How to Read Buy Sell Signals on TradingView: Rules, Timing and Evidence

Distinguish a continuing state from a one-bar event. RSI below 30 is a state; crossing back above 30 is an event. Price above a moving average can remain true for many bars, while a crossover occurs only when the series changes sides under the chosen comparison rule. Record the source, lookback, threshold, chart interval and whether equality qualifies. For a hypothetical RSI sequence of 28, 29, 31 and 33, a cross above 30 occurs at 31. The later value of 33 is still above the threshold but is not another cross.

RSI compares smoothed gains and losses. Stochastic compares the close with a recent high-low range, then may smooth that position into K and D. A Stochastic cross can be filtered by whether both lines, one line or the prior bar was inside an extreme zone; those are different rules. For divergence, ask how swing points are selected. Comparing two adjacent bars is not the same as pairing confirmed pivots. A pivot with two bars to its right is known only two bars later. If a label is drawn back on the pivot, its chart position can precede the time a trader could have observed it.

In the common EMA form, MACD subtracts a slower EMA from a faster EMA. Its signal line smooths MACD, and the histogram is MACD minus that signal line. A signal-line cross and a zero cross are different comparisons; neither is automatically the stronger trading rule. A growing positive histogram means that difference is increasing, not a measured acceleration of future returns. TradingView ta.supertrend uses negative direction for its bullish state and positive direction for its bearish state. A bullish flip crosses below zero. Colors are author choices, so inspect the direction logic rather than assuming a green label proves a specific implementation. Moving-average crossovers likewise describe their input histories, not a guaranteed future trend.

VWAP is a volume-weighted average from a chosen anchor. A close above it describes position relative to that average; the anchor, session and volume feed determine the reference. OBV adds volume after an up close and subtracts it after a down close. It does not identify the buyers, sellers or institutions involved. Volume Profile groups reported activity into price rows. The selected interval, bin size and allocation method affect POC and value-area levels. Historical high-volume rows are not a live order book. A price break accompanied by higher volume can be a defined filter, but it does not prove that the break will hold.

An open-bar indicator can change on every price update. A condition that waits for the chart bar close uses a later, more stable observation. Higher-timeframe conditions need a separate decision about using their developing bar or the previous completed bar. Closing a five-minute bar does not close the current hour. TradingView alerts run from a saved snapshot of the script and settings. Updating chart inputs does not update an existing alert; recreate it when needed. A chart marker, an alert event, a simulated strategy fill and a broker fill are distinct outcomes. Check which one the claimed result measures.

Adding a filter selects a different subset of events. It can exclude winners as well as losers. RSI, moving averages and MACD share price inputs, so agreement is not automatically independent evidence. Specify how conflicting conditions are handled and whether all confirmations must occur on the same bar. A performance claim needs full entries, exits, costs, position sizes, symbol, date range and trade count. Compare results on data that was not used to choose parameters. Defaults do not establish repeatable performance, and no fixed small number of trades proves an advantage. For a hypothetical rule with a 40% win rate, an average win of 2 risk units and an average loss of 1, the gross average is 0.4 × 2 - 0.6 × 1 = 0.2 risk units per trade. Costs or worse fills can remove that margin. This arithmetic is an illustration, not a result for any indicator on this page.