Every tool on the chart — where it came from and what it actually says

Where it came from

Averaging is older than the chart. The arithmetic mean over a window was standard in economic statistics in the nineteenth century, and once daily price series were printed in newspapers traders simply carried the idea across. The specific numbers people argue about — 20, 50, 200 — are conventions, not discoveries: they come from the number of trading days in a month, a quarter and a year.

What it actually measures

The mean close of the last N bars, redrawn at every bar, so the line moves by dropping the oldest close and adding the newest.

How it is read

• Slope is the whole message: a rising average says the recent closes are higher than the ones it replaced.
• Its position relative to price is the second read — price above the line means the last N bars have been rising on balance.
• Two averages of different lengths separate when the shorter one turns first. The gap widening is acceleration, the gap closing is the move losing its push.
• Price crossing one average is noise on its own. Price, the short average and the long average all agreeing is the version worth looking at.

Where it misleads

• Averages are lagging by construction: they can only describe bars that have already closed.
• In a range, price cuts through the same average repeatedly and every cross looks meaningful in hindsight. That is the tool failing, not the market hiding something.

Where it came from

The exponential average weights the newest close heaviest and never fully drops an old one out of the calculation. It became practical when charts moved from paper to screens in the 1970s and 1980s: the arithmetic was no harder for a machine, and traders wanted an average that turned sooner than a plain one.

What it actually measures

A running average where today's close counts more than yesterday's, decaying smoothly the further back you look.

How it is read

• Faster than an SMA of the same length, so it turns earlier and also whipsaws more.
• Read the fan: when 21 sits above 55 and both are climbing, the short-term average of price is above the longer one.
• Distance between the two lines is a rough measure of how stretched the recent move is.

Where it misleads

• Earlier is not the same as better — an EMA gives up smoothness to gain speed, and in chop you pay for it in false turns.
• A single close through it settles nothing; the average itself has not changed direction until several bars have.

Where it came from

VWAP came out of institutional execution, not chart reading. Big buyers needed a yardstick for whether their own fills were good, and the fair answer is the average price everybody paid, weighted by size. By the 1980s desks were benchmarked against it, and the line later migrated onto charts.

What it actually measures

The average price every unit actually traded at since an anchor point, weighting each bar by its volume rather than treating all bars alike.

How it is read

• Price above a session VWAP means the marginal buyer since that anchor is, on average, in profit.
• A long stretch with price pinned around it is a market where both sides are filling near fair value — agreement, not direction.
• Where the anchor sits changes everything. This one starts at the first bar on screen, so read it as 'average since the chart began', not as a daily VWAP.

Where it misleads

• VWAP resets. Compared across different anchors it is a different number with the same name.
• Thin volume makes it jumpy — a couple of small prints can drag it more than a hundred quiet ones.

Where it came from

Goichi Hosoda, a Japanese newspaper journalist, spent roughly three decades before the Second World War working out a chart he called Ichimoku Kinko Hyo — 'one glance equilibrium chart' — and published it in 1969. Every one of its five lines is the midpoint of a high-low range rather than an average of closes, which is why it behaves differently from anything built on closing prices.

What it actually measures

Five mid-points of past ranges: the nine-bar and twenty-six-bar midpoints (tenkan and kijun), and two spans drawn twenty-six bars forward, the space between which is the cloud.

How it is read

• The cloud is drawn forward, so it is the one part of this chart that sits in the future — it is where the mid-points land if price only holds still.
• Price above the cloud, cloud ahead of it: the whole tool is arranged the same way.
• Thick cloud means a wide past range; price has to travel further to cross it, and it is where the tool tends to stall.
• Tenkan crossing kijun is the fastest of its three reads and the one that fires most often without meaning much on its own.

Where it misleads

• The forward displacement confuses people: the cloud under today's price was computed from bars twenty-six back.
• On a 1-minute crypto chart the 52-period span is under an hour of data, so the 'cloud' is a much shorter memory than the same tool on a daily stock chart. Settings do not carry across timeframes or assets.

