Reference entry
Algorithmic Trading
Algorithmic trading is the use of computer programs to place, slice and manage orders according to predefined rules. In institutional asset management it is above all an execution discipline: it turns a target portfolio into trades at the lowest achievable cost and market impact.
What algorithmic trading is - and what it is not
The term covers two different things. In the narrow, institutional sense, algorithmic trading is the execution layer: a portfolio manager or model decides to buy 50,000 shares, and an algorithm decides how to work that order across the trading day. In the broader sense it also includes fully automated strategies that generate their own signals.
Algorithmic trading is not the same as high-frequency trading. High-frequency firms hold positions for milliseconds to minutes and compete on latency. Institutional execution algorithms work orders over minutes to a full session and compete on cost. A systematic investment process decides what to hold; algorithmic trading decides how the resulting orders reach the market.
Common execution algorithms
VWAP-to-close: how SINTRO executes a rebalance
SINTRO's strategies take their decisions on closing prices, so execution is aimed at the close rather than at the whole session. An order is worked along the U-shaped intraday volume profile, back-loaded towards the end of the day, with a large share - around 40% - placed in the closing auction, where liquidity is deepest and the fill matches the official closing price.
Quality is measured from both sides: the average fill against the market VWAP and against the closing price, in basis points. The live illustration in the engine section of the SINTRO homepage shows this schedule for a simulated order.
Measuring execution quality
- 01
Implementation shortfall
The difference between the paper portfolio at the decision price and the real portfolio after all fills, fees and missed trades.
- 02
Slippage versus benchmark
Average fill compared with arrival price, VWAP or close - the benchmark must match the strategy's own decision price.
- 03
Market impact
How far the order itself moved the price. It grows with order size relative to volume and with urgency.
- 04
Transaction cost analysis (TCA)
A post-trade review of every order that feeds back into the choice of algorithm, timing and venue.
Risk controls in algorithmic trading
Automation removes manual error but can multiply a single mistake. Controls are therefore enforced by the system before an order leaves it.
- 01
Pre-trade limits
Maximum order size, notional, price collars and position limits per instrument and portfolio.
- 02
Kill switch
Immediate cancellation of all working orders when limits, connectivity or data quality fail.
- 03
Reconciliation
Positions, cash and fills are reconciled against the broker and the model's target after every session.
- 04
Testing and governance
Algorithms are tested before deployment and changes are documented, in line with MiFID II requirements for firms using algorithmic trading.
Frequently asked questions
What is algorithmic trading?
Algorithmic trading is the use of computer programs to place and manage orders according to predefined rules for timing, price and quantity. Institutional investors use it mainly to execute large orders at low cost and with minimal market impact.
What is the difference between VWAP and TWAP?
VWAP slices an order in proportion to the expected trading volume through the day, so it trades more when the market is busiest. TWAP spreads the order evenly over time regardless of volume. VWAP usually achieves lower impact in liquid markets; TWAP is simpler and more predictable.
Why trade in the closing auction?
The closing auction concentrates a large share of the day's liquidity at a single price, which is also the official close used for fund valuation and index calculation. Strategies that decide on closing prices reduce tracking error by trading there.
Is algorithmic trading the same as high-frequency trading?
No. High-frequency trading is a subset that holds positions for very short periods and competes on speed. Institutional execution algorithms work orders over hours and compete on cost and impact.
Is algorithmic trading regulated?
Yes. In the European Union, MiFID II and its regulatory technical standards set requirements for firms that engage in algorithmic trading, including testing, risk controls, kill functionality and record keeping.
Related reading
- Systematic-Traded Funds (STFs)Definition, structure and origin of the STF investment sub-class.
- Systematic InvestingRule-based portfolio decisions under an explicit risk budget.
- Quantitative InvestingFactor models, statistical methods and quantitative portfolio construction.
- Quantitative StrategiesMomentum, mean reversion, carry, low volatility and regime models compared.
- Systematic TradingStrategy families, system architecture and the pitfalls of live models.
- STFs vs. Hedge FundsHow Systematic-Traded Funds compare with quantitative hedge funds.
- Institutional Asset ManagementSystematic mandates for pension funds, insurers, foundations and family offices.