What happens when timeframes disagree?
You checked structure on the daily chart and got an uptrend. You checked the weekly and got a downtrend. Someone on the internet is looking at the hourly and calling it a range. Nobody is lying — and understanding why is what turns timeframe-hopping from confusion into a method.
Charts are nested, not competing
A weekly candle is five daily candles compressed. A daily candle is a whole day of hourly candles compressed. So a pullback on the weekly chart — one or two red candles inside a rising sequence — is, on the daily chart, a fully formed downtrend with lower highs and lower lows, lasting a fortnight and feeling entirely real while you live through it.
Both descriptions are correct at their own resolution. The daily downtrend is the weekly pullback. They are the same event, measured with different rulers.
A worked example
Illustrative weekly swing points on an index: lows at 4,000 → 4,300 → 4,550, highs at 4,200 → 4,600 → 4,900. Clean weekly HH/HL.
Zoom to the daily chart during the drop from 4,900 down toward 4,550. Inside those three weeks you'd find daily swing highs at roughly 4,880 → 4,780 → 4,690 and daily swing lows at 4,750 → 4,640 → 4,560. Lower highs, lower lows: a textbook daily downtrend, 7% deep, with every structural box ticked.
The two schematics below are that pair: the coarse sequence, and then the inside of one of its pullbacks.
Both labels are accurate at the same moment. "The index is in a weekly uptrend and a daily downtrend" is not a contradiction — it is a complete sentence, and a far more useful one than either half alone.
The method: pick a primary, use a secondary
Practitioners generally resolve this by deciding in advance:
- A primary timeframe — the resolution at which they describe the situation. This is chosen to match the horizon of the question being asked, exactly as Markets Foundations put it: question first, zoom second.
- One timeframe up — for context. What does the pullback you're staring at look like when it's one candle?
- One timeframe down — for detail. What actually happened inside today's candle?
Two or three views, chosen deliberately. The failure mode is the drift: you start on the daily, don't like what you see, flick to the four-hour, then the weekly, then back — until one of them shows the picture you wanted. That's not multi-timeframe analysis, it's shopping. The tell is that your conclusion arrived before your timeframe did.
Why this matters for everything that follows
Every tool in this domain — levels, indicators, patterns — inherits its timeframe from the chart you drew it on. A support level from a weekly chart and one from an hourly chart are different objects with different lifespans, and mixing them without saying which is which produces analysis that cannot be checked, by you or anyone else.
In the data
Coarser bars are folded out of finer ones. Switch the chart below between Daily, Weekly and Monthly: each weekly bar takes the first open, the highest high, the lowest low and the last close of its days, with their volume summed.
One seam is not visible in the folding. Intraday bars (hourly, five-minute) are kept as traded and never restated for later splits or dividends, while a long daily history is often shown adjusted. So a daily chart and a five-minute chart of the same stock are not always one series at two resolutions; check which prices each one carries before you compare levels across them.
Try it now
- Label the structure on each schematic above with your fixed swing rule, then write the one sentence that holds both labels at once. Neither is wrong and neither is complete.
- Now on real prices, with both rulers on the same page. Label the structure below on Daily bars, then press Weekly and label the identical window again. The disagreement, if there is one, is information about where you are in a larger move — and it is one button away rather than a second request.
- Decide, in writing, which single timeframe is your primary for this instrument and why. Next unit: the horizontal lines that traders watch across all of them.