Smart Order Router

Not always, but default to limit orders. Use limit orders for planned entries, larger orders, illiquid instruments and any non-urgent situation. Use market orders for true emergencies (a stop-loss exit), extremely liquid instruments with tight spreads, or when you absolutely must be filled now. A practical middle ground is a marketable limit slightly above the ask (for buys) or below the bid (for sells): it usually fills immediately like a market order but protects against a sudden spike. On fragmented UK shares, a smart router should also check Cboe/Aquis/SIs for a better touch before crossing.

Compare the spread to the price as a percentage. Guidelines: the FTSE 100 future spread (about 1 point, ~0.01%) is normal; if it is several times that, be careful. FTSE 100 index options ATM: under ~1-2% is acceptable; OTM are wider. Single-stock UK options can show 15-25% spreads and need real patience. Compare the spread in pounds to your expected profit - if it is more than ~10% of your target, execution quality matters a lot. On cash equities, remember the best spread may sit on Cboe or Aquis rather than the LSE.

Several reasons: 1) Queue position - other orders at the same price were placed before yours and fill first; the price may touch your level without enough volume reaching you. 2) Time priority - at the same price, earlier orders fill first. 3) The price moved too fast - it touched and bounced before filling. 4) Your quantity was large - some filled before the price moved. In the UK there is a further twist: your size may have filled on one venue while the price moved on another, so a smart router checks all venues. Solutions: be slightly more aggressive (1-2 ticks better), accept some misses, and have a contingency (raise the limit after a timeout).

Illiquid single-stock options can have 15-25% bid-ask spreads. Example: bid GBP 0.45, ask GBP 0.55 (about 20% of mid). Market-buy at GBP 0.55 then market-sell at GBP 0.45 and you have lost GBP 0.10 - roughly 20% immediately. Even if the option is fairly priced at GBP 0.50, you paid ~10% to enter and lose ~10% to exit: a round trip of ~20%. Solution: never use market orders in illiquid options; use limits at mid-price and wait. Many UK retail traders instead take liquid directional exposure via FTSE 100 index options, CFDs or spread bets.

Start tracking: 1) For every trade record the decision price (when you decided), the order type, the execution price, the venue you filled on, and the slippage. 2) After a month analyse average slippage, which instruments are worst, which times are worst, and market vs limit performance. 3) Identify patterns - perhaps the FTSE 250 has more slippage than the FTSE 100, or afternoon trades are worse, or LSE-only routing underperforms smart routing. 4) Act: adjust timing, be more patient with limits, route across venues, avoid the worst conditions. 5) Track the improvement. Simple spreadsheet tracking can save you 0.5-1% on trades over time, which compounds significantly.

Use TWAP when you have no view on volume patterns, the structure is unpredictable, an equal spread of execution is fine, and simplicity is valued. Use VWAP when you want to match the market's pattern, you have good volume predictions, the benchmark is VWAP (common institutionally), and you want to hide in normal flow. Practical point: VWAP needs a volume profile (historical or predicted), and in the UK that profile must be aggregated across all venues. If you lack reliable data, TWAP is safer. For retail sizes (a few contracts) the difference is small - focus on the basics (right order type, right timing, right venue) first.

Strategies: 1) Slice the order - smaller pieces have less impact. 2) Time the slices over 30-60 minutes so the book refills between them. 3) Use limit orders to avoid sweeping the book. 4) Execute during high volume so your order is smaller relative to the market. 5) Use iceberg orders to show only part of the size. 6) Be patient - less urgency means less impact. 7) Read the consolidated book and lean on the deeper side. 8) Avoid trading right after a large move when the book is thin. Uniquely useful in the UK: route part of the size to dark venues (Turquoise Plato, Cboe dark) or periodic auctions to cut your visible footprint.

It depends on urgency. Aggressive (want a fast fill): place at or slightly across the spread - for a buy, the ask + 1-2 ticks - which ensures a fill while protecting against a spike. Moderate: place at the midpoint, which often fills when the market moves to you. Passive: place at the bid + 1 tick (for buys) for a better price if filled, but you may miss it. Very passive: below the current bid, catching only dips - may never fill. Practical approach: start passive, set a timer (say 5 minutes), and if unfilled move toward aggressive. On UK shares, post to the venue showing the best touch, and track your fill rate at different aggressiveness levels.

