Country: Russia
You are in the Russia edition. Type your country to switch.
Nigeria
Europe
Loading every country…
No country matches. Check the spelling, or choose "Any country" for the Global edition.
You are learning in the Russia edition, with examples in roubles and from Russia.
Your learning profile
Advanced path · Free course
Algorithmic trading
Educational only: what algorithms do in markets, from an idea to rules, data hygiene, backtesting properly, costs and capacity, execution, risk controls and how trading algorithms are regulated. No code that connects to a broker and no performance figures.
- Advanced level
- 8 lessons
- 2 hours of study
- 1 final test
- $10 certificate
- Russia edition
Education, not advice: this course explains how products and rules work. It does not recommend any product, trade or provider.
Before you enrol
- To doPass the final test of Risk & psychology: Intermediate
- To doPass the final test of Trading foundations: risk, costs and process
- To doAnswer the suitability self-check
- To doConfirm you are 18 or over
What you will learn
- By the end you can tell execution algorithms from algorithms that decide what to trade.
- By the end you can turn a loose trading idea into exact rules a computer could follow.
- By the end you can explain point-in-time data and why revised figures flatter a backtest.
- By the end you can separate in-sample and out-of-sample data and explain why the final test must be used only once.
- By the end you can list every cost that separates a backtest from live trading, including your own market impact.
- By the end you can explain how queue position decides whether a limit order fills, and how tick size affects queues.
- By the end you can list the main pre-trade and real-time controls that firms place around trading algorithms.
- By the end you can describe the main duties MiFID II places on firms that trade algorithmically in the EU, and the UK's similar regime.
Skills you will practise
Syllabus
-
What algorithms do in markets: execution, market making, trend, arbitrage and high frequencyMost orders in large markets are now sent by computer programs. Some only decide how to buy or sell what a person has already chosen; others decide w…
11 min · 5 steps
- What an algorithm is in trading
- Execution algorithms: TWAP, VWAP and shortfall
- Market making and high-frequency trading
- Trend, reversion and arbitrage
- What algorithms cannot do
- Lesson quiz
-
From idea to rules: a hypothesis, data and a test that could prove you wrongAn algorithm needs rules precise enough for a computer: what to trade, when to enter, how much and when to leave. Before writing them, a careful rese…
11 min · 5 steps
- An idea is not a rule
- Why should it work? Sources of edge
- Writing the rules
- A test that could prove you wrong
- Keeping a research log
- Lesson quiz
-
Data hygiene for backtests: point-in-time data, survivorship, look-ahead and corporate actionsA backtest is only as honest as its data. Prices that were adjusted later, companies that vanished from the list, figures that were published weeks a…
11 min · 5 steps
- A backtest is only as honest as its data
- Point-in-time data and look-ahead bias
- Survivorship: testing only the winners
- Corporate actions
- A data checklist before any test
- Lesson quiz
-
Backtesting properly: out-of-sample tests, walk-forward, multiple testing and overfittingEvery strategy can be tuned until its history looks excellent. Honest backtesting separates the data used to build a strategy from the data used to j…
11 min · 5 steps
- Fitting the past is easy
- Keeping data aside
- Walk-forward testing
- Multiple testing: the luck of the best
- Warning signs of an overfitted backtest
- Lesson quiz
-
Costs and capacity: slippage, market impact and why paper results shrinkA backtest trades at prices that were on the screen. A live strategy has to cross the spread, wait in queues and push prices as it trades. Those cost…
11 min · 5 steps
- Why paper results shrink
- Every cost, named
- Market impact and the square-root rule
- Turnover turns small costs into large drags
- Capacity and measuring real costs
- Lesson quiz
-
Execution and microstructure for algorithms: the order book, queues and latencyAlgorithms live inside the order book. Where an order sits in the queue, how fast it reaches the exchange and which venue it is sent to all change th…
11 min · 5 steps
- Algorithms live inside the order book
- Queues and tick sizes
- Latency and the race for speed
- Adverse selection: who fills your order
- Venues, routing and order types
- Lesson quiz
-
Risk controls for algorithms: kill switches and limits, Knight Capital (2012) and the Flash Crash (2010)An algorithm that misbehaves can lose more in minutes than a person could in years. Firms protect themselves with pre-trade limits, price checks and …
12 min · 5 steps
- Losses at machine speed
- The control toolkit
- Knight Capital, 1 August 2012
- The Flash Crash, 6 May 2010
- Market-wide safeguards and lessons
- Lesson quiz
-
How trading algorithms are regulated: MiFID II, the US market-access rule and retail algorithmsRegulators treat algorithms as part of a firm's conduct, not as a technical detail. In the EU and UK, firms that trade algorithmically must test thei…
11 min · 5 steps
- Software does not carry the responsibility
- The EU and UK: MiFID II
- The United States
- Market abuse by algorithm
- Retail algorithms and the rules where you live
- Lesson quiz
Final test
A fresh set of questions drawn from every lesson of this course, timed, with the options in a new order every time. You need 85% to pass, and passing earns your certificate.