Strategy types overview
An overview of the common families of strategies: momentum, mean reversion, searching for cheap stocks, pairs trading and automated bots. Meant for understanding, not for copying.
These pages are deliberately built differently from most of what you'll find about trading strategies. They don't tell you which rules to follow. They tell you what kinds of strategies exist, what research has found about them over decades and where the weaknesses lie.
The reason is simple. A rule like “Buy when the price is above the average of the last 200 days” sounds like instructions. Without context, though, it's useless or even dangerous. What matters here is understanding why a family of strategies worked at all over long periods and in many markets, and why that often fades over time.
Each of the following pages covers one such family: the basic idea, a well known research result, the explanations for why the effect exists at all, and the reasons it's harder to use in practice than on paper.
One pattern runs through almost all of them: the strategies worked better in studies than in an individual investor's portfolio. Trading costs, taxes, slippage and competition from professionals eat up much of the measured excess return. That's no accident. In a market where many people are hunting for exactly these edges, little is left for the individual.
Studies of trading strategies almost always work with old prices and simplified cost assumptions. The measured excess return refers to broadly diversified portfolios of hundreds of stocks, often rebalanced only once a month or quarter. It can't simply be transferred to a single position of a private investor with little money.
Research also shows again and again that effects weaken after they're published. As soon as an effect appears in a study, money tries to exploit it. That money changes prices and makes the excess return smaller afterwards. A large study by McLean and Pontiff found that effects were on average more than half weaker after publication. This has been documented for several of the families covered here.
For you as a private investor this means: a strategy supported by research doesn't promise you any particular future return. It describes a pattern from the past with a plausible but not fully settled explanation. Everything in the lessons Backtesting and Overfitting applies to each of the following pages.
Summary
- These pages explain kinds of strategies, they don't recommend any.
- Study results apply to broadly diversified portfolios over long periods, not automatically to single positions.
- Many effects weaken once they're known and many people use them.
Did you get it?
What sets these pages apart from a trading guide?
They explain how a family of strategies works and what research says about it, but give no rules to copy.
Why do academically supported effects often weaken in practice?
Because money trying to exploit the effect changes prices and makes the future excess return smaller.
What kind of portfolio do most study results apply to?
For broadly diversified portfolios held for a long time, not for single positions with little money.
Related
- Pairs trading and statistical arbitrageStrategy Types
- Momentum and trend followingStrategy Types
- Why most people should invest, not tradeStage 1