Vortex Indicator: The VI+/VI- Trend-Direction Oscillator (vs. ADX)
The Vortex Indicator measures the distance between each bar's extreme and the prior bar's opposite extreme, sums both directions, and divides each by summed true range. The formula worked over five hand-checkable bars, the exact structural difference from Wilder's DMI that no rewrite bothers to state, the 14-versus-Fibonacci settings question answered the way the creators framed it, the crossover read, the resting-order entry from the original paper, and the range conditions where the two lines cross all day and mean nothing.
The honest headline first, because it is the thing most write-ups bury under nine paragraphs of enthusiasm: the Vortex Indicator is not a standalone system, and in a sideways market its two lines will cross back and forth all session while sounding exactly as confident as they do at the start of a genuine trend. Every reference source says some version of this, and the creators' own published examples include real false crossings on real charts. With that stated up front, the rest of the indicator is worth your time, because the construction is genuinely different from the directional tools you already run. It measures the distance between each bar's extreme and the prior bar's oppositeextreme, which is a deliberately odd thing to measure, and the reason it behaves differently from Wilder's Directional Movement Index falls directly out of that choice.
Quick Answer
The Vortex Indicator is a two-line trend-direction tool built by Etienne Botes and Douglas Siepman, published in the January 2010 issue of Technical Analysis of Stocks & Commodities. For each bar, +VM = |current high - prior low| and -VM = |current low - prior high|, alongside TR = max(high - low, |high - prior close|, |low - prior close|). Sum each of the three over the chosen period, then +VI = sum of +VM divided by sum of TR and -VI = sum of -VM divided by sum of TR. Both are unbounded ratios that sit around 1.0, not a 0-to-100 scale. +VI crossing above -VI is the bullish signal and -VI crossing above +VI is the bearish one, with the winning line holding control until the next cross. The default period is 14, chosen to match Wilder's DMI parameter, and the creators themselves suggest 21, 34 or 55 for longer timeframes or to cut false signals. It whipsaws badly in ranges, so it is paired with a trend filter rather than traded alone.
Everything below is the detail that paragraph compresses: the formula worked over five hand-checkable bars so you can verify the arithmetic yourself, the precise structural difference from DMI that the secondary write-ups skip entirely, the settings question answered the way the creators actually framed it rather than flattened into one number, the crossover read plus the entry technique from the original paper, and the specific chart conditions where the two lines produce confident signals about nothing.
What Is the Vortex Indicator in Trading?
Two lines in a sub-pane below price. One tracks upward pressure, one tracks downward pressure, and the one sitting higher is the side currently in control. Platforms label them either VI+ and VI- or +VI and -VI depending on who wrote the study; the two notations refer to the same two values and nothing about the calculation changes between them. This post uses +VI and -VI, because that is the form the formula naturally produces.
The attribution is unusually clean for a technical indicator. Etienne Botes and Douglas Siepman published it in the January 2010 issue of Technical Analysis of Stocks & Commodities, Vol. 28:1, pp.20-30, and the paper is freely available, which means you can read the creators' own reasoning rather than a fourth-hand summary of it. That matters more here than usual, because several of the most useful details about this indicator appear only in the original and got dropped somewhere along the chain of rewrites.
The design idea came from water. The paper credits Viktor Schauberger's observation of vortex flow in streams, where the current moving in one direction along one part of the channel is paired with a counter-current moving the other way, and the relative dominance of the two determines the net movement. Translated to a price chart, the two VM terms are the two currents, and the indicator reports which one is winning. Whether you find the metaphor illuminating or decorative, the arithmetic is the part that actually decides anything, so it is worth going through carefully.
One placement note before the formula. The Vortex Indicator belongs to the same conceptual family as ADX and DMI, and the three answer genuinely different questions about the same trend. ADX answers how strong the trend is, and its full DX walkthrough plus the argument over the 25 threshold lives in the dedicated ADX guide. Vortex answers who is currently winning and on which bar control changed hands. And Aroon, which counts bars rather than distances, answers how long it has been since the last extreme. Running all three is not triple confirmation. It is three different questions with three different answers, and treating them as votes is how people end up very confident and very wrong.