Where it came from

Published by the mechanical engineer J. Welles Wilder Jr. in his 1978 book New Concepts in Technical Trading Systems — the same book that introduced RSI, ATR and ADX. The name says what it draws: a dot that accelerates toward price along a curve, and flips to the other side when price reaches it.

What it actually measures

An accelerating trailing stop. Each dot is a function of the last extreme price and how far the move has run, so the dots tighten as a trend extends.

How it is read

• Dots below price mark the tool's current up state, dots above mark down. The flip is the whole event.
• The dots are a distance, not a level: they show how far price has traveled since the tool last turned.
• Two consecutive sets of dots on the same side, each closer to price than the last, is the tool admitting the move is stretched.

Where it misleads

• It flips constantly in a range. That is its known failure and the reason it is usually paired with a filter rather than used alone.
• It is not a support level. Price trading through a dot is the tool turning, nothing more.

Where it came from

Richard Donchian, a futures trader often called the father of trend following, ran a simple rule in the 1950s and 60s: buy when price makes a new high over the last N periods, sell when it makes a new low. The two lines are just that rule drawn — the highest high and lowest low of the window.

What it actually measures

The extremes of the last N bars, nothing else. No averaging, no weighting.

How it is read

• The channel edges are the boundaries of the recent range: inside them, price has not exceeded anything.
• Its width is a plain read on how much range this market covers over N bars.
• Price closing outside the channel is by definition a new N-bar extreme — the tool's event, and the cheapest possible definition of a breakout.

Where it misleads

• Every strong move starts with a channel break; most channel breaks go nowhere. The tool cannot tell you which, only that an extreme happened.
• Long windows lag badly, short windows fire constantly. The setting is a choice about which mistake you would rather make.

Where it came from

Wilder's 1978 answer to a specific complaint: moving averages tell you direction but not whether there is any trend worth following. The ADX measures the strength of the move without saying which way it goes; the two directional lines beside it carry the direction.

What it actually measures

How consistently price is making progress in one direction, on a scale from nothing to very strong, regardless of up or down.

How it is read

• Low and flat is a market going sideways; the number is deliberately direction-blind, so a violent selloff and a violent rally can both read high.
• Rising ADX with +DI above −DI and rising ADX with −DI above +DI are the two arrangements; the two DIs say which side is being paid.
• The lines crossing is the directional half of the tool. The level of ADX is the strength half.

Where it misleads

• A high ADX does not mean a good entry; it means the move has already been strong. Very high readings describe the end of trends as readily as the middle.
• It is slow by design. Reversal readings arrive after a good part of the new move has printed.

Where it came from

John Bollinger, a technical analyst who worked through the 1980s, wanted bands that adapted to volatility instead of sitting at a fixed distance from price. He set them two standard deviations either side of a moving average and published the method properly in Bollinger on Bollinger Bands in 2001. The statistical idea underneath is far older: standard deviation as a measure of spread.

What it actually measures

The middle line is a 20-period average; the outer lines are that average plus and minus two standard deviations of the same 20 closes. So the bands measure recent dispersion, not direction.

How it is read

• The bands widen when price is moving a lot bar to bar and pinch when it is quiet — that squeeze is the tool's most honest message.
• Touching a band is a statement about distance from the average, not about being too high or too low. In a strong trend price can ride an outer band for many bars.
• A close back inside the band after a ride is where the tool says the move has paused.

Where it misleads

• The classic beginner error: buying the lower band as if it were a floor. It is a measure of spread, and spread expands in the direction of the move.
• With the standard settings a squeeze does not say which way the break comes — only that the market has been unusually quiet.

Where it came from

Chester Keltner published a channel built on a moving average of the typical price in 1960. It was rebuilt in the 1980s by Linda Raschke and others using an exponential average and an ATR-based offset, which is the version almost every platform plots today — worth knowing, because the two tools share a name and not much else.

