CustomTradingPlatform .com
Log in Get started free
Home / Strategies

Strategies

Systematic Trading Strategies

Six rule-based strategies, each specified precisely enough to backtest, with the same method applied to every one.

TrendMoving Average CrossoverEnter when a fast average crosses a slow one, exit on the reverse cross.
Mean reversionRSI Mean ReversionEnter on an oversold reading, exit as the indicator normalises.
BreakoutRange BreakoutEnter on a close outside a defined range, stop back inside it.
MomentumMomentumHold the instruments with the strongest recent return, rebalance on a schedule.
TrendTrend FollowingEnter in the direction of a longer-term trend filter, trail the stop.
Mean reversionMean ReversionEnter on a statistically large deviation from a moving average, exit at the mean.

Not yet backtested

These six are specified as rules. None has a published backtest result on this site yet. Individual strategy pages, each with entry, exit, costs, expectancy, drawdown and a verdict, are the next thing to build here.

See how a strategy is tested.

What a strategy has to specify before it can be tested

Each of the six above is a name and a sentence. That is not yet a strategy — it is a family of them. Two people implementing "moving average crossover" from that sentence will produce different results, and neither will be wrong.

A strategy is testable once it fixes all five of these, and every gap left open is a place where a backtest quietly becomes a different strategy:

  1. Entry. Which bar, at which price. The close that triggered the signal is not available to trade at; the next open is.
  2. Exit. A fixed horizon measures the raw move. A stop and target measure a system, and mix the rule with the exit logic.
  3. Position size. Fixed lots make the widest stop the largest loss. Fixed fractional risk keeps the loss constant and varies the size — see the position size formula.
  4. Universe and period. Which instruments, which timeframe, which years. Choosing them after seeing results is how a backtest gets fitted to its sample.
  5. Parameters. The fast and slow lengths, the oversold threshold, the range window. Each is a choice, and trying many and keeping the best is overfitting.

Worth knowing

The strategies here are published as descriptions rather than as signals, and the distinction matters. A signal asks you to trust the sender. A rule asks you to check it — and a rule you can check is the only kind whose result can be reproduced.

Why the same method is applied to all six

Comparing strategies is only meaningful when the differences between them are the strategies. If one is tested with a stop and another without, one on a decade and another on two years, the comparison measures the test design rather than the rules.

So the same entry convention, the same cost treatment and the same reporting apply to every one. That makes the weaker ones look weaker than a flattering test would, which is the point.

What none of them will tell you

Whether it works for you. A strategy with positive expectancy still delivers long losing runs, and whether you survive one is decided by position size and by temperament rather than by the rule. The arithmetic of that is in how much to risk per trade and how drawdown affects recovery.