ICT Unicorn Model

ICT Unicorn Model runs on MT4 and MT5 and looks for one arrangement: a breaker block and a fair value gap occupying the same prices, inside a four-point swing structure.

All three of those things are ordinary on their own. Together they're not, and the whole design of this tool is about refusing to draw anything until all three line up.

The four points

Structure comes first. The tool tracks pivots and looks for four of them, labelled a, b, c and d on the chart.

For a bearish setup: d lower than b, c higher than a, and each point at least Swings bars apart from the one before it. That spacing rule is doing real work — without it the same four labels get pinned to four candles inside one consolidation and every squiggle becomes a structure.

For a bullish setup it's the mirror image: c is the lower low, d is the higher high.

The breaker and the gap

With the structure in place, the tool goes looking for the breaker candle at point b. It takes the candle at b if it's the right colour, and if it isn't, walks back up to four bars to find one that is. That block's high and low become the breaker's range.

Then it scans forward for a fair value gap. By default the gap has to overlap the breaker's range — either the gap's top or its bottom must sit inside the block. That overlap is the unicorn. Switch FVG valid type to Any closet FVG and the overlap requirement disappears, which gives you many more setups and a considerably weaker filter.

The gap also has to be taller than a multiple of ATR, so a two-point hole between two candles inside a range never qualifies.

What you get when one is found

A dotted box for the breaker, a solid box for the gap, and two coloured boxes stacked at the entry: risk below, reward above.

• Entry is configurable four ways: the midpoint of the breaker, its edge, the midpoint of the gap, or its edge. The default is the breaker's midpoint.
• Stop is the c pivot, the structural point that would invalidate the whole read.
• Target is entry plus the stop distance scaled by your reward-to-risk setting.

Then there's a clock on it. The setup stays live for Limit Order Timelife in bar bars, five by default, and if price hasn't reached the entry by then it's dropped and you get an "invalidated by time" alert. This is the part people underestimate: the model expires. It isn't a level that sits on your chart for a week waiting to be right.

The two filters that are off by default

Only unicorn that break previous session requires the setup to have taken out the previous session's high or low. The three killzones are drawn with their own start and end times and a timezone offset.

Only unicorn that break previous swing structure requires it to have broken the last swing point from a separate, slower pivot detector, defaulting to 30 bars left and 3 right.

Both are off out of the box. Turn either on and the number of setups drops sharply. That's the trade the tool leaves to you rather than deciding for you.

Do adjust Timezone before you judge the session filter. The session times ship as they are for one particular server offset, and a broker two hours away puts the London killzone in the wrong place without warning you.

The panel keeps score on your chart

There's a panel that tallies how the setups on your chart resolved: whether price reached the target before the stop, counted separately for long and short.

Read it for what it is. It's an arithmetic count of what happened on the bars you have loaded, on one symbol, on one timeframe, with your settings, and it assumes a fill at the marked entry with no spread, no slippage and no partial exits. It is not a backtest and it isn't a performance claim. What it's genuinely good for is comparing settings against each other on the same data: switch entry from breaker midpoint to gap edge and watch which way the count moves.

You won't find a win rate table in this listing, and I'd be careful with any indicator listing that has one.

It publishes real buffers

Six plotted buffers: entry, stop and target for both directions. That means iCustom gets you the actual numbers, which is unusual for a structure tool and the reason people build around this one.

The boxes, labels and arrows are chart objects and don't come out that way. The three prices do.

Where it's weak

The breaker search stops at four bars back. If the last opposite-colour candle before the push is further back than that, no breaker is found and the setup is skipped entirely. There's no setting for the depth.

Pivot detection needs bars to its right. Structure is confirmed after the fact, which is what makes it reliable and also means a setup appears a few bars after the swing that created it.

The entry, stop and target are geometry, not a trade plan. The stop is always the c pivot regardless of how far away that is, so on a wide structure the risk box can be enormous relative to the reward box. The tool draws it; it doesn't judge it.

Session times are one broker's. Adjust Timezone first, before deciding the killzone drawing is wrong.

It's quiet. On many symbols and timeframes you'll get a handful of setups a week. If your chart is empty, that's usually the tool working, not failing.

Questions you'll have

Does ICT Unicorn Model repaint? Setups are formed on closed bars, and a drawn model stays where it was drawn. What does change is a live setup's state: it can be filled, or dropped when its bar clock runs out.

Why does it draw nothing on my chart? The overlap requirement. A breaker and a gap both have to exist and share prices, inside a spaced four-point structure. Setting FVG valid type to Any closet FVG will give you far more setups and a much weaker filter.

What are a, b, c and d? The four structure pivots. For a bearish model c is the higher high and d is the lower low; bullish is the reverse. c is also where the stop goes.

Why did a setup disappear before price got there? Limit Order Timelife in bar expired, five bars by default. You'll have had an "invalidated by time" alert.

Can an EA read the entry and stop? Yes. Entry, stop and target are plotted buffers for both directions, so iCustom works. The boxes and labels are chart objects and aren't readable that way.

Is the panel percentage a win rate I can expect? No. It counts, on the bars you have loaded, whether the target was reached before the stop, assuming a perfect fill. Use it to compare your own settings, not as a forecast.

Best timeframe? M15 and H1 give a workable number of setups. On M5 the pivot spacing rule is the first thing to raise.

MT4? Both platforms are included, one download each. Same structure rule, same overlap requirement, same six buffers.

Support

Message me through the site with a screenshot, the symbol and the timeframe, and I can tell you which of the three conditions a candidate failed. Updates are included for as long as your subscription runs.

Analysis tool. It places no trades and makes no profitability claim. Past price behaviour is not a guide to future price behaviour.