What Is the Vortex Indicator Formula?
+VM = |current high - prior low|
-VM = |current low - prior high|
TR = max(high - low, |high - prior close|, |low - prior close|)
Sum all three over the chosen period, then:
+VI = sum of +VM over N bars / sum of TR over N bars
-VI = sum of -VM over N bars / sum of TR over N bars
Read the two VM lines slowly, because the crossed-over comparison is the whole trick. +VM measures from the current bar's high down to the previous bar's low. That distance is large when today reached well above where yesterday bottomed, which is what upward pressure looks like when you measure it that way. -VMdoes the mirror image, measuring from the current bar's low up to the previous bar's high. Neither term compares a high to a high or a low to a low, and that choice is the reason the indicator is not simply a repackaged DMI.
TR is the ordinary true range, the same one that feeds ATR and the same one the ATR band that flips sides on a close through the line is built from. It exists here purely as a normaliser, so that a $400 stock and a $4 stock produce comparable readings. Without it you would be comparing raw dollar distances across instruments, which tells you nothing.
Here is the arithmetic worked over five bars so you can check it by hand. Real charts use 14, but five is enough to see every term do its job, and the numbers are chosen to stay clean. The prior bar before bar 1 had a high of 100.00, a low of 98.00 and a close of 99.00.
| Bar | High / Low | +VM (high vs prior low) | -VM (low vs prior high) | TR |
|---|---|---|---|---|
| 1 | 101.00 / 99.00 | |101.00 - 98.00| = 3.00 | |99.00 - 100.00| = 1.00 | 2.00 |
| 2 | 102.00 / 100.00 | |102.00 - 99.00| = 3.00 | |100.00 - 101.00| = 1.00 | 2.00 |
| 3 (an inside bar) | 101.50 / 100.20 | |101.50 - 100.00| = 1.50 | |100.20 - 102.00| = 1.80 | 1.30 |
| 4 | 103.00 / 100.50 | |103.00 - 100.20| = 2.80 | |100.50 - 101.50| = 1.00 | 2.50 |
| 5 | 104.00 / 101.80 | |104.00 - 100.50| = 3.50 | |101.80 - 103.00| = 1.20 | 2.20 |
| Summed over the 5 bars | - | 13.80 | 6.00 | 10.00 |
| Divide each sum by the TR sum | - | +VI = 13.80 / 10.00 = 1.38 | -VI = 6.00 / 10.00 = 0.60 | - |
+VI at 1.38 against -VI at 0.60 says the up-current has been carrying roughly two and a third times the down-current's distance across those five bars, normalised against how much range the instrument produced. The two do not total exactly 2.00, and they are not supposed to; that is covered in the FAQ because the assumption is common enough to be worth killing directly.
Note bar 3 in particular. It is an inside bar, printing a lower high and a higher low than bar 2, and both of its VM terms still record real distance. It contributes 1.50 to the up sum and 1.80 to the down sum, so on that bar the down-current actually edged ahead. Hold onto that, because it is about to be the entire DMI comparison.
The Lines Oscillate Around 1.0, Not 0 to 100
This trips people up constantly and it is worth stating flatly. +VI and -VI are ratios. Each is a sum of distances divided by a sum of true ranges, which means neither has a ceiling, neither has a floor other than zero, and neither is scaled into a percentage. In practice both spend their time in the rough neighbourhood of 1.0, with the winning line above it and the losing line below. A reading of 1.25 is not 1.25% of anything. It means the accumulated up-distance over the window came to 125% of the accumulated true range over the same window.
So there is no 70 and no 30 here. There is no overbought zone, no oversold zone, and no midline at 50. Traders who arrive from RSI or from the usual intraday indicator shortlist and try to apply a bounded-oscillator read will wait for levels that the construction simply cannot produce. The reference level is 1.0, and the only thing worth watching is where each line sits relative to that and relative to the other one.
Vortex Indicator vs DMI: What Actually Differs?
Most answers to this question stop at "Vortex gives you two lines while ADX needs a third one for direction," which is true and almost useless. The real difference is mechanical, the creators stated it themselves in the original paper, and it changes what each indicator can see.
