Measuring Trend in Gold
Counting red and green candles is not trend analysis. Here is what to measure instead, and how the disagreements between timeframes are handled.
Direction and structure are two different things
A timeframe can be moving down while still printing higher highs. That is not a contradiction — it is a pullback inside an uptrend, and it means something quite different from a frame that is moving down and printing lower lows.
So they are measured separately. Direction is net movement normalised by that frame's own volatility, which prevents a $2 drift in a quiet market counting the same as a $2 drop in a violent one. Structure is the swing pattern: higher highs and higher lows, lower highs and lower lows, or neither.
Where the close sits inside the range
A daily candle that opens at the low and closes at the high is a different statement from one with the same direction that closes mid-range after being rejected. Measuring the close's position within the bar's range captures conviction that a simple up/down classification throws away.
The same applies across several bars. Three daily closes at the top of their ranges is a stronger reading than one large green candle followed by two indecisive ones, even where the net movement is identical.
Combining timeframes without double counting
Weekly, daily, four-hour and one-hour readings mostly agree, and when they all agree it is tempting to treat that as four pieces of evidence. It is closer to one. This system therefore scores six components and lets them vote, but keeps a separate measure of whether the weekly and daily agree, and reports conviction alongside the grade rather than folding it in.
Strength attenuates, it does not reverse
Trend strength, volatility regime and news can only reduce conviction in a reading. None of them can flip its sign. This is a deliberate constraint: a system where enough secondary indicators can overturn the primary structural read will eventually find a combination that turns a clear downtrend into a buy signal, and it will do so at the worst moment.
Grades fixed in advance
The thresholds separating neutral from directional, and directional from strong, were written down before the measurements were run. It is easy to nudge such a boundary until a backtest improves, and the result is a number that describes the past accurately and the future not at all. The current reading and its components are published live.
Next: risk management and position sizing.
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.