Indicator reference
What each indicator measures, how to read it, and the mistake it usually invites.
An indicator is a rearrangement of the price and volume you already have — it adds no information the market has not already printed. What it adds is a consistent way to describe a condition, which is what makes a rule testable. Read these as definitions, not as recommendations: none of them predicts anything.
At a glance
| Indicator | Available in | Type | Parameters |
|---|---|---|---|
| SMA | Charts · Conditions · Sweep · Python · AI Strategy | Trend | period 20 |
| EMA | Charts · Conditions · Sweep · Python · AI Strategy | Trend | period 50 |
| RSI | Charts · Conditions · Sweep · Python · AI Strategy | Momentum | period 14 |
| MACD | Charts · Conditions · Sweep · Python · AI Strategy | Momentum | fast 12, slow 26, signal 9 |
| Bollinger Bands | Charts · Conditions · Sweep · Python · AI Strategy | Volatility | period 20, stdDev 2 |
| ATR | Charts · Conditions · Python · AI Strategy | Volatility | period 14 |
| Stochastic | Charts · Conditions · Sweep · Python · AI Strategy | Momentum | kPeriod 14, dPeriod 3, smoothK 3 |
| ADX | Charts · Conditions · Sweep filter · Python · AI Strategy | Trend | period 14 |
| OBV | Charts · Python · AI Strategy | Volume | none |
| VWAP | Charts · Conditions · Python · AI Strategy | Volume | anchor 1, anchorTime 0, period 20 |
| VWAP Bands | Charts · Conditions · Sweep · Python · AI Strategy | Volume | anchor 1, anchorTime 0, period 20, multiplier 2 |
| Williams %R | Charts · Conditions · Sweep · AI Strategy | Momentum | period 14 |
| Ichimoku | Charts · Conditions · Sweep · AI Strategy | Trend | conversion 9, base 26, spanB 52, displacement 26 |
| MFI | Charts · Conditions · Sweep · Python · AI Strategy | Volume | period 14 |
| Donchian Channel | Charts · Conditions · Sweep · Python · AI Strategy | Volatility | period 20 |
| Heikin-Ashi | Charts · Conditions · Python · AI Strategy | Trend | none |
| Volume Profile | Charts · Conditions · Sweep · Python · AI Strategy | Volume | period 30, bins 24, valueArea 70 |
| Fibonacci Retracement | Charts · Conditions · Sweep · Python · AI Strategy | Volatility | period 30 |
| Pivot Points | Charts · Conditions · AI Strategy | Volatility | anchor 1 |
| Supertrend | Charts · Conditions · Sweep · Python · AI Strategy | Trend | period 10, multiplier 3 |
| ATR % | Charts · Conditions · Sweep filter · Python · AI Strategy | Volatility | period 14 |
| CCI | Charts · Conditions · Sweep · Python · AI Strategy | Momentum | period 20 |
| Rate of Change | Charts · Conditions · Sweep · Python · AI Strategy | Momentum | period 12 |
| Stochastic RSI | Charts · Conditions · Sweep · Python · AI Strategy | Momentum | rsiPeriod 14, stochPeriod 14, smoothK 3, smoothD 3 |
| Aroon | Charts · Conditions · Sweep · Python · AI Strategy | Trend | period 25 |
| Keltner Channel | Charts · Conditions · Sweep · Python · AI Strategy | Volatility | period 20, atrPeriod 10, multiplier 2 |
| Parabolic SAR | Charts · Conditions · Sweep · Python · AI Strategy | Trend | step 0.02, maxStep 0.2 |
| WMA | Charts · Conditions · Sweep · Python · AI Strategy | Trend | period 20 |
| Hull MA | Charts · Conditions · Sweep · Python · AI Strategy | Trend | period 16 |
| Chaikin Money Flow | Charts · Conditions · Sweep · Python · AI Strategy | Volume | period 20 |
| TRIX | Charts · Conditions · Sweep · Python · AI Strategy | Momentum | period 15, signal 9 |
| Ultimate Oscillator | Charts · Conditions · Sweep · Python · AI Strategy | Momentum | short 7, medium 14, long 28 |
| Volume SMA | Conditions · Python · AI Strategy | Volume | period 20 |
Reference
SMA
The average closing price over a fixed number of bars, recalculated each bar. It smooths price into a single line you can compare the current price against.
Price above the line means the recent average is being paid up; below means the opposite. The length sets the lag: a 20-bar average turns within days, a 200-bar average within months, and that lag is the point rather than a flaw.
EMA
A moving average that weights recent bars more heavily than old ones, so it turns sooner than an SMA of the same length. The trade-off is that it also reacts to noise sooner.
