Which inputs does Multi Timeframe Trend use?

What “Multi Timeframe Trend” uses as inputs

Multi Timeframe Trend (MTT) is a trend-bias approach that forms a decision framework from multiple chart timeframes instead of relying on a single timeframe. In plain terms, the “inputs” are the data and the settings used to (1) determine trend direction on each timeframe and (2) combine those directions into a single, higher-level bias.

Because different platforms and providers implement MTT differently, you should treat the exact input list as “depends on the specific rules.” This article describes the common, stable input categories that you can verify in your own implementation.

Mechanism and definition: common input categories

MTT generally starts with these input categories.

Price data (the raw dataset)

Market price series such as open, high, low, close (OHLC) and/or a derived series (for example, averages derived from those prices).
The source of the series matters: some systems compute indicators from the same displayed candles; others may use internal feeds or slightly different candle construction. Even if you see the same timeframe labels, the exact bars can differ.

Timeframe selection (which charts to use)

At least two timeframes: a “higher” timeframe that represents broader market structure and a “lower” timeframe that represents timing or confirmation.
The specific timeframe pair (for example, any higher/lower choice) is an input parameter that can change the behavior materially.

Trend definition method (how “trend” is measured)

Many implementations define trend direction by applying a trend-measure technique to each timeframe’s price data.
Common ways include: comparing moving averages, evaluating whether price is consistently above/below a reference, or estimating directional bias using slope/structure rules.
The method you choose becomes a key input because it changes what qualifies as “up” or “down.”

Thresholds and parameters (the rule settings)

If the trend definition uses thresholds, they are inputs. Examples of parameter types include lookback length for whatever trend measure is used, smoothing length, and any “buffer” conditions.
These parameters are usually integers or floats inside the rule, and changing them can flip classifications.

Combination logic (how multiple timeframes become one bias)

Inputs also include the decision rule for combining results: for instance, whether the lower timeframe must agree with the higher timeframe, or whether disagreement cancels the bias.
Some rules require “alignment” across both timeframes; others weigh one timeframe more.

Evidence or example: what to list when you verify inputs

A practical way to describe MTT’s inputs is to write them as a checklist. Here is a self-contained example template you can adapt (without assuming any specific platform’s defaults).

Assume:

Two timeframes are used: higher timeframe H and lower timeframe L.
Trend direction on each timeframe is determined by comparing a reference trend measure.
The system outputs a bias only when H and L match.

Then your input list should explicitly include:

Data: OHLC series used for H and L (and the timezone/session assumptions if your chart uses them).
Timeframes: the exact selection of H and L.
Trend measure: the method used on each timeframe (for example, a moving-average comparison or a structure rule).
Trend parameters: the lookback/smoothing/threshold values.
Combination rule: whether it requires agreement (H up AND L up) or allows partial agreement.

Even without live prices, you can independently verify the inputs by replaying the same fixed settings on historical data and checking whether the classification matches your described logic.

Limitations and risks: material failure modes

MTT’s biggest risks come from mismatches between assumed inputs and what is actually computed.

Regime changes: a market can shift from trending to ranging (or vice versa). When trend definitions rely on historical continuity, classifications can lag.
Data and candle construction differences: two implementations can label timeframes the same way but compute slightly different candles, producing different trend measurements.
Parameter sensitivity: lookback lengths, thresholds, and the alignment logic can materially change which periods qualify as “trend.”
Confirmation delay: requiring multiple timeframes to align often reduces early detection but may reduce false positives in some regimes; the trade-off is not universal.

A limitation that matters for every calculation or example: historical relationships do not establish future results. Also, outcomes vary with market conditions, costs, execution, and jurisdiction, even when the inputs are identical.