Wilder's Directional Movement Index compares each bar's high to the previous high and each bar's low to the previous low, then discards the smaller of the two moves by setting it to zero. On every bar, at most one of +DM and -DM survives. If a bar prints a higher high and a higher low, the down-move is zeroed. If it prints a lower high and a lower low, the up-move is zeroed. And on an inside bar, where the high is lower and the low is higher, both get zeroed and the bar contributes nothing directional at all.
The Vortex Indicator never substitutes a zero anywhere. Both +VM and -VM are computed and summed on every bar without exception, because both terms reach across to the prior bar's opposite extreme and that distance is always positive. An inside bar still records a real number on each side. Go back to bar 3 in the worked table: DMI logs 0 and 0 there, Vortex logs 1.50 and 1.80. Same bar, same data, and one indicator has thrown away information the other kept.
Same five bars. One indicator throws a bar away, the other never does
Whether keeping that information is an improvement depends on what you think an inside bar means. If you read consolidation as genuinely non-directional, Wilder's zeroing is a feature and the Vortex is adding noise. If you read it as two-sided pressure that still tells you something about who is stretching further, the Vortex is preserving signal that DMI deliberately drops. The two positions are both defensible, and nobody has settled it. What is not in dispute is that the two indicators are computing different things, which is the part the shallow comparison obscures.
| Mechanic | Vortex Indicator | Wilder's DMI / ADX |
|---|---|---|
| What each bar contributes | Both +VM and -VM are computed and summed on every single bar, always | At most one of +DM and -DM is non-zero on any bar. The other is set to zero |
| What happens on an inside bar | Both terms still record real distance, because both reach back to the prior bar's extremes | Both terms are zero. The bar contributes nothing directional to either running sum |
| Which two points the up-move is measured between | Current high against the previous low | Current high against the previous high |
| Which two points the down-move is measured between | Current low against the previous high | Previous low against the current low |
| What the plotted lines are | Two lines, +VI and -VI, each a ratio against summed true range | Two lines, +DI and -DI, plus ADX derived from the gap between them |
| Scale | Unbounded ratios that sit around 1.0 in practice | 0 to 100, because the DI values are multiplied by 100 |
| Where trend strength comes from | Read it off the separation between the two lines. There is no third line | ADX, the smoothed normalised gap between +DI and -DI, carries strength on its own |
| Smoothing | Plain sums over the period in the original construction | Wilder's own smoothing applied to the sums, then smoothed again for ADX |
| Attribution and age | Etienne Botes and Douglas Siepman, January 2010 | J. Welles Wilder, 1978, so it has had three extra decades of scrutiny |
| Shared failure mode | Whipsaws in sideways, trendless conditions | Whipsaws in sideways, trendless conditions |
The last row is the one to carry away. Both degrade in the same market condition, so running the pair does not protect you from chop. It gives you two independent descriptions of a trending market and two simultaneous false readings in a flat one, which is the same trap counting genuinely independent signals rather than counting lines on the screen is meant to catch. The age gap in the row above it is worth a second of thought too: 1978 against 2010 means DMI has had three extra decades of people trying to break it in public.
What Are the Best Vortex Indicator Settings?
The default is 14, and the reason is documented rather than arbitrary. Botes and Siepman chose 14 specifically to match the parameter Wilder used for the Directional Movement Index, so that a trader comparing the two studies would be comparing them over the same window. Every charting platform inherited that number, and the StockCharts ChartSchool entry on the Vortex Indicator states it alongside the longer platform variants.
