Smart money concepts, often called ICT after the trader who popularised them, is a way of reading a chart built on the idea that large institutions leave footprints: order blocks where they accumulated positions, liquidity pools above old highs and below old lows that price is drawn to sweep, and breaks of structure that confirm a change of direction. Retail traders learn to mark these zones and trade the reactions at them, usually on short timeframes.
The claim is that price moves to where the orders are, that institutions engineer sweeps of retail stops before the real move, and that a trader who reads the structure can enter where the institutions did.
We tested the mechanical forms of these ideas, order block reactions, liquidity sweeps and structure breaks, across 17 years of data with real trading costs, and none of them showed an edge. Their results were indistinguishable from entering at random on the same instruments over the same period, and after costs they lost.
The reason is that the concepts describe things that are true of almost every chart in hindsight. Every reversal has an order block behind it; every move sweeps some level. A rule that can find a setup on any chart cannot distinguish the ones that work from the ones that do not, and the losing setups are as common as the winning ones once they are counted rather than remembered.
If you trade it anyway, define the rules exactly, count every setup they produce over a year rather than the ones you noticed, and measure the result after spread. The method's appeal is in the story it tells about each trade. The test is in the trades it produces.