Read it like an SMA, but expect earlier turns and more of them. Crossovers between a fast and a slow EMA are the usual way it is traded, with the slower line acting as the trend and the faster one as the trigger.
RSI
RSI compares the size of recent gains to the size of recent losses over a lookback window and scales the result to 0–100. It measures the balance of buying and selling pressure, not the size of the move.
Above 50, gains outweighed losses over the window; below 50, the reverse. The conventional 70/30 lines mark unusually one-sided stretches — they are not a ceiling and a floor. In a strong trend RSI can sit above 70 for weeks.
MACD
The gap between a fast and a slow EMA, plus a signal line that smooths that gap. The histogram is the distance between the two, so it grows while momentum builds and shrinks as it fades.
The histogram crossing zero means the fast average has crossed the slow one — the trend change everyone watches for. Bars growing above zero say the move is still accelerating; shrinking bars say it is running out of push even while price still rises.
Bollinger Bands
A moving average with a band drawn a set number of standard deviations above and below it. The bands widen when the recent range is volatile and squeeze together when it is quiet.
About 95% of closes fall inside two-standard-deviation bands, so touching one is normal, not rare. A squeeze — bands drawing close together — says the market is coiled, without saying which way it will uncoil.
ATR
The average size of a bar's full range, including gaps from the previous close. It is a measure of how much this market moves, in the price units of the symbol itself.
ATR is in the symbol's own price units: an ATR of 900 on Bitcoin and 0.9 on a $30 stock describe similar volatility. That is why it is used for position sizing and stop distances rather than as a buy or sell line — no single threshold works across symbols.
Stochastic
Where the close sits inside the high–low range of the last N bars, as 0–100, with a smoothed second line. Near 100 the market is closing at the top of its recent range, near 0 at the bottom.
%K above 80 says closes are clustering at the top of the recent range, below 20 at the bottom. The %D line is %K smoothed; crossings between them are read as momentum turning, and in a trend the oscillator can stay pinned at an extreme.
ADX
How strong the current trend is, on a 0–100 scale, with no opinion on direction. The +DI and −DI lines it is built from carry the direction; ADX itself only says how decisive the move is.
Below about 20 the market is ranging and trend-following rules tend to whipsaw; above 25 a trend is considered established. Because ADX has no direction, it is used as a gate on other signals rather than as one itself — which is exactly how the sweep's filter axis uses it.
OBV
A running total that adds the bar's volume on up closes and subtracts it on down closes. Its level is arbitrary — only the direction it drifts carries information.
Compare its slope to price. Both rising says the move is being carried by volume; price rising while OBV flattens says fewer participants are behind it. The absolute number means nothing — it depends on where the calculation started.
VWAP
The average price paid over the session, weighted by volume, so heavily traded prices count more. It is the benchmark many desks measure their own fills against.
Price above VWAP means buyers are paying more than the session's volume-weighted average, and below means the opposite. It resets with the anchor you choose — session, week, month, or a bar you pick yourself on the chart — so a VWAP is only meaningful inside the period it is anchored to. Anchoring to an event you can see (a gap, a swing low, the day a trend began) is how it is used away from the intraday session.
VWAP Bands
VWAP with bands drawn a chosen number of standard deviations above and below it. Unlike Bollinger Bands, both the centre line and the width are weighted by volume, so prices that actually traded in size set the band rather than prices merely touched.
Read the distance from the centre in standard deviations, not in price: ±1σ covers the ordinary two-thirds of trading, and a close beyond ±2σ says this bar traded where almost nothing else did. The width itself is the second reading — a band that narrows through the session says the day is agreeing on a price.
Williams %R
The mirror image of the Stochastic, scaled −100 to 0: it measures how far below the recent high the close sits. −100 is the bottom of the range, 0 is the top.
−20 and above is the top of the recent range, −80 and below the bottom. It moves almost identically to the Stochastic %K, so use one or the other rather than treating agreement between them as confirmation.
Ichimoku
A set of five lines that read trend, momentum and support in one picture: two averages that cross, a shaded cloud projected ahead of price, and a lagging line drawn behind it.
The standard reading: the conversion line crossing the base line is the trigger, price above the cloud is the bullish regime and below it bearish, and a thick cloud marks a zone the market has recently disagreed about. The cloud is projected forward, so it shows where support would sit if price gets there.
MFI
RSI's calculation with volume folded in, on a 0–100 scale. Two bars with the same price change score differently if one traded far more, so it reacts to conviction rather than price alone.
The 20/80 lines are the convention, tighter than RSI's 30/70 because volume weighting makes extremes rarer. Divergence is the classic use: price making a new high while MFI does not says the new high is being made on lighter money.