In the same paper the creators then say plainly that longer parameters are more robust and more accurate, and name 21, 34 and 55 for longer timeframes or for filtering out false signals. This is not a disagreement between sources that needs resolving. It is the designers describing a tunable knob and telling you which way they lean, and flattening it into a single "the" setting misrepresents what they wrote. TradingView's own support page for the study describes a working range of roughly 14 to 30, which sits consistently between the two positions.
| Period | Where the number comes from | What it does to signals | Who reaches for it |
|---|---|---|---|
| 14 | The creators' own baseline, chosen to match the parameter Wilder used for DMI | The most crossings of any common setting, and the earliest ones | The reference default and the number every platform ships. Start here before changing anything |
| 21 | The first of the three Fibonacci alternatives named in the original paper | Noticeably fewer crossings. Each VM spike carries less weight against a longer sum | The mildest step up when 14 is producing more signals than the instrument justifies |
| 34 | The middle Fibonacci option from the same paper | Slow enough that short pullbacks inside a trend stop generating crosses | Swing timeframes, and intraday traders who only want the session's dominant swing |
| 55 | The longest option the creators name, recommended for longer timeframes | Very few signals, all of them late. A single bar barely moves the sums | Position and longer swing work. Close to useless as an intraday trigger |
| 14 to 30 | The parameter range TradingView's own support documentation describes | Spans roughly from the reference default up past the first Fibonacci step | A reasonable search band if you are testing rather than picking a number off a page |
| 23 days or 26 weeks | Platform variants that appear in the StockCharts reference treatment | Longer windows aimed at daily and weekly charts rather than intraday bars | Longer-horizon chartists. Neither number is a day-trading setting |
The Fibonacci framing is the creators' own aesthetic preference and nothing in the formula cares. A 22-period Vortex behaves almost identically to a 21-period one. What the period genuinely controls is how much a single bar can move the sums, and therefore how often the two lines are allowed to swap places. Longer window, fewer crossings, later entries. That is the entire trade-off, it is the same trade-off every smoothing parameter presents, and lifting a number out of an article and running it on an instrument nobody tested it on is covered at length in the broader technical analysis overview.
The lines crossed. That is one indicator agreeing with a decision you had already half made.
Upload the screenshot and SnapPChart reads that single image against a fixed rubric, then returns a setup grade, an entry, a stop with the reasoning behind the level, targets, and the reward-to-risk those levels imply. Arguing with the grade is cheaper than arguing with the fill.
Grade this chartHow Do You Trade a Vortex Crossover?
The signal is stated identically by every source. +VI crossing above -VI is the bullish signal. -VI crossing above +VI is the bearish one. The line that ends up on top holds control until the next cross, which means the indicator is always in one state or the other and never neutral. There is no flat reading, no no-trade zone, and no built-in way for it to tell you that nothing is happening. Read that sentence twice, because it is the root of the whipsaw problem further down.
The resting-order entry from the original paper
Botes and Siepman do not enter at the close of the crossing bar. Their described technique is to leave a good-till-canceled stop order at the high of the crossing bar for a long, or at the low of the crossing bar for a short, and let price come and take it. If the cross was real, price clears that extreme and the order fills. If the cross was one of the many that reverse immediately, the order never triggers and the signal costs nothing except the alert.
It is a filter that uses price rather than a second indicator, which is why it is worth knowing even if you do not adopt it. The cost is honest and predictable: on the trades that run hardest, you buy higher than the crossing close, so your stop distance widens and your reward-to-risk on the same target shrinks. This is one named technique from one paper, not a universal rule, and it should be tested on your own instrument before it becomes a habit.
Stops, and why the paper argues for trailing them
The same paper recommends trailing stops derived from ATR over fixed-dollar stop-losses. The reasoning is internally consistent: the Vortex is already normalised by true range, so a stop that also scales with true range keeps the exit logic on the same footing as the entry logic across instruments and across volatility regimes. A flat dollar stop is far too tight on a name that moves three dollars a day and far too loose on one that moves thirty. The mechanics of ratcheting a stop behind a live position, and the specific point at which people start moving it for emotional reasons rather than structural ones, are covered in the walkthrough of trailing-stop rules and where they break down.
One published systematic rule-set
For a concrete example of what pairing looks like in practice, one published systematic treatment goes long only when +VI crosses above -VI and price is above its 50-day moving average, with MACD positive as a third confirmation. The trend filter answers the question the Vortex cannot, which is whether the instrument is trending at all, and the MACD histogram read adds a momentum check on top. That is one source's rule-set rather than a consensus, and every added condition means fewer trades and later entries on the ones that survive. Filters do not improve signals. They make them rarer.
Does the Vortex Indicator Actually Work?
