Methodology — How Each Gold Signal Is Decided and Measured

The higher-timeframe reading is a single graded score built from six independent components, each voting on direction, and three further readings that can only attenuate the result and never flip its sign. The components are weekly structure, the last few daily candles, daily structure, the four-hour balance, the one-hour structure, and the progression of daily swings.

The attenuators are trend strength, volatility regime and news. A contracting market or an important scheduled release reduces conviction; neither can turn a downward read into an upward one. The bands that separate neutral from directional, and directional from strong, were fixed in advance rather than chosen to make a backtest look better.

Direction alone is not a trade. A downward reading says nothing about whether the current price is a good place to sell — selling after a long decline into support and selling a rejection from the day's high are opposite bets with the same "direction".

So the system maintains a map of the day's reference levels and measures where price sits relative to them, in ATR units. It distinguishes a level that was simply traded beyond from one that was reached for and given back, which is the difference between a trend and a sweep.

Every signal must define where it is wrong before it defines what it hopes to make. The stop sits behind structure the market has already respected, and the distance from entry to that stop must fall within a sane band relative to the risk unit — too tight and the trade is noise, too wide and the reward is not worth the exposure.

Cost is a gate. The spread as a fraction of ATR is checked before a signal is allowed, and the measured round-trip friction of about 0.19R is subtracted from every result reported on this site. Every result in the signal history is after costs.

It does not forecast. It does not assign a probability to the next move. There is no confidence percentage attached to a signal, because a fabricated confidence number reads as meaningful and is not.

The system measures magnitude — how much of a normal day's range has been used, how extended a move is — more confidently than it measures direction, and the published record is arranged so you can check that claim rather than take it.

Every figure on this page comes from live trading. Past 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.