Skip to content

3. SMT, PSP & Strength Switch

7 min read

Maintained by Nix Young

This chapter treats SMT, Precision Swing Points and Strength Switches as related confirmation tools for reading correlated assets, swing formation and reversal timing. It separates one-stage and two-stage forms, explains the sequential SMT definition, and places each tool within the wider context of key levels and draw liquidity.

Questions this chapter answers
What is SMT used for in this framework?
SMT is assessed at a relevant high or low and is filtered by context, including the direction of draw liquidity and the quality of the surrounding structure. Read this section
What is a Precision Swing Point?
A PSP combines a swing point with SMT. The chapter ranks swing, PSP and two-stage PSP structures as progressively stronger confirmation. Read this section
What is a Strength Switch?
A Strength Switch is a change in the relative leading and lagging relationship between correlated assets, used here as a re-synchronisation and timing confirmation. Read this section
What does SSMT mean here?
SSMT is defined as sequential SMT against the immediately previous high or low on the timeframe being used, rather than against an older candle. Read this section
Community insights
  • There's a form of SMT at every single retracement and reversal in the market. — Garrett — "Precision Swing Point", 12:30
  • Every single retracement and every single reversal carries some form of SMT. — Garrett — "Precision Swing Point", 22:15
  • A "sequential SMT" is simply an SMT between quarters — for example between the Q2 and Q3 90-minute candles. "Logical liquidity target" is just another name for equilibrium: the 50% retracement of the higher-timeframe expansion being traded. — Afyz — trade recap, 5.5R YM short, 11:34
  • A "hidden SMT" is a divergence where the compared highs/lows on two correlated assets weren't actually printed on the same day — one asset's relevant high sits on an earlier day than the other's, making the divergence easy to miss unless highs/lows are checked across days rather than just within the current session. — Lathyrus Masterclass — Supplementary Lessons — Afyz Mini-Lesson: Asset Synchronization & Phases of Price, 1:20
  • Spot an 'SMT inside a fair value gap' without dropping to a lower timeframe: since a gap's boundary is itself formed by a high/low, checking whether price reaches above that internal boundary on one asset but not the other is functionally a lower-timeframe SMT read, done directly on the higher-timeframe chart. — TTrades — "SMT Divergence - ICT Concepts", 15:35
  • For this session-specific strategy, upgrades the standard 'no SMT, no trade' rule to require a sequential SMT specifically — divergence between adjacent quarters/candles on the 6-hour or 90-minute timeframe — not just any SMT at any key level. — $niper — "Steal My Asian Session Trading Strategy (Works in all Markets)", 2:05
  • When one correlated pair taps a fair value gap but the other pair fails to tag the equivalent gap, that failure-to-tag is treated as confirmation the gap is a genuine breakaway gap (price already ran the liquidity it needed and won't have to come back) — extends the standard SMT-divergence check from swing highs/lows specifically to gap tags. — $niper — "Where to Place your Stop Loss in Trades", 7:28
Community insight
  • Reversals usually require 2-stage SMT.
  • 15m gaps are only suitable for early reversal, preferably in discount.
  • Do not use 15m gaps in premium as a reversal basis.
  • To confirm the daily high / low, usually require:
    • a swing point;
    • SMT;
    • sometimes 2-stage confirmation.
  • For recent gaps, 1-stage SMT can be used for continuation.
  • For old gaps, 2-stage SMT is usually required.
  • For 2-stage SMT, the first stage should ideally occur at the key level.
  • In a gap, usually only 1 stage is needed.
Community insights

If there is no SS / HTF open / new key level to time expansion, but the market has a lagging-asset sequence, understand the condition as:

  • a negative condition;
  • SMT-break direction becomes more important;
  • prioritise the lagging asset’s catch-up / SMT-break sequence;
  • do not force the original reversal idea.
Community insights

Low-probability SMT:

  • SMT high / low is not printed at roughly the same time.
  • SMT fill creates EQL / EQH, unless there is a V-shape recovery.
  • Double sweep is valid, but low probability.
  • Close-proximity SMT in premium with a swing is usually more like a retracement.
  • SMT away from a key level is lower quality.
  • Valid SMT should ideally be at a key level.
  • Exception: a failure swing with SMT can still be tradable if other conditions support it.
  • If daily / weekly highs and lows are wickless on both assets and only one later trades into that opening price to create a wick, the SMT is low quality — the other asset still has a reason to return to its level.
  • Only SMTs aligned with the direction of your draw liquidity are worth tracking. An SMT that forms on the opposing side of the range is irrelevant to the trade — it is only ever going to produce a retracement, not a reversal of your draw, so it does not need to be filtered out or feared.

Community insights
  • The market can't reverse without a swing point, and it can't reverse without SMT — PSP is swing point plus SMT. — Garrett — "Precision Swing Point", 01:55
  • 2-stage PSP (SMT+PSP or PSP+SMT) ranks above plain PSP, which ranks above a plain swing point. — Garrett — "Precision Swing Point", 04:55
  • A two-stage crack doesn't always need its first-stage PSP to occur exactly at the key level itself — the PSP can form on an earlier candle away from the level, with price only reaching the actual gap/key level afterward, and the combination still counts as a valid two-stage confirmation. — Anomaly — Advanced course — Lesson 5: Candle Confirmations, 5:29

PSP = swing point + SMT.

The market cannot make a high-quality reversal without both a swing point and SMT.

Probability ranking:

  1. Swing point
  2. PSP
  3. 2-stage PSP
    • SMT + PSP
    • PSP + SMT

Key rules:

  • PSP should ideally occur at the swing point.
  • 2-stage PSP = 2-stage SMT.
  • SMT + PSP is the highest-probability reversal structure.
  • When SMT breaks, it may create a true reversal.
  • A driver candle forming PSP is a time-based CiC and has higher probability.
  • After decoupling, the 6:00am 4h close may often form a PSP, which becomes important for the 10:00am 4h re-sync.