Donchian Channel
The highest high and the lowest low of the previous N bars, drawn as a channel with a midline. The rails are built from bars before the current one, so a close beyond a rail is a genuinely new extreme.
The upper rail is the price that would make a new N-bar high, and the lower rail the price that would make a new N-bar low. A wide channel means the recent range is broad; a narrowing one means the market has been coiling.
Heikin-Ashi
Each bar's close is replaced by the average of its own open, high, low and close, and its open by the midpoint of the previous Heikin-Ashi bar. The result is a smoothed body that keeps one colour through a trend and flips less often: over a year of daily BTC the real body changes colour 155 times and this one 85.
Read the colour, not the level. A run of same-colour bodies is the trend this transform exists to show, and the flip is the signal. The two lines drawn here are the smoothed open and close; they are a body, not a price channel, and the gap between them says how convinced the bar is rather than where support sits.
Volume Profile
Slices the range of the last N bars into bins and asks how much volume traded in each. The busiest bin is the Point of Control; the value area is the contiguous band around it holding the given share of the total. Returns three levels — POC, value-area high and value-area low. The window excludes the current bar.
Read it as where the market has agreed to do business. Price inside the value area is trading where it has been trading; above the high or below the low means it has left that zone, and the POC is the level it most often comes back to. Unlike every other entry in this catalog the reading is a function of price rather than of time — it does not say what happened at bar i, it says which levels the last N bars used.
Fibonacci Retracement
The highest high and lowest low of the last N bars, with the space between them cut at 23.6%, 38.2%, 50%, 61.8% and 78.6%. Levels are measured down from the high, so 23.6% sits just under it and 78.6% just above the low. The window excludes the current bar, which is what keeps the top line reachable.
Read it as depth of pullback. Close above the 23.6% line means the market has given back almost nothing of the window's range; below 61.8% means most of it is gone. The 50% line is the halfway point, and it is exactly Donchian's middle line reached from the other direction - the same two extremes, subdivided instead of averaged.
Pivot Points
Seven flat levels derived from the previous completed session's high, low and close: the pivot itself at (H+L+C)/3, three resistances above it and three supports below. They do not move for the whole session, which is the point - floor traders needed a level they could lean on all day rather than a line that shifts under every bar. The first two sessions in any window stay blank, because the first one may have arrived half-finished and levels built on it would change with the requested period.
Seven horizontal lines that do not move until the session rolls. The middle one is the previous session's (high + low + close) / 3; R1 and S1 sit the same distance out that yesterday's range was, and R2/R3 and S2/S3 extend it. Read the pivot as the day's dividing line - above it the session has been paid up relative to yesterday, below it marked down - and the outer levels as places the crowd has agreed to watch rather than as forecasts.
Supertrend
A trailing stop line set an ATR-based distance from price, which flips to the other side once price closes through it. Below price it marks an uptrend, above it a downtrend, and the flip is the whole signal.
The line is the level that would flip the trend, so it doubles as a trailing stop: in an uptrend it ratchets up under price and never moves down until price closes through it. A larger multiplier or a longer period both mean fewer flips and a wider stop.
ATR %
The same measurement as ATR, divided by the price and expressed as a percent. That one change makes it comparable across symbols: 1.5% means the same thing on a 0 stock and on Bitcoin, where an ATR of 900 and an ATR of 0.9 do not.
Read it as ‘this market moves about X% per bar’. On daily bars, under ~1% is quiet and over ~3% is turbulent for most liquid markets; intraday the same numbers describe something busier. Compare a symbol against its own recent range rather than against another symbol's absolute level.
CCI
How far the typical price sits from its own average, scaled by how far it usually sits. The ±100 lines mark readings that are uncommon rather than impossible — unlike RSI there is no ceiling, so the scale keeps going.
Zero means the typical price is sitting exactly on its own average. The +/-100 lines are where readings start to be uncommon rather than where they stop: about a fifth to a third of bars fall outside them, and a strong trend will hold price beyond +100 for weeks. Read distance as "further from normal than usual", not as a ceiling.
Rate of Change
The percent change from a set number of bars ago. The simplest momentum measure there is, and the one to read first: the rest of the momentum family is a smoothing or a bounding of this idea. Zero is its only meaningful level.
The number IS a percentage: ROC(12) of +4 means the close is 4% above where it was twelve bars ago. Zero is the only structural level, so the reading is used as a sign (are we higher or lower than we were?) and as a magnitude (by how much?), rather than as a band.
Stochastic RSI
The Stochastic formula applied to RSI instead of to price. RSI rarely reaches its own extremes, so measuring where it sits inside its own recent range restores a full 0–100 swing — at the cost of being twice removed from price, which makes it faster and noisier than either parent.
