-asset impact. 2) Relative value preservation: for pairs/arb strategies, execute both legs in sync to maintain spread. Don't let one leg get far ahead. 3) Margin management: execution across assets affects margin. Sequence to stay within margin limits. 4) Currency considerations: if multi-currency, consider FX impact and timing. 5) Cross-venue: if assets trade on different venues (e.g., equities on NYSE/Nasdaq vs futures on CME Globex), coordinate timing with each market's hours. 6) Risk buckets: group trades by risk factor (equity beta, rates, volatility). Execute to manage overall portfolio risk during transition. 7) Completion synchronization: aim to complete all legs around same time to minimize period of partial implementation. Implementation: build execution management system that tracks all orders, aggregates positions, and coordinates across assets/venues.