Multi Timeframe Trend (MTT) in plain terms
Multi Timeframe Trend is an approach that compares trend direction and/or trend strength across more than one chart timeframe (for example, a higher timeframe like H4 alongside a lower timeframe like M15). Instead of treating one timeframe as the only “truth,” it uses multiple time resolutions to judge whether the market has aligned direction, conflicting direction, or mixed conditions.
Because this concept is used in different ways, the first piece of “data” you need is a written definition of what “trend” means in your method (direction only, direction plus momentum, direction plus moving-average slope, or another measurable rule). Without a definition, you cannot consistently assess anything.
Direct answer: what data you need
To assess Multi Timeframe Trend in a way that you can explain and independently check, you need four categories of inputs.
1) Market price inputs (the raw data)
You need historical price series with consistent fields such as open, high, low, close, and timestamps. Choose one price source and keep it consistent across all timeframes. Also decide whether your method uses candle closes only or uses intrabar values.
Material details that matter:
- Timeframe set: the specific higher and lower intervals you compare (e.g., H4 and M15).
- Session and timestamp convention: how the provider labels candle boundaries.
- Data completeness: missing candles, duplicated timestamps, or incomplete history.
2) A trend measurement definition (the measurable rule)
You need rule-based measures that convert price data into a “trend state.” Examples of measurable components you might define include:
- Direction rule: “up” if a defined series is above its reference, “down” if below.
- Strength rule: a threshold for slope, rate of change, distance, or momentum.
- Aggregation rule: how you combine multiple confirmations (for instance, majority alignment, all-align requirement, or weighting).
This is not about a named indicator as a standalone signal; it is about specifying the exact calculations and thresholds your method uses so others can reproduce them.
3) Provenance and timeliness (what the data claims)
Even for historical work, provenance matters. Record:
- Data origin: which platform or feed the price series comes from.
- Symbol specification: how the instrument is defined (naming, contract conventions).
- Versioning: if your platform revises historical candles (some sources can correct or adjust data).
For “timeliness,” separate two ideas:
- Whether the historical dataset is complete up to the point you analyze.
- Whether the dataset could change later due to provider updates.
4) Calculation settings and assumptions (what can change results)
You need all method settings that affect output. This usually includes:
- Lookback lengths (how many candles you use).
- Any smoothing or resampling steps (how lower timeframes are produced from the same source).
- Handling of timezones and daylight saving differences.
If you make an example, state assumptions explicitly (for instance, “using candle close values only,” “using these two timeframes,” and “defining up/down by whether X is above/below Y”).
Mechanics: how the data is used
A reproducible MTT assessment generally follows a workflow:
- Select timeframes and extract the corresponding candles from the same data source.
- Compute trend states for each timeframe using your defined rule.
- Align signals by time using candle timestamps so higher-timeframe context matches lower-timeframe periods.
- Classify the multi-timeframe outcome (e.g., alignment vs conflict) according to your specified combination rule.
A key operational limitation is that timeframes overlap differently. If you align by candle start times instead of candle close times (or vice versa), results can differ. So your “data” must include the timestamp alignment logic, not just prices.
Evidence or example (with explicit assumptions)
Consider a reproducible example where your trend definition is purely directional:
- Assumption A: Use candle close prices only.
- Assumption B: Higher timeframe is H4; lower timeframe is M15.
- Assumption C: Trend on each timeframe is “up” when a chosen reference series is rising across the most recent N candles; otherwise “down.”
Data you would need:
- A historical OHLC dataset that contains M15 candles at the correct timestamps.
- A method to compute/derive the H4 series consistently from the same source (or extract H4 directly from the same provider).
- The value of N and the exact rising condition (for example, strictly higher each candle close vs non-decreasing).