What it actually measures

The same distance idea as Bollinger, but the width comes from the actual average bar range (ATR) rather than from standard deviation of closes.

How it is read

• Bollinger and Keltner together are a genuine read: Bollinger inside Keltner is a squeeze in both measures of volatility.
• A close outside the channel is a bar range event — price moved more than two average bars away from the average.

Where it misleads

• Standard deviation reacts to a single outlier bar; ATR reacts to it too but spreads it over the window. The two channels disagreeing is information about which kind of quiet the market is in.
• Neither channel is a ceiling. Both are drawn at a chosen multiple and price has no obligation to respect a chosen multiple.

Where it came from

Also from Wilder's 1978 book. He wanted a gap-aware measure of range: a bar that opens far from the previous close has moved further than its own high minus low suggests, so true range takes the largest of three distances and ATR averages it.

What it actually measures

The average distance price travels in a bar, in dollars — not percent, not direction. It is the volatility number the rest of a chart is often scaled against.

How it is read

• Read it as an expectation: if ATR is 0.0040, a normal bar covers about four-tenths of a cent.
• Rising ATR means bars are getting bigger; falling ATR means the market is compressing.
• It is the honest denominator for everything else — a stop, a target or a gap measured in ATR is measured against what this market actually does.

Where it misleads

• It says nothing about direction, and a high ATR is as consistent with a collapse as with a rally.
• Dollar ATR does not compare across assets; percent ATR does. On XRP at four cents, a half-cent ATR is enormous.

Where it came from

Gerald Appel built it in the late 1970s as a way to see a moving-average crossover continuously instead of waiting for the lines to actually cross on the price chart. Two exponential averages are subtracted, and the result is smoothed once more with a signal line. The histogram came later, as a way to see the gap between the two.

What it actually measures

The distance between a 12-period and a 26-period exponential average — positive when the fast one is above the slow one — plus a 9-period average of that distance.

How it is read

• The line crossing zero is the two averages crossing. The line crossing its signal is a shorter-term turn within that.
• The histogram is the gap between them: bars growing means the two averages are separating, bars shrinking means the move is losing its slope.
• Shrinking histogram while price still makes new highs is the most-watched divergence on any chart, and the least reliable.

Where it misleads

• Every setting is lagging twice over — averages of averages. By the time MACD turns, a good part of the move is priced.
• Divergence is a description of momentum, not a reversal signal. Price can keep going with the histogram shrinking for a long time.

Where it came from

George Lane developed it in the 1950s and spent decades teaching it. The premise is narrower than RSI's: over the last N bars price has some range, and where the last close sits inside that range tells you whether buyers or sellers finished the window on top.

What it actually measures

The position of the latest close within the recent high-low range, as a percentage. %D is simply a short average of %K.

How it is read

• Near 100 means the close was near the top of the recent range; near 0 means the bottom. In a range that is the whole story.
• The two lines crossing in the extremes is Lane's original read; the cross in the middle is much weaker and fires constantly.
• Flattening at an extreme while price keeps pressing is the tool saying closes are no longer finishing near the edge — the move is being absorbed.

Where it misleads

• In a trend it pins at the extreme and stays: an overbought stochastic in a strong move is a description of strength. Trading the cross in a trend is how people get run over.
• It uses the close inside the range, so a single wick to a new extreme resets the window and the reading jumps.

Where it came from

Donald Lambert published it in Commodities magazine in 1980, designed for cyclical commodities. It measures how far typical price has strayed from its own average, scaled by the average deviation — a deliberately crude cousin of a z-score.

What it actually measures

Distance from the 20-period average of typical price, divided by mean absolute deviation, so the ±100 lines are a normalization rather than a percentage.

How it is read

• Readings beyond ±100 mean price is unusually far from its own average for this window, by this measure.
• Its original purpose was cyclical: Lambert was looking for recurring swings in commodities, not for a ceiling on price.
• Because it uses mean deviation rather than standard deviation, a few wild bars move it less than they would move Bollinger width.

