SMC Indicator Comparison: Three Questions, One Index

I have read enough of these comparisons to notice the shape they share. Ten products, a paragraph each, a winner at the bottom, and a table whose columns quietly change between the products being compared. Somewhere in the middle sits ICT, which is not a product at all. The page argues with itself and the reader is the one who pays for it later, in a subscription that answered a question they were not asking.

This page is the index for that mess rather than another entry in it. It separates the three questions, gives you the criteria an honest comparison has to meet, and then points at whichever page settles the one you actually have.

What does an SMC indicator comparison actually compare?

Three different things, and they are never interchangeable. One is a way of reading price, one is a piece of software, and one is where that software lives. Each is settled by its own kind of evidence: an argument for the first, a test for the second, and your own working day for the third.

The category error is easy to make and expensive to keep. ICT is a body of teaching. Smart Money Concepts is the broader vocabulary most of that teaching now travels under. Price action is older than both. None of them is software, so none of them can be scored against a script the way a listicle scores them. When a page puts a school of thought in the same column as a subscription product, the column has stopped measuring anything.

Method comes first because it decides the contents of every other column. Say you accept a reading where structure only counts once inducement has been taken. A tool that labels highs and lows without it is then a different product, not a cheaper one, and switching to it changes what you trade rather than what you pay. Reject that reading and the same tool is fine, so do not pay for machinery you have decided you do not want. The order the layers are learned in matters more than the brand you end up with, and it is the part nobody sells you.

Only once the method is fixed does the tool question become answerable at all. Then it stops being about taste: either the implementation marks what the method requires, or it does not, and history on your own chart shows which. This is the one layer a stranger can reproduce, so it is the only part of your comparison worth writing down for anyone else.

Then there is the platform, which sounds like an afterthought and decides more purchases than either of the others. A tool that reads structure beautifully on a charting platform you have already left is worth nothing on Monday morning. This is where "best" turns into "available", and it is why two traders who agree about everything else still buy different products.

Why do most tool lists not settle anything?

Because they hand you a verdict instead of a method. A ranking you cannot reproduce is an opinion with a table around it, and the ordering can move for reasons that have nothing to do with the products: an affiliate arrangement, a review cycle, a page rewritten by someone who never loaded any of them.

Watch for the tells. A list that never states its criteria has none, and the products in it were chosen by whatever the writer could get access to. If every product in it wins a different superlative, the thing has been arranged so nobody comes last. And when a page scores a competitor without describing the test, it is quoting marketing copy at you. That applies to us as much as anyone: we have not run these tests on other vendors' builds, so we do not score them.

The honest version is duller and more useful. It names the criteria in advance, describes how each one is checked, fills in what it can verify, and leaves the rest of the grid empty for you. Seven such criteria and a protocol you can run in an evening are already published on this site, and the index further down points at them; this page does not repeat them, because the point here is which question you are answering, not how to score one candidate.

Two more habits are worth keeping. Score against your own instrument and timeframe rather than the vendor's demo chart, because a tool that is excellent on a trending pair can be noise on the one you actually trade. And write the results down as you go: a tool you liked on Tuesday will score differently once the criteria are in front of you, and the written row is the only thing that survives a good sales page.

Is ICT a rival to an indicator, or a different layer?

A different layer, and the confusion is the most common way this comparison goes wrong. ICT teaches a way to read liquidity and structure. An indicator is one implementation of whatever reading its author accepted. You do not choose between them any more than you choose between grammar and a keyboard, and a page that lists them side by side has skipped the only question that matters: which reading do you want implemented?

Once that is clear, the real fork appears, and it is about who does the marking. You can draw the structure yourself, or you can let software draw it and spend your attention elsewhere. Both are defensible, and the trade is not subtle.

Handing the marking to a tool inverts each of those.

There is a middle position and it is the one most working traders end up in: let the tool mark structure and zones, then check the two or three places where the method is strict before you act. That is also the honest answer to whether an SMC or ICT indicator on TradingView can replace learning the method. It cannot, and any vendor telling you otherwise is selling the wrong promise. The rules of the markup itself are the thing you are actually buying an implementation of.

Where do two readings of the same chart disagree?

At one candle, usually. Two tools can watch the same level and reach opposite conclusions depending on how price interacted with it: a wick through the level is a sweep of liquidity, a body closed beyond it is a break of structure. Everything downstream, including which zones are drawn and which trend you are told you are in, follows from that single call.

This is the disagreement worth checking before any feature list. Load a tool, find a labelled break on history, and look at the candle that earned the label. If a wick through the level was enough, the tool will keep printing breaks that never happened, and its trend read will be wrong for long stretches while looking entirely confident. There is one deliberate exception in the method: inducement is taken on any touch, so a wick counts there. Every other level waits for a body close beyond it.

