Direct answer: verifying RSI and MACD information
Information about RSI (Relative Strength Index) and MACD (Moving Average Convergence Divergence) can be verified by separating stable indicator mechanics from variable market and provider details. First confirm the definition and formula each indicator uses, then reproduce the indicator output from the same input data and settings, and finally evaluate limitations such as parameter mismatches and implementation differences. No single source should be treated as authoritative for future performance; verification focuses on whether the indicator is calculated correctly under stated assumptions.
What RSI and MACD are (mechanism and definitions)
RSI is a momentum oscillator that summarizes recent price movement strength relative to gains and losses over a chosen lookback window (often called a period). A typical RSI calculation uses average gains and average losses over that period, then transforms the ratio into an index value.
MACD is based on differences between moving averages. Commonly, it uses a fast moving average minus a slow moving average to produce the MACD line, and then compares that line to a signal line (often another moving average of the MACD line). Because MACD depends on moving average type (for example, whether the implementation uses EMA or SMA) and on chosen periods, the same price series can produce different results if settings or formulas differ.
Evidence and reproducible verification steps
Step 1: Fix your assumptions before checking claims
Write down: (1) the indicator type (RSI vs MACD), (2) the parameter values (RSI lookback period; MACD fast/slow periods; signal period), and (3) the data source and sampling rule (for example, candle close prices on a specific timeframe). If any of these are unspecified, treat resulting comparisons as unverified.
Step 2: Verify the formula used
Check whether the source’s described computation matches the standard definition at a mechanics level:
- RSI must reference gains and losses over the chosen period and convert them into a bounded index.
- MACD must define two moving averages used to create a difference line, and specify how the signal line is derived.
This step verifies the “what” and “how,” not any performance claims.
Step 3: Recalculate on the same historical series
Use one fixed, historical price series (for example, a publicly accessible dataset or an export from your own platform) and compute the indicators using your recorded assumptions. Then compare your computed values against the values shown by the source you are evaluating.
To make comparisons meaningful, use the same input series (close price vs typical price), the same timeframe, and the same parameter settings. If your outputs differ, the difference is evidence that either assumptions differ or the provider’s implementation is not aligned with the stated definition.
Step 4: Isolate where differences appear
If RSI differs, check whether the implementation uses the same smoothing approach for average gains/losses. If MACD differs, check moving average type and indexing conventions (for example, whether the first values are offset due to warm-up periods). Identifying the first timestamp where values diverge helps pinpoint the specific mismatch.
Step 5: Cross-check with at least two independent calculators
Use two independent ways to compute the same indicators under the same settings and data. Agreement increases confidence that the mechanics are being applied correctly; disagreement indicates a likely difference in parameters, data inputs, or formula details.
Limitations and material failure modes
RSI and MACD are sensitive to inputs and settings. Common failure modes include:
- Lookback or period mismatch: changing RSI’s period or MACD’s fast/slow/signal periods can change the entire shape.
- Different price series: using close versus another price definition (even on the same timeframe) can alter results.
- Implementation differences: moving average type and RSI averaging/smoothing methods can vary across tools.
- Warm-up effects: early indicator values may be undefined or less reliable until enough data points exist for the longest lookback.
- Confusing verification with prediction: past relationships between indicator behavior and outcomes do not establish future performance.
Because costs, execution quality, and jurisdictional constraints affect trading outcomes, verifying “indicator accuracy” is not the same as verifying “profitability.” Verification should remain focused on correct calculation under stated assumptions.