Algorithmic Trading and DMA
Execution algorithms and direct market access, from the desk.
Summary
Johnson, an execution specialist, wrote a technical reference on how orders are actually executed in electronic markets, covering market microstructure, execution algorithms and direct market access.
The microstructure sections explain how modern markets work: order types, the limit order book, matching rules and priority, market fragmentation across venues, the different market models including quote-driven and order-driven markets, and the differences across asset classes and geographies.
The cost framework is the analytical foundation. Johnson decomposes execution cost into explicit components, meaning commissions, fees and taxes, and implicit components, meaning bid-ask spread, market impact from the order's own effect on price, timing risk from price movement during execution, and opportunity cost from unexecuted quantity. He explains implementation shortfall, which measures the difference between the price when the decision was made and the price actually achieved, and argues it is the correct benchmark because it captures everything.
The algorithm sections cover each family and what it optimises. Time-weighted and volume-weighted average price algorithms spread an order over a period to match a benchmark, which controls tracking error rather than cost. Percentage of volume algorithms participate at a set rate. Implementation shortfall algorithms trade off market impact against timing risk explicitly, trading faster when volatility is high. Liquidity-seeking algorithms search across venues including dark pools. He covers the parameters each exposes and how to select among them given order size, urgency and the security's liquidity.
The book is technical and written for practitioners. It is dated in market specifics since 2010, particularly on dark pools, on venue structure and on the regulatory framework in both the United States and Europe. The cost framework and the algorithm logic have not changed.
Key takeaways 7
- Execution cost decomposes into spread, market impact, timing risk and opportunity cost, and reducing one usually increases another.
- Implementation shortfall measures the difference between the decision price and the achieved price, which captures every cost component.
- Volume-weighted average price algorithms control tracking error against a benchmark rather than minimising cost.
- Trading faster reduces timing risk and increases market impact, so the correct speed depends on volatility and on order size relative to volume.
- Market impact rises with order size relative to average volume, which is the binding constraint on strategy capacity.
- Fragmentation across venues means the best price may exist somewhere the order does not reach without a liquidity-seeking algorithm.
- Algorithm selection should follow from order urgency, size and the security's liquidity rather than from convention.
Who should read it
Traders and portfolio managers who need to understand what execution actually costs and how algorithm choice changes it.
Similar books 6
Chosen by shared topics, the entries that hold them, and the people who recommended both.
Plain-language description of what quant trading systems do.
How a trader uncovered high-frequency front-running and built the IEX exchange.
Recommended by: Mohnish Pabrai, Nat Eliason
A trader in his parents' London house, and his link to the 2010 flash crash.
GameStop and Citadel Securities' order flow; dramatised.
Why standard machine learning fails on market data, and how to avoid false backtests.
Where it sits in the reference 2
Topics