Open course in algorithmic trading
Algorithmic Trading: An Open Introduction
What is algo trading? A plain account of what an algorithm decides, the main families of strategy, the costs that eat returns, and what it takes to test an idea honestly.
What is algo trading? It means letting a program make some or all of the trading decisions under rules written in advance. Most trading volume on major electronic exchanges is now placed by algorithms, much of it by firms executing large orders for clients rather than by speculators.
What is algo trading, in one paragraph?
An algorithm takes data in, usually prices and sometimes much more, applies fixed rules, and sends orders out. The rules might decide what to buy, when to sell, how large a position to hold, or simply how to split a large order so that it moves the price less. A person still chooses the rules, and is still responsible when they fail.
Why do most beginner algorithms lose money?
Three reasons account for most of it. Costs: a strategy that trades often pays the spread and commission every time. Overfitting: a rule tuned to past data learns its noise. And unrealistic testing: a backtest that assumes every order fills at the last price. This course explains each one in plain terms.
Where does this course lead?
It is the free starting point for the algorithmic trading track. Algorithmic Trading Foundations is the next step, then the practitioner courses and the Certificate in Algorithmic Trading.
What you will be able to do
- Explain what algorithmic trading is and how it differs from discretionary trading.
- Name the main strategy families and the time horizons they work on.
- List the costs and risks that most beginner strategies ignore.
- Describe what a fair test of a trading idea requires.
Syllabus, module by module
What an algorithm decides
What to trade, when, how much and how to place the order; and which of these the algorithm controls.
Who trades with algorithms
Execution desks, market makers, systematic funds and individual traders, and what each is trying to do.
Strategy families
Trend following, mean reversion, statistical arbitrage, market making and execution, by horizon.
Costs and risks
Commissions, spreads, slippage, data, infrastructure, and the risk of a bug.
Testing an idea
Backtests, overfitting, out-of-sample data and paper trading, and where to go next.
How the certificate is earned
Self-check questions at the end of each lesson. The course is not assessed. Completing every lesson earns a Certificate of attendance, which states that the course is not assessed. Read how certification works.
Questions applicants ask
Is algorithmic trading profitable?
For some firms, yes, over long periods. For most individuals testing simple public strategies, costs and overfitting remove the edge. The course shows why, so you can judge claims for yourself.
Is this course really free?
Yes. There is no charge and no payment details are asked for. You receive a certificate of attendance on completion.
Is algorithmic trading legal?
Yes, in the major markets, subject to the same rules as any trading. Some practices, such as spoofing, are illegal whether a person or an algorithm does them.