A Practical Taxonomy of Systematic Trading Strategies
Summary
The document surveys several ways to group systematic trading strategies. Strategy families include trend and reversal approaches, price or technical methods, fundamental, yield, growth, and quality signals, as well as arbitrage, market making, sentiment based trading, and event driven research. It also mentions statistical and machine learning methods, including cointegration and regression, as ways to generate signals.
These categories overlap rather than form one official taxonomy. Strategies can also be classified by implementation frequency or by the data they use, such as market prices, fundamentals, and sentiment. The examples help an analyst build a broad map of the field, but the document does not define the categories consistently or compare their performance. It also notes that portfolio construction, position limits, and risk monitoring are separate considerations that a strategy label alone does not capture.
Key ideas
- Systematic trading includes strategy families beyond momentum and mean reversion, such as arbitrage, market making, and event trading.
- Signal methods can use statistical models or machine learning as well as price and fundamental analysis.
- Strategy taxonomies are not official, and the listed categories can overlap.
- Trading frequency and input data provide additional ways to classify a strategy.
- Portfolio construction and risk controls remain important beyond the signal category.
Tags
Full text
# Categories of systematic trading strategies? # Categories of systematic trading strategies? What are the main categories of systematic trading strategies (e.g. momentum, mean reversion), as might be considered by an index or fund-of-fund analyst? Are there any common sub-strategies? ## Answer by Meh (score 12, accepted) https://quant.stackexchange.com/a/301 There are other strategy types not covered by mean-reversion/trend following: - arbitrage - keep correlated assets close in price (SPX index versus the 500 stocks contained in it, or Gold trading in London versus Gold trading in New York) - market making - buy on bid, sell on ask, gain the spread - liquidity rebate - some venus pay you for putting limit orders in the book. Put in a limit order to buy, when it's hit try to sell at the same price that you bought at (or better) and gain the rebate. Works best on high volume, low price assets. - predatory trading - seek big hidden liquidity in the market and front-run it - behavioral trading - quantify market sentiment and trade on it (analyze tweets, determine global/regional mood and use known psychological theories to predict the effect on market behavior) - event trading - analyze news (electronic, paper, blogs, twits) and predict market impact of new relevant facts (litigation, new products, new management, ...) ## Answer by chrisaycock (score 9) https://quant.stackexchange.com/a/294 There is no official taxonomy of quant trading models. After all, "valuations" are inherently subjective, no matter how much math we put behind them. But there are some industry-standard terms that might be helpful. Inside the Black Box has the following break-down: - Price - Trend - Reversal - Fundamental - Yield - Growth - Quality It's also possible to break-down by implementation: And these don't even get into portfolio construction, position limits, risk monitoring, etc. As for what works, keep this maxim in mind: > Bulls make money, bears make money, but pigs get slaughtered. And lastly, comparing chartists to quants is like comparing astrologists to astronomers. ## Answer by Ishan Shah (score 0) https://quant.stackexchange.com/a/44892 The categories for systematic trading strategies are - Chart Patterns - Technical Indicators - Quant Trading Strategies - Machine Learning/Artificial Intelligence The description of these categories is below. - Quant Trading Strategies: Advanced tools such as statistics are used to generate the trading signals. Example, statistical arbitrage using cointegration - Machine Learning Strategies: Different machine learning algorithms are employed such as very basic linear regression to more advanced LSTM (neural network) With each of these main categories, there are different styles based on the frequency of trades such as low-frequency trading (LFT), medium frequency trading (MFT) and high-frequency trading (HFT) There can be further subcategories categories based on the data used such as - Price data (OHLCV) - Fundamental Data - Sentiments Data
Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.