Read it exactly like Stochastic, on a different input: above 80 the RSI is near the top of ITS recent range, below 20 near the bottom. Because RSI is already a smoothed measure, the reading reaches both extremes often — several times a week on an hourly chart is normal, not remarkable.
Aroon
How recently the window's highest high and lowest low were set, each as a percentage. 100 means the extreme is the current bar. Where Donchian reads the values at the edges of a window, this reads their timing, so the two lines crossing marks a change in which extreme is fresher.
Each line is a countdown from 100 to 0: 100 means the extreme is this bar, 0 means it is the oldest bar in the window. Both high means the market is making new highs AND new lows recently — an expanding range. Both low means neither extreme is fresh, which is a market going nowhere.
Keltner Channel
A moving average with rails set an ATR multiple away. Read it against Bollinger Bands: both draw a band around an average, but Bollinger measures the spread of closes while this measures the size of bars, gaps included. They disagree exactly when volatility is changing shape.
The rails sit an ATR multiple from the average, so their distance describes how big the bars have been rather than how scattered the closes are. Compare the two channels: Bollinger inside Keltner is the classic "squeeze", a market whose closes have clustered while its bars have not shrunk as much.
Parabolic SAR
A stop that accelerates toward price and never retreats. Read it against Supertrend: both trail, but this one tightens every time the trend makes a new extreme rather than as volatility falls, so it closes in on a long quiet trend and is knocked out by ordinary noise in a choppy one.
The dots-turned-line sits below price in an uptrend and above it in a downtrend, and the side it is on is the whole reading. Each new extreme in the trend speeds it up, so the gap closes as the move matures — a stop that starts loose and finishes tight.
WMA
A moving average whose weights fall off in a straight line, reaching zero exactly at the edge of the window. It sits between the other two: SMA gives every bar the same say, EMA never fully forgets an old bar, and this one forgets completely but favours the recent.
Read it as a faster SMA of the same length. Because the weights reach zero exactly at the window edge, a bar leaving the window has no residual influence — unlike an EMA, where it never quite goes away. That makes WMA turn sooner than SMA and settle sooner than EMA after a shock.
Hull MA
A moving average that subtracts most of its own lag instead of shortening its window. Doubling a half-length average and taking away the full-length one overshoots the current move; smoothing that overshoot puts it back on price without giving the lag back. Measured on a year of BTC dailies at length 16, the best-matching price lag was 7 bars for SMA, 5 for EMA, 4 for WMA and 2 for this.
Read the slope, not the crossings: the line is smooth enough that its direction is meaningful bar to bar, which is unusual for an average this responsive. On a year of BTC dailies at length 16 the best-matching price lag was 2 bars, against 7 for an SMA of the same length.
Chaikin Money Flow
Where each bar closed inside its own range, weighted by that bar's volume and summed over a window. The volume family's third question: OBV reads the close-to-close direction of the whole bar, MFI reads the typical price, and this one asks only whether the close finished near the high or the low. Ranges from -1 to +1.
Positive means the window's bars, weighted by volume, closed nearer their highs than their lows. The scale is small in practice: readings past +/-0.25 are uncommon on liquid markets, so zero is the level that matters and the magnitude is a strength reading rather than a threshold.
TRIX
The percent change of an EMA that has been smoothed three times, with a signal line of its own. Where MACD subtracts one length from another, this takes a single length and smooths it repeatedly, which removes cycles shorter than the period almost entirely. Because it is a percentage, its zero line means the same thing on every symbol.
It is a percentage per bar, so +0.1 means the smoothed trend is rising a tenth of a percent per bar. Zero is a genuine level and comparable across symbols, which MACD's is not. The signal line crossing is read the way MACD's is — earlier than the zero cross, and noisier.
Ultimate Oscillator
One reading built from three lookbacks at once, weighted 4:2:1 toward the shortest, so an extreme requires the short, medium and long views to agree. Buying pressure is measured from the lower of this bar's low and the previous close, which puts gaps inside the measurement. Bounded 0-100; the conventional lines are 30 and 70.
Above 70 all three lookbacks agree the market has been bought; below 30 they agree it has been sold. Because agreement is required, extremes are rare: on a year of BTC dailies it went above 70 twice and below 30 four times. A reading in the 40s and 50s is the normal state, not an absence of information.
Volume SMA
The average traded volume over a fixed number of bars. It gives you a baseline to call the current bar's volume unusually high or low.
Compare the current bar to the line: two or three times the average is a genuine surge, while a bar below it says the move happened on thin participation. Volume seasonality is real — the same multiple means different things at a session open and mid-afternoon.