TradingView indicators. One per question, tested.

Educational and general. This article explains how common technical indicators work and how to test them; it is not personalised investment advice and nothing here is a recommendation to buy, sell, or hold any security. Shishin is not affiliated with, endorsed by, or sponsored by TradingView; we mention its platform because it is where most people meet these indicators.

Open the indicator menu on a TradingView chart and you are offered hundreds of built-ins and a community library of more than 100,000 scripts. The question people actually ask is which few are worth keeping. Our answer comes from building a ranking system that uses the same raw ingredients, and it is shorter than most lists: pick one indicator per question, and test it before you trust it.

The short version. No single TradingView indicator is reliably “the most accurate.” Useful indicators each answer one question: trend (moving averages, the 50/200-day stack), momentum (RSI 14, MACD 12/26/9, rate of change), volatility (ATR 14, Bollinger Bands 20/2), volume (relative volume, OBV, VWAP) and relative strength versus an index. Use one per question, because RSI, stochastic and CCI largely repeat each other. Decades of research show single rules rarely survive costs and data-snooping tests, so test any indicator out of sample.

What indicators does TradingView offer, and how many can you use for free?

TradingView groups its library into built-in technical indicators, fundamental metrics, and community scripts written in its Pine Script language. As of October 2026 the free Basic plan allows two indicators per chart, and the paid tiers raise that to 5, 10, 25 and 50.

The Indicators dialog in Supercharts is split into tabs. The Technicals tab holds the built-ins: moving averages, oscillators, bands and channels, volume studies, plus built-in strategies and chart patterns. The Fundamentals tab plots company data such as price-to-earnings or price-to-book. Community scripts are the user-written library, which TradingView’s own help centre puts at over 100,000 custom indicators. The per-chart limits, read from the TradingView pricing page as of October 2026, are below. Plans and limits change, so check the current page before you rely on them.

A two-indicator limit sounds stingy. We would argue it is closer to the right number than fifty. The rest of this piece explains why.

Which indicators matter, grouped by what they measure?

Almost every indicator is a transformation of the same five inputs: open, high, low, close and volume. Grouping them by the question they answer is the fastest way to see which ones duplicate each other.

Trend: moving averages and the moving-average stack

A simple moving average (SMA) is the mean close over the last N bars; an exponential moving average (EMA) weights recent bars more heavily. The question it answers: is price, on balance, rising or falling over this horizon? Common lengths are 20 (about a month of trading days), 50 and 200. A “stack” checks several at once: price above the 50-day, the 50-day above the 200-day, and the 200-day itself rising, which we explain in the moving-average stack. The common misuse is treating a single crossover as a timing signal. Moving averages lag by construction; in sideways markets they cross back and forth and generate a string of small losses (“whipsaws”). They describe the regime well and time entries poorly.

Momentum: RSI, MACD and rate of change

Momentum indicators ask how fast price is moving and whether that speed is changing. The Relative Strength Index (J. Welles Wilder, 1978) compares average gains to average losses over 14 bars and scales the result from 0 to 100, with 70 and 30 as the conventional bands. MACD (Gerald Appel) is the 12-period EMA minus the 26-period EMA, with a 9-period EMA of that difference as the signal line. Rate of change (ROC) is simply the percentage change versus N bars ago; TradingView’s built-in defaults to 9.

The classic misuse of RSI is selling or shorting whenever it crosses 70. In a strong uptrend RSI can stay above 70 for weeks, and an “overbought” reading is often a sign of strength rather than exhaustion. MACD’s misuse is the mirror image of the moving average problem: it is built from moving averages, so its crossovers inherit their lag and whipsaws. Rate of change is the plainest of the three and, over longer windows (months rather than days), it is the raw material of the momentum effect documented in the academic literature.

Volatility: ATR and Bollinger Bands

Volatility indicators ask how far price typically moves, not which way. The Average True Range (also Wilder, 1978) averages the true range, the largest of today’s high minus low, the gap from yesterday’s close to today’s high, or the gap to today’s low, typically over 14 bars. Bollinger Bands (John Bollinger, 1980s) draw a 20-period moving average with bands two standard deviations above and below it.

ATR’s misuse is reading it as directional: a rising ATR means bigger moves, up or down. Its best use is sizing risk, for example placing a stop a multiple of ATR away so that it sits outside normal noise. Bollinger Bands’ misuse is treating a touch of the upper band as a sell signal; in a trend, price can “walk the band” for a long time. Their more interesting reading is the squeeze: when the bands narrow, volatility has contracted, and contraction tends to precede expansion, though it does not tell you in which direction.

Volume: relative volume, OBV and VWAP

Volume indicators ask whether the move has participation behind it. Relative volume compares current volume to its recent average; TradingView’s built-in Relative Volume at Time does this intraday, comparing volume so far in the session to the average at the same point in past sessions. On-Balance Volume (Joseph Granville, 1963) is a running total that adds the day’s volume on up closes and subtracts it on down closes. VWAP is the volume-weighted average price, by default anchored to the start of each session.