Where it misleads

• The ±100 lines are not overbought and oversold in the RSI sense; they are distance markers, and in a trend CCI sits beyond them for long stretches.
• The 0.015 constant is cosmetic scaling from the original article. It is why the numbers land near ±100, and nothing more.

Where it came from

Wilder's volume-weighted counterpart to RSI, published in the same 1978 book. It takes the RSI question — how one-sided was the recent movement — and asks it with money instead of price: up-days weighted by their dollar volume against down-days weighted by theirs.

What it actually measures

The ratio of money flowing on up-bars to money flowing on down-bars over the window, mapped to 0–100. Same scale and shape as RSI, different input.

How it is read

• Where RSI and MFI disagree, the disagreement is the read: price rising while dollar flow falls says the up-bars were smaller in size than the down-bars.
• Above 80 and below 20 are its conventional edges, and they are further out than RSI's because volume makes the series more extreme.
• It needs volume to exist at all, so on a thin book it reads whatever few prints there were.

Where it misleads

• It is not the same as RSI plus volume being 'more accurate'. It is a different series with its own noise, and on crypto exchange feeds volume includes wash prints.
• Venue volume is not market volume. A single exchange's MFI is that exchange's tape.

Where it came from

The simplest momentum measure there is: today's price against the price N bars ago, as a percentage. Momentum studies of this kind appear in technical writing from the early twentieth century, long before computers, because it needs nothing but two numbers and a subtraction.

What it actually measures

Percent change over the lookback. Not smoothed, not averaged — the raw distance travelled.

How it is read

• Crossing zero is the closing price passing the close N bars ago, which is the plainest definition of a turn there is.
• Its height is speed: a big reading says the move covered a lot of ground in N bars.
• Because it is unsmoothed it turns before any average-based tool does, at the cost of being noisy.

Where it misleads

• A high ROC after a big move is arithmetic, not a sell signal — the same number means 'fast' and 'already happened'.
• Percent change from a low base is misleading: a recovery off a spike low produces enormous readings that mean little.

Where it came from

The oldest tool on this page. Tickertape readers in the 1900s judged a move by how much stock changed hands behind it, and Dow theory formalized the idea that volume should confirm price. Every modern volume tool is downstream of that one observation.

What it actually measures

Base units traded in the bar — XRP in this case, not dollars. Nothing about direction on its own.

How it is read

• Compare a bar with the bars around it, never with another asset: a quiet bar and a busy one are only meaningful relative to the same market.

Where it misleads

• Exchange volume includes wash and bot activity, and it is one venue's tape, not the whole market's.
• It is not an intensity measure: on a 1-hour chart, a large volume bar is an hour of trading, not a spike.

Where it came from

Joseph Granville published it in Granville's New Key to Stock Market Profits in 1963. His claim was blunt: volume precedes price. The line adds a bar's volume when price closes up and subtracts it when price closes down, so the total drifts with the balance of up and down bars.

What it actually measures

A running total of signed volume. It has no scale of its own — only its own slope and its level relative to where it started matter.

How it is read

• Read the slope, and read it against price's slope. OBV climbing while price goes nowhere says up-bars carried more volume than down-bars did.
• The reading is cumulative from the first bar on screen, so its absolute value is arbitrary — the shape is the tool.
• Long flat stretches mean up and down volume cancel, which is a genuinely quiet market.

Where it misleads

• One enormous bar permanently shifts the level of the line, and it never decays out of the total.
• It ignores the size of the move: a close a tick higher and a close a dollar higher both add the full volume.

Where it came from

The sequence itself is medieval — Leonardo of Pisa introduced it to Europe in 1202 in Liber Abaci, and the ratios near 0.618 and 0.382 come out of it. Applied to markets it is much newer: R. N. Elliott built his wave counts on Fibonacci relationships in the 1930s, and charting software in the 1980s and 90s turned it into lines you can drag onto any chart.

