SMT Divergence Explained

SMT divergence, short for smart money technique, is a mismatch between two normally correlated instruments where one makes a new high or low and the other fails to follow. ICT traders read that failure to confirm as a sign the move lacks broad participation and may be more vulnerable to a reversal. It is a comparative read across two charts rather than a signal from any single one, and it sits inside the broader smart money concepts framework.

What is SMT divergence?

SMT divergence is a smart money technique that reads two instruments which normally move together and looks for the moment they disagree. The core idea is confirmation: if two closely correlated markets are genuinely being pushed in the same direction, they should make matching new highs or lows. When one makes a fresh extreme and its partner refuses to, that lack of confirmation is the divergence.

The reasoning ICT traders attach to it is that a broad, well-supported move should show up across correlated markets at once. A new high in one index but not the other suggests the push is narrow rather than market-wide, which they interpret as a warning that the extreme may be a liquidity grab rather than the start of a lasting leg. It is fundamentally a relative read: neither chart means much in isolation, but the difference between them is the message. This makes SMT a cousin of intermarket analysis, which studies how related markets confirm or contradict each other.

Which pairs are used for SMT divergence?

SMT divergence only makes sense between instruments with a strong, stable correlation, because the whole method assumes they should move together. The two classic families are correlated equity index futures and correlated currency pairs. In each case, the tighter the historical relationship, the more informative a divergence becomes.

In US equities, traders compare the ES (S&P 500 futures) with the NQ (Nasdaq 100 futures). The two indices overlap heavily, so they usually make highs and lows in step. In FX, common pairs are EURUSD and GBPUSD, which share the US dollar as the quote currency and therefore tend to move together, or a pair and its inverse such as EURUSD against USDCHF. The shared driver is what creates the expectation of confirmation. When two currency pairs that share a leg stop agreeing, the odd one out is doing something its correlation would not normally allow. That anomaly is the signal SMT traders are hunting for.

How do you read a bullish or bearish SMT?

You read SMT divergence by checking whether both correlated instruments confirm the same new extreme, and treating any failure to confirm as directional. The direction of the divergence tells you which side looks unconfirmed. The two mirror cases are straightforward once you anchor on which instrument made the extreme and which one did not.

A bearish SMT appears at a high: one instrument makes a higher high while its correlated partner makes a lower high, failing to confirm the new peak. The failure suggests the upside push is not broad, hinting the rally may be running into a liquidity sweep rather than genuine strength. A bullish SMT is the mirror at a low: one instrument makes a lower low while the other holds above its prior low, refusing to confirm the new trough, which hints the downside is unconfirmed. The chart below shows the price paths of two correlated instruments where the second fails to make the matching new high.

Because SMT is about confirmation, it pairs naturally with a structural read. A divergence that lines up with a swept level and a shift in market structure trading carries more weight than one floating on its own.

How does SMT relate to liquidity and structure?

SMT divergence is most persuasive when it coincides with a liquidity event, because the two ideas describe the same moment from different angles. A new high or low that one instrument makes but the other does not is often exactly where stops sit, so the unconfirmed extreme frequently doubles as a liquidity grab. One chart sweeps the liquidity above a prior high; the correlated chart declines to, and the divergence is born.

Read that way, SMT is a confirmation layer stacked on top of structure and liquidity rather than a standalone method. ICT traders typically want three things to agree: a run into an obvious pool of liquidity, a failure of the correlated instrument to confirm the extreme, and then a shift in structure that signals the move is turning. When all three line up, the read is treated as stronger. When only the divergence is present, it is treated as a hint to watch more closely, not to act. This layering is why SMT rarely appears alone in ICT material and almost always sits beside liquidity and structure concepts.

Is SMT divergence reliable?

SMT divergence is not a reliable standalone signal, and its single biggest weakness is that correlation is not constant. The method assumes two instruments should move together, but the strength of a correlation drifts with the macro backdrop. A divergence between ES and NQ can reflect a genuine loss of breadth, or it can simply reflect money rotating between large-cap tech and the broader market, which is a normal, meaningful move rather than a warning.

Two further cautions matter. First, SMT is discretionary: which highs and lows you compare, and over what window, changes the read, so two traders can see different divergences on the same charts. Second, like the rest of the ICT toolkit, it is anecdotal and popular in retail circles but unbacked by academic evidence, and the appearance of a divergence at some past turns does not make it predictive. It works best as one input, cross-checked against relative strength and structure, rather than a mechanical trigger.

Putting SMT divergence in context

SMT divergence is best understood as a confirmation tool that asks a single sharp question: are two markets that should agree actually agreeing? When they do not, the odd one out flags a move that may lack the broad support a durable trend needs. Used alongside liquidity and structure, it can sharpen a read at a potential turning point. Used alone, it is easily fooled by ordinary rotation and shifting correlations.

The discipline is to verify the correlation first, then treat any divergence as a prompt to look harder rather than a signal to act. Anchored to the same risk control that governs the rest of technical analysis, SMT becomes a useful cross-check within an ICT workflow. It connects naturally to the Power of Three model, where the manipulation phase often prints exactly the kind of unconfirmed extreme SMT is designed to catch.