The misuses: OBV’s absolute level is meaningless (it depends on where the chart starts), so only its slope and divergences say anything. VWAP resets each session by default, so it is an intraday execution benchmark, not a swing-trading trend line; a daily chart VWAP is easy to misread. Raw volume without normalisation is the simplest mistake of all, because a large-cap stock always trades more shares than a small-cap one. We cover that normalisation in dollar volume and liquidity.

Relative strength versus an index

Relative strength asks whether a stock is beating the market, not just rising. On TradingView you can plot it by charting a ratio symbol such as a stock divided by SPY: a rising line means the stock is outperforming the index. Its misuse is confusing it with RSI, which, despite the name, compares a stock only to its own past. We unpick the two in the relative strength rating. For swing trading, relative strength is arguably the most under-used question on this list, because in a falling market the stocks that hold up best are often the ones that lead when it turns.

Indicator cheat sheet: the question, the setting, the pitfall

One row per indicator. Settings are the widely used defaults, not recommendations; changing them to fit past data is exactly the habit the evidence section warns about.

Why does stacking more indicators not help?

Because most indicators of the same type measure the same thing, and adding a second copy of a measurement adds confidence without adding information. Statisticians call this multicollinearity.

RSI, the stochastic oscillator (14, 1, 3 on TradingView) and the Commodity Channel Index (20) are a good example. RSI compares recent gains to losses; the stochastic locates the close within the recent high-low range; CCI measures distance from a moving average in units of mean deviation. Different formulas, same underlying question: has price recently moved up a lot relative to its own recent history? On most charts the three peak and trough together. When all three flash “overbought” at once, that is one observation reported three times, not three independent confirmations. MACD and a pair of moving averages overlap in the same way.

John Bollinger made the same point in his own book: confirm a signal with indicators that are not correlated with each other (momentum with volume, for instance), not with several momentum oscillators. The practical rule is one indicator per question. A chart with a trend measure, a momentum measure, a volatility measure and a volume measure asks four different questions. A chart with six oscillators asks one question six times, which is a big part of why the free plan’s two-indicator limit costs less than it seems.

What is repainting in TradingView, and why does it matter?

Repainting is when a script calculates or plots differently on historical bars than it did in real time, so the chart shows signals that were never actually available when they appear to fire. TradingView’s own Pine Script documentation defines it as script behaviour causing historical versus realtime calculations to differ, and notes that the large majority of indicators repaint in some form.

TradingView sorts repainting into three tiers. Widespread and usually acceptable: values on the still-forming bar change until the bar closes. Potentially misleading: scripts that plot into the past, such as pivot and zigzag tools that only confirm a high several bars after it happened, or strategies recalculated on every tick. Unacceptable: scripts that leak future information onto historical bars. The usual culprit is request.security() called for a higher timeframe with lookahead = barmerge.lookahead_on and no one-bar offset, which lets a historical daily bar “see” that day’s close before the day was over. The docs’ recommended pattern is to pair lookahead with a [1] offset so only confirmed values are used, and to act only on confirmed bars (barstate.isconfirmed).

This matters because the most impressive community scripts are disproportionately the ones that repaint: an indicator that “never misses a top” on history is often one that drew the top after it happened. Before trusting any community script, read its source for higher-timeframe calls and past-plotting, and watch it live on a few bars to see whether old signals move. The same flaw in a backtest is look-ahead bias.

What are the most accurate trading indicators, according to the evidence?

The research answer is that no single indicator has a durable, after-cost edge that survives proper testing. Some simple rules looked strong in long historical samples, then weakened in later data and largely disappeared once data snooping and trading costs were accounted for.

• Brock, Lakonishok and LeBaron (1992), in the Journal of Finance, tested moving-average and trading-range-break rules on the Dow Jones Industrial Average from 1897 to 1986 and found buy signals were followed by higher, less volatile returns than sell signals. It remains the most cited case for simple technical rules. It did not deduct trading costs.
• Sullivan, Timmermann and White (1999), also in the Journal of Finance, expanded that universe of 26 rules to thousands and applied White’s Reality Check, which adjusts for the fact that the best rule was picked out of many. Over the original sample the best rule still looked significant, but in the following decade of data it no longer did.
• Park and Irwin (2007), reviewing the literature in the Journal of Economic Surveys, counted 58 of 92 modern studies with positive results, but found most suffered from data snooping, ex post rule selection, or weak handling of risk and transaction costs, and that profits were concentrated before the early 1990s.
• Bajgrowicz and Scaillet (2012), in the Journal of Financial Economics, revisited the Dow from 1897 to 2011 with a false-discovery-rate method and concluded an investor could not have picked the future best rules in advance, and that even in-sample performance was wiped out by low transaction costs.