Community insights

SS = the lagging asset becomes strong, or the relative strength relationship between leading and lagging assets changes.

Common forms:

  • SSMT
  • PSP
  • Gap-size difference
  • Displacement-speed difference
  • Candle-body difference
  • Bullish / bearish candle bias difference
  • Double sweep, low probability

SSMT (sequential SMT) is a time-based CiC: it only asks whether the high/low that just printed SMT is the previous candle’s high/low on some specific timeframe — the previous 90-minute high/low, the previous 6h high/low, the previous daily high/low, or the previous weekly high/low. Nothing more than that; it is not evaluated against any candle further back than the immediately previous one on the timeframe being used.

  • SS overrules all other confluence.
  • If assets decouple, wait for SS before entering.
  • SS is the asset re-sync trigger.
  • SS confirms daily ERL ↔ IRL, meaning DH / DL.
  • If SS appears at a reversal, usually all assets will reach the draw.
  • If there is no SS, all assets need a V-shape before a reversal can be considered.
  • When timing expansion, SS has priority over HTF open and new key level.

Community insight
  • Two-stage SMT shows up either as a same-timeframe 'roof' (a close-proximity switch soon after the sweep) or as a large higher-timeframe SMT confirmed by a smaller lower-timeframe one — either way, don't trust it until a strength switch actually confirms it. — Garrett — "A Deeper Dive into SMT Divergence", 9:22

Two-stage SMT can take more than one shape:

  • Same-timeframe form: the asset that swept switches strength at a close-proximity high / low soon after, sometimes shaped like a roof.
  • Higher-timeframe / lower-timeframe form: a large-range, higher-timeframe SMT, for example a daily swing high, is confirmed by a smaller-range, lower-timeframe SMT.
  • Either form should still be confirmed by a strength switch before it is trusted. Do not assume the first-stage SMT is holding until the strength switch appears.

Community insights
  • Mechanically, look at the closes: a PSP is candle 2 closing back within candle 1's range on both correlated assets, but one closes bullish and the other bearish. That opposing close is the crack in correlation. — Garrett — "GxT | Precision Swing Point | Pt.1", 27:35
  • A Continuation PSP forms in the retracement after a true reversal, not at the reversal itself — one asset prints a swing point in that retracement while the other doesn't, and that SMT is what ends the retracement and resumes the original expansion. — Garrett — "GxT | Precision Swing Point | Pt.1", 29:56
  • A Strength Switch PSP is one of the highest-probability sequences: the asset that sweeps the high or low is normally the stronger asset in the triad, so when it closes against its expected bias, that mismatch is a Strength Switch layered onto the PSP. — Garrett — "GxT | Precision Swing Point | Pt.1", 31:39
  • Mechanically, a PSP is a candle 2 that closes back within candle 1’s range on both correlated assets, but in opposing directions — one closes bullish, the other bearish. That opposing close is the CiC.
  • Continuation PSP: forms during the retracement that follows a true reversal, not at the reversal itself. One asset creates a swing point in that retracement while the other doesn’t, forming an SMT — this is the trigger for the retracement to end and the original expansion to resume, not a new reversal signal.
  • Strength Switch PSP: the asset that sweeps a high or low is normally the stronger asset in the triad, but instead closes against its expected bias. That mismatch is itself a Strength Switch, layered onto the PSP, and is one of the highest-probability sequences to trade.

3.6 Confirmation Ordering for a True Reversal

Section titled “3.6 Confirmation Ordering for a True Reversal”
Community insights

A true reversal always needs two CiCs, but which one forms first changes whether candle 2 itself is tradeable or whether you have to wait for candle 3:

  • PSP confirms SMT: a precision swing point closes first at the key level, and a consecutive-candle SMT then confirms it while candle 2 is still open. Because the two-stage crack is already complete, candle 2 itself can be traded as the reversal.
  • SMT confirms PSP: an SMT forms first, but the swing is only confirmed once that candle closes as a PSP. Since the two-stage crack isn’t complete until the close, you cannot trust candle 2 in real time — wait for candle 3 and trade it as the continuation instead.
  • Two-stage PSP with two-stage SMT: a two-stage PSP forms, and the high or low it creates is itself then run and reconfirmed by a second, separate two-stage SMT. This three-stage combination is rarer but a very high-probability signature when it appears.
  • If there is no two-stage CiC at all — for example only a single PSP or a single SMT with no confirmation — it is not a true reversal, and the fractal should not be traded.
  • As a rule of thumb, the number of CiCs required drops by one at every step away from the reversal: a true reversal itself needs two; the higher-timeframe continuation gap it creates needs one (an SMT fill or single PSP/SMT closure is enough); the lower-timeframe reversal signature traded inside that gap needs none beyond the V-shaped reversal signature itself, since the higher-timeframe crack has already done the filtering.

A candle’s own timeframe does not always show the two-stage crack directly. There are three cases:

  1. Same-timeframe confirmation: the candle itself prints the two-stage CiC at its own high / low — trade it directly, no further profiling needed.
  2. Lower-timeframe profiling: the candle shows no two-stage crack on its own timeframe, but a lower timeframe (for example a 4h swing profiling a daily candle’s wick) shows the two-stage crack instead. Use that lower-timeframe confirmation to trust the higher-timeframe wick.
  3. No confirmation on any relevant timeframe: neither the candle’s own timeframe nor an appropriate lower timeframe shows a two-stage crack. There is no true reversal here — wait for the candle to close, then trade continuation from the new invalidation it creates instead.

Personal study notes, shared as-is and in good faith. Educational material only — nothing on this site constitutes financial advice.