Direct answer: what data you need
To assess Mcginley Dynamic in a way that another reader can independently verify, collect four categories of information: (1) the exact input data series, (2) the formula and parameter settings used to compute it, (3) the provenance and timeliness of the input bars or ticks, and (4) the quality and limitations of both the data and the implementation. If you lack any of these, you cannot reliably explain why the indicator behaves as it does or whether results are comparable across sources.
Mechanism and definition: what is being assessed
Mcginley Dynamic is a moving-average style measure. Assessment therefore starts with the same question you would ask for any moving-average calculation: what price series is the average applied to, and how is the update rule defined?
Concretely, you need:
- The input series definition (for example, whether the indicator uses close prices, typical price, or another price definition).
- The calculation specification (the exact mathematical update rule, including parameter values and any initialization method for the first value).
- The data frequency and bar boundary convention (for example, whether values are computed per candle and based on the bar close).
A stable concept to distinguish is the indicator’s mechanics versus variable conditions. Mechanics are fixed by the formula and settings. Variable conditions include market regime, volatility level, trading hours/rollovers, and data provider construction of candles.
Evidence and example checks: how to verify independently
Because no single market outcome proves correctness, the best evidence is reproducibility. Use these data checks so another person can confirm the same behavior:
- Verify input alignment
- Record the exact time range you used.
- Ensure the time zone and session handling match the source. Even when timestamps look identical, session rules can shift the effective bar boundaries.
- Document how missing candles are treated (dropped, forward-filled, or kept as gaps).
- Verify calculation settings
- Record the parameter(s) used by the implementation.
- Record the starting point or initial value approach (how the indicator seeds its first computable value).
- Confirm whether the implementation is evaluated “in order” with no look-ahead (computed using only information available up to each bar).
- Verify numerical consistency across sources Without assuming any broker or platform is correct, you can compare outputs:
- Run the same calculation using the same input bars from at least two independent data sources.
- Compare the resulting Mcginley Dynamic values for equality or small tolerated differences.
If values differ materially, the mismatch likely comes from input definition, bar construction, frequency, or parameter/settings rather than “indicator meaning.”
- Use simple, documented test assumptions When you provide an example in your own assessment, state assumptions explicitly:
- Use a clearly defined historical window.
- State the price definition used as input.
- State whether your comparison uses bar closes only.
This turns your explanation into something others can reproduce.
Limitations and risks: what can go wrong
At least one material limitation matters for assessment: indicator outputs can be highly sensitive to implementation details and data quality, even when the underlying idea is the same.
Key failure modes to document:
- Data-definition mismatch: one provider may compute candles differently (timezone, session, rollovers), leading to different inputs.
- Parameter sensitivity: changing settings can change responsiveness, so comparisons must use identical settings.
- Initialization effects: early indicator values may depend on the starting method and limited history.
- Missing/noisy points: gaps, bad ticks, or outliers can distort an update rule that reacts to recent movement.
- Historical non-transferability: relationships observed in one period do not guarantee similar behavior later, especially when volatility and regime change.
Outcomes also vary with costs and execution when indicators are used in trading contexts, but for assessment purposes you should separate “indicator behavior” from “trading performance.”
Verification or next question: what to collect for your checklist
If you want a complete assessment package, collect and document:
- Input series definition (price type, time range, timezone/session handling).
- Frequency and bar-close convention.
- Formula and parameter settings, including initialization.
- Data provenance (which dataset or vendor) and evidence of timeliness (whether values are historical and stable, not assumed live).
- Data quality notes (gaps, outliers, and how you handled them).