The pattern is consistent: an indicator rule that looks profitable is usually the survivor of a large search, and the edge shrinks as soon as you account for the search and the costs. That is the multiple-testing problem, and it is why we distrust any “most accurate indicator” claim that comes with a win rate and no out-of-sample record.

How do you test a TradingView indicator yourself?

Write the rule down before looking at results, test it on data it was not tuned on, include realistic costs, and compare it with a simple benchmark. TradingView’s Strategy Tester can do most of this if you configure it honestly.

1. Fix the rule in writing. Entry, exit, position size and stop, with the indicator settings, before you run anything.
2. Use confirmed bars only. Signals should fire on the bar close, with no lookahead in higher-timeframe calls.
3. Split the data. Tune on an earlier period, then run unchanged on a later one. See in-sample vs out-of-sample.
4. Add costs. Set commission and slippage in the strategy properties. Short-horizon rules are the most sensitive.
5. Test many symbols, not one. A rule that works on one chart you picked because it trended is anecdote.
6. Count your attempts. If you tried twenty settings, the best of them is flattered by the search. Our piece on overfitting explains why.
7. Compare with buy-and-hold and random entries. An indicator earns its place only if it beats doing nothing clever.
8. Forward-test. Paper-trade the unchanged rule for months before committing capital.

Which indicators does Shishin use?

Shishin uses the same raw ingredients as the indicators above, but it never relies on one of them alone. It combines them into a single rank across roughly 1,000 US stocks and judges the combination on out-of-sample and live results.

The daily board scores each stock on eight components, all built from price and volume: trend structure in the spirit of the moving-average stack, relative strength against the market, volatility contraction (the idea behind the Bollinger squeeze), volume behaviour, and ATR-scaled risk, among others. We describe them in the eight score components and explain why a blend beats a single reading in composite scoring. Each component was chosen to answer a different question, which is our answer to the multicollinearity problem: eight correlated oscillators would just be one oscillator, loud. We do not use valuation (P/E) or dividend yield; the scores are price-based.

Several design choices map directly onto the pitfalls in this article. Entries fill at the market close, so signals are computed on confirmed daily bars and cannot repaint. Stops are volatility-aware, using ATR rather than a fixed percentage. Data is point-in-time and split- and dividend-adjusted, which closes off the lookahead leaks that inflate many backtests. The four engines (Genbu for quality, Suzaku for momentum breakouts, Byakko for defence, Seiryu for large-cap recovery) apply these ingredients differently by market regime. None of that guarantees an edge, so we publish the evidence: the paper-traded results, misses included, are on the track record, and every signal is committed to the tamper-evident log on the verification page. The ranked board itself is on the rankings page; it is research, not advice.

Common mistakes with TradingView indicators

• Indicator soup. Five oscillators that agree are one opinion. Pick one per question.
• Tuning settings to the chart in front of you. An RSI length that “works” on one stock’s last year is fitted to noise. Defaults are not magic, but they are at least not fitted.
• Trusting a community script’s history. If it repaints, its past signals are fiction. Watch it live first.
• Reading volatility as direction. ATR and band width say how far, not which way.
• Ignoring costs and the benchmark. A rule that beats nothing after costs is a hobby, not an edge.
• Treating any of this as certainty. Our own composite approach has limits: it is late at turning points by design, it can and does have losing stretches, and a combined ranking is still built on past price behaviour. It is a more disciplined use of the same ingredients, not an escape from uncertainty.

Sources & further reading

• TradingView, plan comparison and pricing (per-chart indicator limits read 21 October 2026).
• TradingView, Pine Script User Manual, “Repainting”.
• TradingView Help Center, “TradingView indicators: simple steps to get started” and “Relative Volume at Time”.
• Brock, W., Lakonishok, J. & LeBaron, B. (1992). Simple technical trading rules and the stochastic properties of stock returns. Journal of Finance, 47(5), 1731 to 1764.
• Sullivan, R., Timmermann, A. & White, H. (1999). Data-snooping, technical trading rule performance, and the bootstrap. Journal of Finance, 54(5), 1647 to 1691.
• Park, C.-H. & Irwin, S. H. (2007). What do we know about the profitability of technical analysis? Journal of Economic Surveys, 21(4), 786 to 826.
• Bajgrowicz, P. & Scaillet, O. (2012). Technical trading revisited: false discoveries, persistence tests, and transaction costs. Journal of Financial Economics, 106(3), 473 to 491.
• Wilder, J. W. (1978). New Concepts in Technical Trading Systems. Trend Research. Bollinger, J. (2001). Bollinger on Bollinger Bands. McGraw-Hill.
• Composite scoring on why eight features beat one indicator.
• Why most backtests lie on overfitting and look-ahead bias.
• In-sample vs out-of-sample on the only test that counts.

Shishin Research