Stop Loss Placement, Technical Structure Versus Arbitrary Percentages
A stop loss order exists to answer one specific question objectively, before emotion has a chance to interfere. At what price does the original reason for entering this trade become invalid? Many traders instead default to a round, arbitrary percentage below their entry, something like 5%, 10%, or a fixed dollar amount, without connecting that number to any specific feature of the chart or the thesis behind the trade. That produces a stop that is either too tight, getting hit by normal noise before the thesis has a chance to play out, or too loose, allowing far more damage than the setup actually justified.
Why Arbitrary Percentage Stops Underperform
A fixed percentage stop loss applies the same way regardless of the asset's typical volatility, the specific technical level the trade is based around, or current market conditions. A 5% stop might be far too tight for a highly volatile small-cap asset that regularly swings 8 to 10% within a single session on no particular news, guaranteeing the stop gets hit by ordinary noise long before any real trend has a chance to develop. The same 5% stop might be far too loose for a large-cap asset with typical daily ranges under 2%, letting a losing trade run much further than the setup's actual invalidation point would have justified.
Anchoring Stops to Market Structure Instead
A sturdier approach places the stop loss at the price level where the specific reason you entered the trade would be proven wrong, rather than at a round number chosen independently of the chart. Common structural anchor points include the low of the candle or swing that triggered entry, a prior support or resistance level the thesis depends on holding, or the boundary of a chart pattern, such as below a breakout level, if the trade was based on that breakout holding.
| Entry Reasoning | Structurally Sound Stop Placement |
|---|---|
| Breakout above resistance | Just below the broken resistance level (now expected to act as support) |
| Bounce off support | Just below the support level itself, allowing for minor wicks |
| Trend continuation entry | Below the most recent relevant swing low in the trend |
| Range trade at range low | Just below the established range boundary |
Reconciling Structural Stops With Position Sizing
A common objection to structure based stops is that the resulting stop distance sometimes turns out inconveniently large or small relative to a trader's preferred position size. The right response to this tension is not to move the stop to a more convenient distance, since that disconnects the stop from the actual invalidation point of the thesis. Instead, adjust position size, using the fixed fractional method to solve for the appropriate position size given the structurally correct stop distance, however large or small that distance turns out to be.
A Simple Test for Any Stop Loss Before Entry
Before finalizing any trade, ask yourself this. If price reaches this stop level, does that specifically mean my original reason for entering was wrong? If the honest answer is no, if the stop is just a comfortable distance away with no particular structural significance, the placement should be reconsidered. A stop that isn't tied to thesis invalidation is really just arbitrary no matter how confidently it was chosen, and arbitrary stops tend to produce inconsistent, hard to analyze results across a sample of trades.
- Identify the specific price level that would invalidate your trade thesis before entering, independent of any position sizing consideration.
- Use the Trade Planner to work out an appropriate position size for that stop distance, rather than adjusting the stop to fit a size you already had in mind.
- Revisit and refine your structural stop placement logic from time to time by reviewing past trades in a trading journal to see which stop types tended to get hit by noise versus genuine invalidation.
Pairing a structurally anchored stop loss with the Trade Planner means both halves of the equation, where the stop belongs and how large the position should be given that stop, each get worked out on their own correct logic, rather than compromising one to fit the other.