Not on its own, and every source that engages with the question says so. It works in trending markets and produces false signals in sideways, range-bound conditions, which is the identical caveat attached to the accelerating dot trail that reverses on contact and to essentially every trend-following construction on this site. When tools built from completely unrelated arithmetic share one failure mode, the failure mode belongs to the market condition rather than to any of the tools.
The Vortex has a specific version of the problem, and it follows from the always-in-one-state design noted above. Because there is no neutral reading, a flat range does not produce a flat indicator. It produces a continuous sequence of crossings, each one formally identical to the crossing that starts a real trend. The creators' own worked examples on 2008 crude oil include false signals shown and explained rather than edited out, which is more honesty than most indicator papers manage.
| Condition | What the lines do | Why it misleads |
|---|---|---|
| A sideways range | Both lines hug 1.0 and cross each other repeatedly | Every crossing reads as a change of control and none of them lead anywhere. This is the whipsaw case every source on the indicator warns about, and the creators' own worked examples include real ones |
| A wide outside bar | Both +VM and -VM spike at once, because the bar reaches past both prior extremes | The two terms inflate together, so the ratio between the lines can barely move while the underlying numbers double. A violent bar can look like a quiet one in the sub-pane |
| An overnight or earnings gap | One VM term jumps to the size of the gap, since it measures against yesterday's extreme | The reading reports the gap, not what happens after it. Gap sessions are where crossover systems collect their worst fills |
| A crossing that reverses on the next bar | The lines touch, swap order, and swap back | Entering at the close of the crossing bar pays for the cross itself. The original paper's resting-order technique exists specifically because this happens often enough to matter |
| A thin, low-volume name | The prior bar's extremes were set by prints nobody was really trading against | There is no volume term anywhere in the formula. A 200-share poke through yesterday's high feeds the sums exactly as institutional size would |
| A slow, orderly trend | The two lines separate a little and then stay parallel | Steady one-way drift produces modest VM terms bar after bar. The separation never gets dramatic even though the trend is real and tradeable |
| A freshly loaded chart | The sums need the full period to fill before either value means anything | The first 14 bars of a 14-period Vortex are computed against an incomplete window, so early history on a short data load is not comparable to the same bars on a long one |
| Any of the above, read as a 0-100 oscillator | The trader waits for an overbought level that does not exist on this scale | There is no 70 and no 30 here. The reference level is 1.0, the lines are unbounded ratios, and importing RSI habits onto them produces a permanent misread |
The one independent head-to-head against DMI, and what it actually showed
There is a single named, dated, independent test worth quoting, and it is useful precisely because it refuses to hand either indicator a win. Richard Denning ran both against each other in the Traders' Tips section of the same January 2010 Stocks & Commodities issue, using the same 14-day parameter for both, on two different universes. The results split by asset class.
| What was tested | Parameter | Reported outcome |
|---|---|---|
| 38 futures contracts, Jan 3 1978 to Nov 6 2009 | 14 | Vortex outperformed DMI |
| 101 NASDAQ stocks, Jan 2 1992 to Aug 14 2009 | 14 | DMI outperformed Vortex |
Anyone quoting only the first row is selling you something. The same researcher, the same parameter, the same year, and the ranking flips depending on whether you point it at futures or at NASDAQ equities. The honest reading is that the winner is asset-class dependent and neither indicator is generally superior, which is also the most useful thing the study could have told you: test on what you actually trade, because the result does not transfer. The details, along with the full citation, sit on Wikipedia's Vortex indicator entry, which is where the Denning figures are catalogued.
Where Vortex Fits on a Real Chart
It answers one narrow question with real precision: across the last N bars, which side has been stretching further past the other side's prior extreme, and on which bar did that flip. Used as a confirmation on a move you already identified some other way, that is a genuinely useful input, and the inside-bar behaviour means it keeps reading during consolidation where DMI goes quiet. Used as the reason to take a trade, it is a line that is always in one of exactly two states and has no way to tell you the market is doing nothing.
