What Data Is Needed to Assess Trend Intensity Index?

Direct answer

To assess a Trend Intensity Index in a way that another reader can independently verify, you need four things: (1) the definition or formula used to compute it, (2) the input data series and preprocessing choices used by that formula, (3) provenance and timeliness details about where the data came from, and (4) quality checks and limitations that explain when the indicator can fail.

Because there is no single universally fixed “Trend Intensity Index” definition, the most important “data” is the exact computation specification—otherwise you cannot tell whether two people are measuring the same quantity.

Mechanism and definition (what to collect)

A Trend Intensity Index is a derived number intended to reflect how strongly price action appears to be trending, based on inputs such as recent price movements. To assess it, collect the following non-variable elements:

1. Indicator definition (calculation specification)

• The formula: which mathematical steps convert inputs into the index value.
• Required parameters: lookback length, smoothing method, scaling/normalization, and any thresholds (if the index includes them).
• Sign conventions: whether the index is absolute (trend strength only) or directional (separate up vs down).

2. Input series used by the formula At minimum, specify the price data used (for example, close only, or open/high/low/close). If the formula uses derived quantities, document those too:

• Returns or differences (how they are computed).
• Trend proxies (such as moving averages or regression outputs, if included).
• Volatility or scaling factors, if the index adjusts for typical variation.

3. Timeframe and alignment assumptions Even with the same formula, results change with timeframe and alignment. Document:

• Bar timeframe (e.g., 1-minute, 1-hour, daily) and whether it matches the index specification.
• How bars are aligned (server time vs local time, and session handling if applicable).
• Treatment of non-trading periods (weekends, holidays) and whether missing bars are dropped or carried forward.

Evidence and example (how to verify inputs)

Since no real-time market data is assumed here, you can still set up a verification method using historical data you already have.

Example verification checklist (applies to any dataset/provider):

• AFVINKPUNTEN (checks that confirm you used the same ingredients):
  • The exact same input fields (e.g., close prices) are used as required by the formula.
  • The preprocessing matches: returns vs differences, log vs arithmetic, and any smoothing window.
  • The timeframe and bar count match the formula’s lookback.
• Bewijs of document (what to keep as proof):
  • A written copy of the formula and parameter values.
  • A short record of the data download settings (symbol, timeframe, start/end dates, and data type if offered).
  • The specific mapping from your dataset columns to the formula’s inputs.

Rode vlaggen (common failure modes):

• Two sources claim the “same” index but differ in parameter values (lookback length, smoothing) or in whether they use close vs typical price.
• One dataset has missing values or different handling of gaps; the index can shift even when the underlying price trend is similar.
• The index is computed on one timezone/session calendar while the inputs were gathered using another, causing bars to represent different moments.

Klaarcriterium (when you can trust the assessment is consistent): You can reproduce the same index values (within expected rounding) when you use the same formula, inputs, parameters, and preprocessing on the same dataset.

Limitations and risks (what can go wrong)

1. No single fixed meaning If the “Trend Intensity Index” definition differs between documents or platforms, comparing values may be meaningless. The index name alone is insufficient.

2. Sensitivity to preprocessing and provider differences Results can change due to rounding, missing bars, corporate actions handling, and differences in how OHLC data is produced.

3. Historical relationships do not guarantee future relevance Even if the index correlates with past outcomes, that does not establish that it will behave similarly later. Market regime shifts, liquidity changes, and changing cost structures can alter the relationship between “trend intensity” and real-world results.

4. Costs and execution context matter If someone later tries to translate the index into decisions, outcomes can vary with bid-ask spreads, commissions, slippage, and operational constraints. These are not properties of the index alone.