What it actually measures

Nothing is measured. The tool divides a chosen swing into fixed fractions and draws those prices as lines, so it produces candidate levels, not readings.

How it is read

• The swing you choose decides every line. This chart takes the high and low of the window on screen and measures back from whichever end price finished nearer.
• The 50% level is not a Fibonacci ratio at all — it is in every retracement tool anyway, and it is traded as a level by people who think it is one.
• Used as a vocabulary for describing where price pulled back to, it is harmless and useful; 0.618 is just a name for 'a bit more than halfway'.

Where it misleads

• Any price can be made to land on a Fibonacci level by choosing a different swing. It will always have levels, because it draws them everywhere.
• There is no mechanism behind it in a market. Treat it as a ruler drawn where a person chose to put it, not as a structural fact.

Where it came from

Perpetual swaps were introduced by BitMEX in 2016 to give crypto a futures contract with no expiry. Without a settlement date nothing forces the contract back to spot, so the exchange invented a periodic payment between longs and shorts — the funding rate — as the tether.

What it actually measures

Which side of the perpetual is paying the other to hold their position, and how much per interval. Positive funding means longs pay shorts.

How it is read

• Persistently positive funding says the leveraged crowd is paying to stay long. That is a statement about positioning, not about price.
• Extreme funding and flat price is the interesting combination: the crowd is paying up and getting nothing for it.
• It is an exchange-specific number; every venue has its own, and they can disagree.

Where it misleads

• High funding is not automatically bearish. Crowded positioning can stay crowded through moves in either direction.
• It reflects only the perpetual market on that venue — not spot buyers or the rest of the world.

Where it came from

As old as organized futures. Open interest is the count of contracts currently open — every one of them has a long and a short — and for a century it has been read next to volume to tell new positioning apart from closing.

What it actually measures

How many contracts are live right now. It rises when new positions open and falls when positions close.

How it is read

• Price up with open interest up is new money taking the up side; price up with open interest down is shorts closing out.
• A sharp fall in open interest during a violent move is the fingerprint of liquidations, not of conviction.
• It is a stock, not a flow: the level matters less than the change against the price change beside it.

Where it misleads

• It does not tell you who is long. Open interest is one number covering both sides and every venue reports its own.
• Rising open interest says nothing about direction; it only says exposure is building.

Where it came from

Delta comes out of order-flow trading, where every print is labelled by which side crossed the spread — a buy that lifted the offer or a sell that hit the bid. Adding that difference up over time gives CVD, a cumulative line of aggressive buying minus aggressive selling. It is not an old indicator; it exists because exchanges publish trade-by-trade data.

What it actually measures

The running total of market buys minus market sells, by size. It measures who was taking liquidity, not who was providing it.

How it is read

• CVD climbing while price is flat says buyers are lifting offers into resting sell orders and not moving price — someone larger is absorbing.
• CVD falling while price holds says the same in reverse.
• Read the disagreement, not the level; the absolute number depends on where you started the sum.

Where it misleads

• It only sees aggressive orders. Large passive players are invisible in it by construction.
• One venue's tape is not the market, and on crypto venues some prints are not real.

Where it came from

Depth is simply the visible part of the limit order book on an exchange — every resting bid and offer with its size. It became a trading tool with electronic markets; before that it was a screen a specialist could see and nobody else could.

What it actually measures

How much is currently being offered or bid at each price, within a chosen distance from mid.

How it is read

• Thin depth on one side means price can travel further on small size; it is a statement about slippage, not about direction.
• Walls are orders that can be pulled. Resting size is an intention, and intentions in a book cancel cheaply.
• Compare the shape, not the total: bids bunched within a cent say something different from the same size spread out.

Where it misleads

• It is one venue's book, and half of a real market's liquidity is hidden or off-venue.
• Spoofing is common: large orders placed to be seen and cancelled before they trade.