Direct answer
To verify information about RSI strategies, use a source hierarchy approach: first confirm the stable mechanics (what RSI measures and how it is computed), then verify any claims that depend on variable conditions (market regime, data quality, costs, and execution). Where a claim cannot be reproduced from stated assumptions, treat it as unverified.
Mechanism and definition
RSI usually refers to the Relative Strength Index, an indicator built from recent price changes. A “RSI strategy” is typically described as a set of rules that uses RSI values (for example, thresholds or patterns) together with specific decision logic (entry/exit timing, lookback length, and how signals are interpreted).
Verification starts by separating two categories:
- Stable mechanics: the indicator’s formula, parameter meaning (such as the lookback window), and how inputs are derived from price series.
- Variable conditions: what market data was used (timeframe, symbol, trading hours), how it was cleaned, and what operational assumptions were made (transaction costs, slippage, execution timing).
If an article or provider description mixes these without stating assumptions, verification becomes unreliable.
Evidence and reproducible example checks
Use reproducible checks that do not require live data.
-
Reconstruct the calculation inputs Assume you are given or you choose a historical price series. Clearly note what “price” means (close-only vs. other price fields), the timeframe, and the period length used for RSI.
-
Recalculate RSI step-by-step Compute RSI using the same lookback length and the same method described by the source you are verifying. Record intermediate values (at least the components used to form the RSI value). If your computed RSI series differs materially, the claim likely uses different assumptions or a different calculation convention.
-
Verify strategy rule interpretation If the “RSI strategy” claims include thresholds (for example, buying when RSI crosses a level), check whether the rule uses crossing direction, strict vs. inclusive comparisons, and whether decisions occur at bar close or at the next bar open. These details can change outcomes even when the RSI numbers are correct.
-
Check any backtest description for missing assumptions A verification-focused backtest review looks for stated costs (spread/fees), position sizing rules, and how exits are timed relative to indicator updates. If those are not specified, you cannot reproduce results.
-
Cross-check with an independent implementation Implement the verified calculation in a separate environment (or compare to another calculator) using the same input data and parameters. Agreement on RSI values is evidence that the indicator mechanics were represented correctly.
Limitations and risks (material failure modes)
Several limitations commonly break verification:
- Calculation mismatches: Different RSI conventions (data source, price field, smoothing method) can produce different RSI values.
- Lookahead bias and timing errors: If signals are evaluated on information that would not have been available at decision time (for example, using future bars), performance claims are not reproducible.
- Missing cost and execution assumptions: Results may change substantially when transaction costs, spreads, or slippage assumptions are altered.
- Market regime dependence: Historical relationships do not guarantee similar behavior in future or different market conditions.
- Unstated data cleaning: Outliers, corporate actions, missing bars, or timezone handling can alter the input series.
A good verification process therefore asks: “Can I reproduce the RSI values, then reproduce the strategy decisions, then reproduce the evaluation method—using the same explicit assumptions?”
Verification or next question to ask
If you want to verify a specific RSI strategy claim, start by writing down the parameter set (timeframe, lookback length, price field), the exact rule logic (crossing definition, evaluation timing), and the backtest assumptions (costs and execution). Then test whether those details are sufficient to reproduce RSI values and the resulting rule triggers.
If the source omits critical assumptions, treat the information as descriptive rather than verified, and focus next on clarifying the missing inputs before using any claimed conclusions.