Futures execution: the FTSE 100 future is very liquid with tight spreads, so you can be more aggressive on timing, and market orders are acceptable in urgent situations. Focus on timing (the open and the close beat the midday lull). Options execution: liquidity varies enormously - FTSE 100 index options are reasonable, but single-stock UK options are thin. Always use limits, never market orders in illiquid strikes; the spread is often your biggest cost, so be patient and place at mid. Consider liquidity when choosing strikes. For multi-leg options, sequence matters - watch leg risk. Many retail traders take directional exposure via CFDs or spread bets instead of thin single-stock options.

Scenario 1: a limit order partially fills and the price moves away. Options: accept the partial and do not chase; adjust the limit to get more; or cancel the remainder and reassess. Scenario 2: while slicing, early slices fill but a later one does not - pause, reassess, and either be patient or adjust the later slices' price. Scenario 3: a multi-leg order where only some legs fill - this is critical: complete the hedge leg or close the filled leg to avoid naked exposure. Best practices: set alerts for partial fills, have predefined rules (e.g., if less than 50% fills in X time, do Y), re-route residual size to an alternate venue, and accept that some orders will not complete as planned.

Framework: 1) Define the objective function (minimise IS, match VWAP, maximise fill rate). 2) Model the market - estimate the volume curve, the impact function (empirically) and spread patterns. 3) Optimise the schedule given the model (analytical for simple cases, numerical for complex). 4) Design adaptation rules with thresholds and actions for when reality deviates. 5) Implement pricing - limit-price logic per slice (passive to aggressive by fill progress). 6) Backtest against the benchmark. 7) Paper-trade in the live market without real orders. 8) Deploy gradually, starting with a small percentage of orders. 9) Monitor and iterate. In the UK, add a venue-routing layer: split each child order to the best-priced lit venue and sweep Systematic Internalisers for price improvement.

Order-flow analysis: 1) Imbalance indicators - compute bid-ask imbalance over a rolling window on the consolidated book; a strong buy imbalance suggests a rise, so execute buys more urgently. 2) Trade direction - estimate buyer- vs seller-initiated trades (a trade at the ask is a buy, at the bid a sell); net buying is bullish pressure. 3) Large orders - detect them from jumps in cumulative volume or large prints. 4) Refresh rate - how quickly depth replenishes after a trade; fast refresh means you can be more aggressive. 5) Integration - build a composite signal and use it to adjust aggressiveness in real time, and sweep SIs when they show price improvement. Challenges: it is data-intensive, needs real-time processing, and the signal can be noisy; start with a simple imbalance metric.

Portfolio rebalancing execution: 1) Net positions first - if you must buy A and sell B, check for natural netting (correlated names can use a pairs approach). 2) Prioritise by urgency - riskier or time-sensitive trades first. 3) Consider cross-impact - selling one stock can move a correlated one, so coordinate. 4) Balance opportunistic vs scheduled execution. 5) Track overall portfolio beta/exposure during the transition. 6) Have a target completion time and be more aggressive as the deadline approaches - the LSE closing auction is ideal for the final transition. 7) Attribute cost by trade to find problem securities. For retail, simplify: reduce risk first, use TWAP for larger legs, limit orders for all, and route across venues.

Evaluation framework: 1) Benchmark selection - IS vs arrival, VWAP or TWAP depending on the algo's goal. 2) Statistical significance - with enough orders, compute average performance with confidence intervals. 3) Decompose cost - split IS into delay, impact, timing and spread; which dominates? 4) Conditional analysis - how does it perform in different conditions (volatile vs calm, trending vs ranging, high vs low volume) and on different venues? 5) Fill rate - does it complete orders, and what is the partial-fill rate? 6) Risk metrics - variance of IS, maximum IS, worst cases. 7) A/B testing - randomly assign orders to two algos and compare with statistical tests. 8) Cost per unit - normalise by size. 9) Trend analysis - is performance improving or degrading as markets adapt? In the UK, always add a venue analysis: did smart routing beat a single-venue route?

Multi-asset execution complexity: 1) Correlation awareness - assets are often correlated, so model cross-asset impact. 2) Relative-value preservation - for pairs/arb, execute both legs in sync to hold the spread. 3) Margin management - sequence execution to stay within margin limits. 4) Currency considerations - for non-GBP lines, account for FX impact and timing. 5) Cross-venue and cross-market - if assets sit on different venues (LSE equity vs ICE futures vs LME metals), coordinate with each venue's hours. 6) Risk buckets - group trades by risk factor (equity beta, rates, volatility) and execute to manage overall portfolio risk during the transition. 7) Completion synchronisation - aim to finish all legs around the same time to minimise partial implementation. Build an execution management system that tracks all orders, aggregates positions and coordinates across assets and venues.