Liquidity in Gold: Why Price Reaches for Obvious Levels
Why the level everyone can see is often the level price trades through, and how to tell a sweep from a genuine break.
Where the orders are
Stop-loss orders are not scattered evenly. They cluster just beyond the levels everyone can see: yesterday's high and low, the previous week's extremes, the high and low of the Asian session, and obvious equal highs on an intraday chart. A stop order is, mechanically, a market order waiting to fire.
So those levels are places where a pool of guaranteed opposing orders sits. Price reaching into one is not a mystical event; it is the market going where there is something to trade against.
A sweep is not a breakout
The distinction matters and is easy to get wrong after the fact. A breakout takes a level and continues. A sweep takes a level, triggers the orders resting beyond it, and then fails — price returns back through the level, and the traders who were filled on the break are now offside.
The two look identical for the first several minutes. That is why "the Asian high was swept" is only meaningful with a definition attached. In this system a level counts as swept only when it sat beyond the reference price, price traded through it by a minimum ATR-normalised distance, and price has since come back. A level that price is simply trading beyond all session is not a sweep, however convenient that reading would be.
Why "sell high, not low" is a liquidity statement
If the bigger trend is down, the tempting trade is to sell immediately. But the price available at that moment is often near the low of the recent move, which means a wide stop and a poor location. The alternative is to wait for an upward excursion — frequently one that takes the Asian high — and sell the rejection from a higher price with a tighter stop.
This is a real pattern, and it is worth being precise about how real. Testing it across six years of development data found that on strongly downward days the important move did follow an early upward excursion about 79% of the time — but the base rate across all days was 82%. In other words the day structure is genuine and it is not conditional on the downward reading. Waiting for that move improved results; the stated reason for waiting was wrong. Publishing that distinction is more useful than publishing the pattern alone.
Measuring it without hindsight
The trap in liquidity analysis is that every chart, read backwards, is full of perfect sweeps. The discipline is to define the level, the penetration distance and the rejection in advance, in units that do not change with volatility, and to record the reading at decision time so it cannot be revised later. Every signal on this site stores its reasoning sealed at the moment of decision.
Next: how the Asian and London sessions shape the day.
Every figure on this page comes from a simulated account. Simulated results are a record of what a system did, not a promise of what it will do. Nothing here is investment advice, and no result is guaranteed.