Four objects carry almost all of the difference between implementations, and you can audit them in a few minutes each:

• Inducement (IDM), marked before the break. The level of liquidity price leaves behind and returns for. No inducement, no label: a high is not a higher high until the liquidity below the last valid pullback has been taken.
• The break, confirmed by a body close. A break of structure (BOS) continues the direction, a change of character (CHoCH) ends it, and both wait for a body beyond the level rather than an intrabar poke.
• The order block that produced the move. Tools differ on whether they tell you which block is tradable or paint every candidate the same colour, which hands the filtering back to you.
• Imbalance, tracked until it fills. A fair value gap that has already been filled is not a reason to act, and a chart still drawing it is training you to ignore your own tool.

One caution about testing any of this on history: history is the view after the fact, and a tool that rewrites closed bars will look perfect in it. Settle that first with a replay pass, because it decides whether anything else you checked was real. The full 100-bar protocol is a ten-minute job, it is linked in the index below, and it is the only one of these checks that cannot be postponed.

Where the tool runs, and how it reaches you

The third layer, and the one that quietly decides most purchases. Three questions live here: which platform runs the markup, how you find out that price has arrived, and what places the order once you have decided. They are independent of each other, and a product can be strong on one and absent on the next two.

Platform first, because it is binary. TradingView runs Pine Script and MetaTrader runs compiled MQL, so a script written for one does not load in the other at all: the logic has to be written twice, in two languages, by the vendor. That is a fact about the software, not a feature to weigh, and it removes candidates from your list before any other column is scored. If you place orders in MetaTrader, what actually runs there is a different shortlist from the charting-platform one, even when the marketing page shows both logos.

Notification is second, and it is the layer people discover they needed after the third missed setup. Marking quality is worthless at 3am if nothing wakes you, and the interesting question is not whether alerts exist but when they fire: as price approaches a zone, or once it is already inside and the entry has moved. The difference between an alert on approach and one after the fact is the whole value of the feature.

Execution is third and is the one most comparisons omit entirely. Marking, alerting and order placement are three jobs, and one product rarely does all three well. A charting indicator draws; a panel in your broker terminal sizes and manages the position. Deciding whether you need the third job at all is cheaper than discovering halfway through a subscription that you bought a drawing tool for an execution problem.

Which comparison settles which question?

Now the index. Find the question you actually have on the left, follow it to the page that settles it, and ignore the rest until it becomes your question. Every one of these is a page that names its criteria, which is the standard this page has been arguing for.

• "Which criteria decide between two tools, and how do I score them?" the seven-column scoring guide, including the ten-minute protocol on your own chart.
• "Do these things repaint, and how would I know?" the replay protocol for the test itself, and what the non-repaint claim covers for what the phrase does and does not promise.
• "Which of them runs in my broker terminal?" the MetaTrader shortlist, where the language a build is written in decides the answer.
• "I am on a funded challenge, does that change the choice?" yes, and mostly in one direction: the demands a rulebook adds push you towards fewer setups rather than more.
• "How do I know a marking is correct at all?" the rules of structural markup, which is the standard every tool on your shortlist is being judged against.

Two kinds of page are deliberately missing here. Head-to-heads against individual vendors, including the ones you have almost certainly met while shopping around, LuxAlgo and ChartPrime, are being written as separate pages, and they will follow the same rule: criteria stated first, cells left empty where we have not verified something ourselves. The other absence is permanent. There is no page ranking any of these products by outcome, because that number does not exist in a form you or anyone else could check, and that includes ours.

Where our own tools sit on this map

One filled row, since ours is the only one I can describe from the inside. On the method layer we implement the reading this page has been describing: inducement marked before the break, the break confirmed on a body close, the decision block kept separate from the extreme block, imbalance tracked until it fills. On the tool layer, Money Hunter is non-repaint by code, with markings finalised at bar close, and SMC ToolBox adds the entry-precision layer on top. On the platform layer it runs on TradingView and on MT4/MT5, alerts arrive through the SMCZone Signals bot on Telegram, and Smart Trade Assistant handles execution and sizing inside MT5.

Pricing is one subscription that carries every product, identical contents on every term: $69 for a month, $169 for three, $289 for six, $499 for twelve. The seven-day trial covers Money Hunter on TradingView and the Telegram alerts; the MetaTrader builds and the execution panel come with a paid term. Nothing is charged while the trial runs and you can cancel at any time.

Whether the method itself pays is a separate question with its own protocol: how to test SMC on your own charts.

FAQ