Arguing which stop is "best" on a finished chart is fan fiction. The stop you would have used changes the size, the hold, and the lesson. Compare types on the same entries, or you are comparing three different trades and pretending they are one debate.
The Triple-Stop Drill does that. Twenty replay entries, three written stop rules, no blending. You learn what each rule costs you in skipped size, early exits, and rule-break temptation — not which one prints money on NSE.
Three rules, stated so they can be scored
Fixed stop. Invalidation is a frozen distance from entry: a point count or a percentage you wrote before the session. Example shape, not a prescription: "8 points" or "0.6% of entry." The number is arbitrary until you test it; the virtue is that it does not move when the candle looks scary.
Structure stop. Invalidation sits beyond a level you can point at before entry: swing low for a long, swing high for a short, range boundary, gap edge. If you cannot name the level in one sentence, you do not have a structure stop. You have a vibe.
Volatility stop. Invalidation is a multiple of a range measure you compute before the session — for example a multiple of yesterday's true range, or of the opening-drive range you defined at 9:30. The multiple is frozen. Recalculating the measure after a spike to "give it room" is not a volatility stop. It is a widened fixed stop in costume.
None of these is a trailing rule. Trailing is a later sequence of decisions; practise it separately in trailing stop practice.
Decision table: what actually changes
| Dimension | Fixed | Structure | Volatility |
|---|---|---|---|
| You decide | Distance, in advance | Which level is invalidation | Which range measure and which multiple |
| Size effect | Stable 1R arithmetic | Wide swings shrink size or force skips | Expands on wild opens, contracts in compression |
| Main lesson | Honesty under identical pain per trade | Whether you can define invalidation without hindsight | Whether you respect the day's range instead of a favourite number |
| Typical NSE failure | 9:15–10:00 noise tags a tight fixed stop that would have been fine after 11:00 | Stop so wide that 1R size becomes token, so you ignore it | Measure chosen after the spike; "ATR" as alibi |
| Emotional trap | Feeling the stop is "too close" on a familiar name | Moving the level after entry because "the real swing is the next one" | Tightening the multiple on a trade you love |
| Skip signal | Distance exceeds your max 1R budget for that product | No clean level, or level implies size you will not honour | Range measure missing or unstable (illiquid name) |
Hypothesis only: opening drives on liquid NSE cash names punish tight fixed stops more than midday ranges do. Your log confirms or rejects that for your execution. It is not a market law.
The Triple-Stop Drill (same entries, three overlays)
Do not run three simultaneous live-style positions. You will cheat the one that is winning. Run one execution rule per week, then overlay the other two in review.
Week A — execute Fixed only. Ten randomised hidden-date NSE sessions, one setup definition, frozen 1R from risk per trade choice. Log 20 entries (skips count as decisions if the setup appeared).
Week B — execute Structure only. Same setup definition, same 1R, same liquidity tier. New random dates. Another 20.
Week C — execute Volatility only. Same again.
Hold constant: session window (open versus midday), product (cash intraday versus delivery), and the rule that size = 1R ÷ stop distance, always rounded down.
Review overlay, after each week, on the same entry prices:
- Where would the other two stops have sat?
- Which would have been hit first on the subsequent candles?
- Which would have forced a skip because size rounded to zero lots?
You are not proving expectancy. You are collecting failure tags.
Log fields:
stop-type: fixed | structure | volatility
stop-distance (points)
computed-size
hit?: yes/no/not-reached
first-hit-bar (if hit)
urge-to-widen: yes/no
skip-reason: none | too-wide | no-level | illiquid
How to choose a type for the next block
Use criteria, not a guru preference.
Choose fixed for a block if your last logs show stop-distance shopping — you keep redrawing structure until the multiple looks pretty. A stupid honest number beats a clever moving one.
Choose structure if your last logs show fixed stops sitting inside obvious noise on the timeframe you actually trade. Then the work is level definition, which belongs with support and resistance practice, not with a tighter percentage.
Choose volatility if the same fixed number is laughably tight at the open and lazily wide at 12:30. Then the day's range should be an input. Write the measure at the start of the replay session, before candles after that timestamp are revealed.
If you cannot choose, you are not ready for three types. Run structure only until invalidation sentences become boring.
Indian-market interpretation
Cash session 9:15–15:30 is not one volatility climate. A fixed 0.4% stop that survives a sleepy midday Bank Nifty-correlated large cap can be random noise in the first half hour of the same name. That is not an argument for "no stops at the open." It is an argument for either a wider written fixed number in that window, a volatility measure that includes the open, or a skip rule: no entries until your clock says the window your plan allows.
Delivery and swing holds face overnight gaps. A structure stop below yesterday's low can be skipped in the gap. Replay on daily bars can show the gap; it will not show your broker's halt, your margin call, or the tax lot. Tag gap-through-stop when it happens. Do not silently assume a perfect fill at the stop price.
Illiquid names make all three types dishonest, because the print you use as "the stop" may not have been tradable. Keep the drill on a liquid tier; liquidity matters in practice for a reason.
Practice fills still ignore brokerage, taxes and slippage. A stop that is one tick beyond structure on a replay candle can be several ticks of live pain. If you later go live, that gap is yours to respect. It is not in the sim.
Failure modes that invalidate the comparison
- Executing Fixed but reviewing as if you had used Structure on the winners only.
- Changing 1R when the volatility stop widens, so you are no longer comparing stops — you are comparing risk appetites.
- Using index-future range to size a single-stock cash stop without writing that translation.
- Trailing any of the three mid-trade during this drill.
After 60 tagged entries (20 × 3), write one paragraph: which type produced fewer urge-to-widen tags and fewer too-wide skips for you? That is a temperament result. It is not a ranking of stop science.
Run the same entries under one stop law at a time. Join the waitlist.
Related reading
- Risk-Reward Practice: Designing Trades Before Entry
- Choosing Risk Per Trade in Practice
- Trailing Stops: The Hardest Thing to Practise
- Why Liquidity Matters When You Practise on NSE Names
Educational note
Educational practice content only. Not investment advice. No guaranteed returns. Practice ≠ live profitability.