The Relative Strength Index (RSI) compresses recent gains and losses into a bounded indicator value. The number does not dictate what to do. One trader may treat a high reading as evidence of momentum; another may test a pullback or reversal condition. Those are competing hypotheses. A reader cannot interpret an “RSI strategy” without knowing the calculation, threshold direction, signal timing, exit and market context.

RSI indicator value to research decisionAn RSI calculation needs enough prior observations under a declared smoothing convention. The illustration separates a computed value, a completed-bar threshold event and a later modeled action; it does not select the threshold or recommend a trade.A BOUNDED INDICATOR VALUE IS NOT YET A TRADING RULE1 · Specifysource · periodsmoothing · warm-up2 · Calculate0–100 valueor unavailable3 · Confirmclosed-bar crossingchosen threshold4 · Modellater entryexit + costsSame RSI line · reversal and momentum rules ask different questionsThresholds are research choices · compare failures and avoid searching only for a winner
An RSI calculation needs enough prior observations under a declared smoothing convention. The illustration separates a computed value, a completed-bar threshold event and a later modeled action; it does not select the threshold or recommend a trade. Illustrative teaching graphic · not live market data

Describe the indicator version

Specify the input price, period, smoothing convention, timeframe and source. RSI implementations can differ in warm-up and smoothing. State the exact formula or software behavior used, and check its first valid values against a small reference example. Until sufficient observations exist, mark the value as unavailable rather than substituting zero. TradingView’s Pine v6 built-in reference documents its ta.rsi() function signature; do not assume a different platform shares its initialization convention without checking.

Reproduce one value before testing a rule

For a fictional first 14-change window, use +1, +2, −1, 0, −2, +3, −1, +2, −3, +1, +1, −2, +2 and −1. The positive changes total 12 points and the absolute negative changes total 10. The initial mean gain is 12 ÷ 14 = 0.8571 and the initial mean loss is 10 ÷ 14 = 0.7143. Relative strength is 0.8571 ÷ 0.7143 = 1.2, so RSI is 100 − 100 ÷ (1 + 1.2) = 54.55. TradingView’s RSI reference describes the same relationship and Wilder-style RMA smoothing for later values. This is a calculation check, not a market observation or a signal. Confirm the first valid bar and zero-gain/zero-loss behavior in the library or platform you actually use.

Then choose one hypothesis. A threshold-reversal rule might log a possible long observation when RSI crosses up through a lower threshold after a completed bar. A momentum rule might instead trigger when RSI crosses upward through a higher level. “Below 30 means buy” leaves open whether a persistent low value creates a signal on every bar, whether a crossing is required, and what proves that the position should close.

Write a sequence with timestamps: the bar on which the threshold condition becomes true, the first later bar at which the decision is allowed, the entry convention, stop or invalidation, exit condition and holding limit. If a stop and target both lie within one candle, state how the model resolves uncertain intrabar order. Every choice forms part of the strategy version.

For example, take three fictional confirmed RSI readings of 28, 27 and 32 with a predeclared lower threshold of 30. A rule that says “RSI is below 30” is true on the first two closes. A rule that says “RSI crosses back above 30” fires once, after the third close. If the test then models entry at the following bar’s open, the signal bar and entry bar remain distinct. TradingView’s strategy broker emulator also applies a next-tick fill convention by default; that is a platform simulation rule, not a promised exchange fill. The threshold and readings here are teaching inputs, not recommended settings.

RSI threshold crossing and later entry timingThree fictional confirmed RSI readings of 28, 27 and 32 are compared with a predeclared lower threshold of 30. A level rule remains true on two closes below 30, while the cross-up rule triggers only when the third reading closes above 30. Any entry is modeled later and requires separate exit and cost rules.A LEVEL CONDITION AND A CROSSING CONDITION CREATE DIFFERENT EVENTSFICTIONAL RSI READINGS · CONFIRMED CLOSES ONLYthreshold 30282732bar 1 closebar 2 closebar 3 closeLevel rule · RSI below 30True at bars 1 and 2Cross rule · 27 → 32Signal confirmed after bar 3 closesModel entry at a later eligible priceA signal is not an exit plan · define holding limit, costs and invalidation · no recommended threshold
Three fictional confirmed RSI readings of 28, 27 and 32 are compared with a predeclared lower threshold of 30. A level rule remains true on two closes below 30, while the cross-up rule triggers only when the third reading closes above 30. Any entry is modeled later and requires separate exit and cost rules. Illustrative teaching graphic · not live market data

Distinguish momentum from reversal

An extreme indicator reading does not ensure an immediate price reversal. In a persistent move, RSI may remain elevated or depressed while price continues in the same direction. A reversal rule can therefore encounter repeated entries too early. A momentum interpretation has its own risks: threshold recrossings may happen after much of a move is complete, and frequent signals may be sensitive to costs.

RSI reversal and momentum hypothesesThe same bounded RSI series supports different testable hypotheses: a cross back above a predeclared lower threshold for reversal, or a cross above a predeclared higher threshold for momentum. The lines show an illustrative indicator path and do not imply recommended levels.THE INDICATOR DOES NOT CHOOSE THE HYPOTHESIS100500lowerhigherReversal testcross back abovechosen lower levelMomentum testcross above chosenhigher levelUse completed-bar values · define later entry, exit and market context · not a buy/sell signal
The same bounded RSI series supports different testable hypotheses: a cross back above a predeclared lower threshold for reversal, or a cross above a predeclared higher threshold for momentum. The lines show an illustrative indicator path and do not imply recommended levels. Illustrative teaching graphic · not live market data

Use a baseline and chronological split. Decide how to compare the rule before testing, such as net results after stated costs, drawdown and exposure, rather than selecting whichever metric later favors it. Record the complete threshold/period search. Test nearby values and other market intervals without presenting the best combination as an original forecast. Once a later interval has informed a setting change, it no longer serves as an untouched evaluation period.

Include instruments and execution assumptions

For spot, account for transaction costs and custody or venue limitations relevant to the actual product. For a perpetual, include the actual settlement funding records during eligible holding intervals, in addition to entry and exit fees, spread and adverse slippage. The sign and interval convention are venue-specific. Leverage changes exposure and liquidation risk; a stop order does not guarantee an exit at its trigger price. Verify current contract, collateral and margin rules for the selected venue.

The CFTC’s virtual-currency advisory describes risks associated with leveraged virtual-currency futures. Its U.S. context does not replace local rules or the specific contract documentation. Keep instrument and jurisdiction explicit in any explanation; do not generalize one venue’s mechanics to all crypto markets.

Check what the product really implements

SuperQuant’s sample RSI reversal rule detects a threshold crossing on a previously closed hourly candle, then uses the engine’s later modeled entry sequence. This is a limited product capability; RSI helpers or research text do not imply a general visual indicator builder. Its example histories are synthetic and the model is not a real exchange. The result can help inspect that configured rule but cannot verify market performance or live order execution.

Keep one immutable record per tested rule, with the data provenance, indicator definition, parameters, trigger and exit, cost assumptions, baseline, results and a failure case. The mean-reversion guide explains why a return-to-average premise can fail; the trend guide treats delayed crossover signals. Use the backtest checklist before interpreting outputs. This article is general education, not a trading signal or personal advice.