The lower timeframe reports. The higher timeframe decides. That is the short answer to the follow-up Davide, one of our members, sent after last week's episode. His question was sharper this time: when the hourly is already overextended, the four-hour is showing an early reversal, and the daily is still undecided, what would confirm that a recovery is becoming a sustainable trend rather than another temporary rebound - and what would cancel it? He asked for both sides, on real charts. Gold, this year, happens to show both.
The rule: timeframes don't vote equally
A conflict between timeframes is not noise. It is a question that has been asked, but not answered yet. The lower timeframe is always the first to notice that something has changed. The higher timeframe tells you whether that change is big enough to matter.
So when the hourly, the four-hour and the daily disagree, read them in order. The hourly asks the question. The four-hour repeats it, louder. The daily answers last. That is why confirmation lives on the daily - and why a turn that exists only on the faster charts is, by definition, still a proposal.
When the daily said yes: June to August
In the summer, the slow clock was ready before the fast one spoke. Through June and July, gold's daily chart built a base. Horizon, our momentum layer, sank to the floor: the histogram and the line at the extreme at the same time. By our definition that is a washout - not a bottom, but a condition. Pulse, the trend layer, spent those weeks green under the bars: a bottom being built, slowly.
On the four-hour, Accumulate stamped again and again in the same price area through July, and Divergence twice marked a regular bullish divergence - price at a new low, momentum not confirming it.
Then, in early August, the move. Price left the area within a few sessions, and Horizon on the daily crossed zero and kept building. The important detail is the order. The four-hour signals did not lead that move on their own. The daily had been getting ready for weeks, so the fast clock's signal had something to stand on.
When the daily said no: October
In October, the fast clock said the same things - and the slow clock never joined. Early in the month the four-hour looked like a turn: two Accumulate stamps a week apart and a regular bullish divergence at the low. Price climbed back to just under 4,200.
But the daily never confirmed it. Its Horizon line stayed below zero the whole way up, and price rose straight into a resistance zone on the daily chart and stalled there. At the high, the four-hour printed a hidden bearish divergence: price held up, momentum didn't. Within a week, gold was back below 4,000.
| July | October | |
|---|---|---|
| 4H Accumulate | stamped again and again | two stamps, a week apart |
| 4H Divergence | regular bullish, twice | regular bullish, once |
| 1D Horizon | washout, then above zero | below zero throughout |
| Outcome | a trend | back below 4,000 |
Same kind of signals on the fast clock. A completely different outcome, because the slow clock never said yes.
Four clocks at Friday's close
On Friday, gold gave three different answers across four timeframes - Davide's situation, almost exactly.
- Hourly: price fell almost 200 points in two sessions. Sentinel reads it Active, with the most extreme trend reading on the whole stack, and Pulse is green under the bars. On this clock, the move is stretched.
- Four-hour: Sentinel says Recovering, and one bar before the close Accumulate printed a fresh stamp. On this clock, a turn is being proposed.
- Daily: Sentinel says Distributing. Price is sitting on a support band built in June and July - the floor of the summer base. Horizon's line is near the bottom of its range, but the histogram is not, so this is not a washout by our definition. Squeeze on the daily is in compression. Friday's close: 3,995.81.
- Weekly: Recovering.
Prism, reading the hourly, the four-hour and the daily at once, put it in one line: the three timeframes do not yet tell one story.
What would confirm it
Four things, in order, from fast to slow.
- The hourly cools off by time, not by price - sideways, instead of fresh lows.
- The four-hour holds its new zone. Phases are sticky: a recovery that is real stays Recovering through the first pullback instead of stamping again, lower.
- The daily changes its answer. Sentinel leaves Distributing, and Horizon's histogram turns up together with the line, the way it did in July.
- Acceptance. Price returns to the daily resistance near 4,200 and is accepted there this time, instead of rejected.
The first two can happen in days. The last two take weeks. If all you have is the first two, you have a rebound in progress - not a trend.
What would cancel it
Three things.
- Acceptance below the support band. If the floor of the summer base gives way and price starts to live under it, the July structure is no longer the reference.
- The next bounce into resistance prints the same thing as October: a hidden bearish divergence at the high on the four-hour.
- The four-hour falls back out of Recovering while the daily stays where it is. The question is withdrawn before it is answered.
Either way, it is information. A confirmation tells you the slow clock agreed; an invalidation tells you it didn't. Neither one is a prediction, and neither one is an entry - those decisions are yours.
A note on the examples: we chose gold because one chart showed both outcomes in the same year. Not every chart reads this cleanly - check it on your own markets.
Frequently asked questions
What is multiple timeframe analysis?
Reading the same market on more than one timeframe - for example the hourly, the four-hour and the daily - and treating each as an answer to a different question. The faster timeframe shows what changed first; the slower one shows whether the change has been accepted.
Which timeframe should you trust when they disagree?
Neither, on its own. The lower timeframe reports a change; the higher timeframe decides whether it matters. A conflict is a question that has not been answered yet - the useful work is defining what an answer would look like.
How do you tell a sustainable trend from a temporary rebound?
Look at whether the slower timeframe joins. On these layers that means the daily status changing, daily momentum turning with both the line and the histogram, and price being accepted at the daily resistance rather than rejected. A rebound usually shows up only on the faster charts.
Where the layers live
Accumulate, Divergence and Blocks are free on TradingView and TrendSpider. Horizon and Pulse live in Compass. Sentinel and Prism - which compared the three timeframes for you - live in Vector.
Davide, thank you for coming back with a harder question. Send us yours: the good ones become episodes.
Educational market commentary - not financial advice. Structura does not issue signals or predictions.
Read us first in your Google results:
Free on TradingView
Get the free structure layer.
Structura Accumulate marks where the market may be rebuilding structure before the move becomes obvious.
Get the free structure layer