Multi-timeframe analysis
Working from the higher timeframe down, and what to do when the timeframes disagree.
10 min read · intermediate · lesson 7 of 8
What this covers
- Assign each timeframe a specific job rather than reading them all the same way
- Work top down so context is fixed before entry detail is considered
- Interpret disagreement between timeframes as information rather than as a problem
- Avoid the search for a timeframe that agrees with a position already held
- Match the holding period to the timeframe the read came from
One market, three questions
There is a single market. Timeframes are not different markets; they are different sampling rates on the same sequence of prices.
What changes is the question each one answers well. A higher timeframe describes the condition that has been in place for weeks. A lower one describes what is happening in the next few hours. Neither is more true.
The failure is asking one timeframe to answer all three questions. A five-minute chart cannot tell you what the last two months have done, and a weekly chart cannot tell you where an invalidation level sits for a position held overnight.
What each level is for
| Level | The question it answers |
|---|---|
| Higher | What condition is in place, and which direction has the burden of proof |
| Working | Where the levels that matter for this decision sit |
| Lower | Whether the move is starting here, and where the invalidation level is |
Choosing the three
The useful separation between levels is roughly four to six times. Daily against four-hour against one-hour is a standard spacing; weekly against daily against four-hour is the same relationship shifted up.
Levels too close together produce three views of the same information and a false sense of agreement. Levels too far apart leave a gap where the context does not connect to the entry.
Which three depends on how long the position is held. A position held for days is read from daily downward. A position held for hours is read from four-hour downward.
The order is not optional
Start at the highest level
Establish the condition: trending, ranging, or in transition. Record it before looking lower.
Mark the levels that matter
On the working timeframe, the boundaries the higher-level condition makes relevant.
Only then go lower
The lower timeframe answers whether the move is beginning at one of those levels, and where it would be proven wrong.
Size from the lower level, hold to the higher
The invalidation distance comes from where the read fails; the holding period comes from the condition that produced the idea.
Starting low produces a view looking for support
Beginning on the lower timeframe produces an impression within minutes, because a short sample always shows something. Every higher timeframe consulted afterwards is examined for whether it permits that impression.
This is why the order matters more than the choice of timeframes. Establishing context first means the lower timeframe is answering a question. Establishing it last means it is defending an answer.
The tell is opening a fourth and fifth timeframe after the first two disagree with the position, and stopping at whichever one agrees.
Disagreement is the information
Timeframes disagreeing is the normal state. A daily uptrend containing a one-hour downtrend is just a pullback described at two sampling rates.
Which of them is in transition is the part that matters. A lower timeframe opposing a stable higher one is ordinary countertrend movement. A higher timeframe losing its structure while the lower one has already turned is a change beginning, and the lower timeframe registered it first because it samples faster.
Alignment across all three is the rarer state and produces the strongest case, at the cost of arriving late in the move. That is the trade being made, and it should be made knowingly rather than by waiting for comfort.
The holding period belongs to the timeframe the idea came from
An idea taken from a daily condition needs days to work. Managing it on a five-minute chart produces an exit on the first adverse hour, which converts a daily position into a five-minute one at daily-sized risk.
The inverse error costs more. An idea taken from a fifteen-minute read, held for a week because it moved against the position, is a position with no invalidation level at all.
Key takeaways
- Timeframes are sampling rates on one market, each answering a different question well.
- Roughly four to six times separation keeps the three views independent.
- Context first, entry detail last. The order prevents a view hunting for support.
- Disagreement is normal; which timeframe is in transition is the useful part.
- The holding period comes from the timeframe the idea came from, not from the chart being watched.
Common mistakes
- Opening the lowest timeframe first and forming a view before context.
- Consulting additional timeframes until one agrees with a position already held.
- Choosing timeframes too close together and reading the repetition as agreement.
- Managing a higher-timeframe idea on a lower-timeframe chart.
Knowledge check
OptionalAn assessment is a read across timeframes rather than a read on one. Where they disagree the state is reported as transitional instead of being resolved into a direction that the evidence does not support.
Related
- What market structure meansMarket Intelligence
- Trends, ranges and transitionsMarket Intelligence