What it structurally cannot contribute is whether the level price is sitting at matters, what the higher timeframe is doing, whether volume confirms the break, or how much you should risk. The ordering that works is the ordinary one: market condition and structure first, then the levels marked before the session starts, then the indicator as the last confirmation to arrive rather than the first input. That is the same sequencing an intraday momentum playbook applies to its own entry triggers, and it is not an accident that every framework arrives at the same order.
What a screenshot read can and cannot see here
Worth being straight about, since this site sells a tool. SnapPChart has no Vortex field. It does not compute +VM, -VM or a summed true range, it holds no lookback window for this indicator, and it carries no +VI or -VI value at all, the way it does for trend and market structure, breaks of structure and changes of character, the moving average stack, the VWAP relationship, MACD, volume behaviour, and support and resistance as price ranges. What the engine does is read a chart screenshot you upload. So if you plot the Vortex on your own platform before taking that screenshot, those two lines are part of what the analysis sees, as geometry drawn in a sub-pane. That is a picture of an indicator, not a verified computed value: it cannot confirm your period is 14 rather than 34, it cannot check whether your platform's implementation added smoothing the original formula does not have, and it cannot tell you whether the crossing bar has closed. What a screenshot-based read genuinely carries and what it only infers from shape is the subject of the wider guide to how AI reads a chart, and a neutral description of what a single chart read covers sits on the AI chart analysis page. If the crossover is the reason you are taking the trade, verifying the crossover stays your job.
For each bar, measure the current high against the prior low and call it +VM, measure the current low against the prior high and call it -VM, take the ordinary true range, sum all three over your period, then divide the two VM sums by the TR sum. That gives +VI and -VI, two unbounded ratios that live around 1.0 rather than on a 0-to-100 scale. Botes and Siepman published it in January 2010. The default 14 exists to match Wilder's DMI parameter, and the creators themselves prefer 21, 34 or 55 for longer timeframes. +VI above -VI is the long signal, the reverse is the short, and the winning line holds control until the next cross. The structural difference from DMI is that DMI zeroes out the smaller move on every bar while the Vortex never substitutes a zero anywhere. The one independent head-to-head split by asset class. And in a flat range the two lines will cross back and forth indefinitely, sounding equally confident every single time.
Frequently Asked Questions
Do +VI and -VI always add up to 2.0?
No, and the fact that they often land close to it is what makes people assume there is a rule. Both lines share the same denominator, the summed true range, so their total is (sum of +VM plus sum of -VM) divided by sum of TR. On any individual bar that ratio can land anywhere. A bar that opens inside the prior range and then expands hard in both directions produces two large VM values against a true range that is smaller than their sum, and the ratio for that bar runs well above 2. A tidy trend bar that gaps slightly in the direction of travel produces one large VM and one small one against a wide true range, and the ratio drops below 2. Averaged across fourteen bars those extremes cancel a lot of the way out, which is why the pair usually sits in the neighbourhood of 2.0 without being pinned to it. The worked five-bar example in this post totals 1.98. Treat the two lines as independent readings that happen to share a denominator, not as a seesaw where one line falling guarantees the other rising by the same amount.
Does the Vortex Indicator repaint?
Not on closed bars, and yes on the bar still forming, which is the same answer every summed-window indicator gives. Once a bar has printed its final high, low and close, its +VM, -VM and TR contributions are fixed forever, so the historical +VI and -VI values you see today are the values you will see next month on a reloaded chart. The live bar is a different matter. Its high and low expand as the bar trades, which means its +VM and -VM terms grow during the bar and its TR grows with them, so the current reading moves continuously. A crossover you saw at 10:41 on a five-minute chart can be gone by 10:44 because the bar pushed to a new high and the +VM term jumped. That is not repainting in the sense of an indicator rewriting history. It is the ordinary behaviour of a value computed from an incomplete bar. If you alert on crossovers, alert on bar close, or you will be pinged repeatedly by states that never survived to become anything.
Is the Vortex Indicator built into MT4, and does it behave differently on forex?
MetaTrader 4 does not ship it in the standard indicator list, so anything called a Vortex indicator on that platform is a custom implementation somebody wrote. That matters more than it sounds, because the formula has two places where implementations quietly diverge: whether the true range term uses the classic three-way maximum with the previous close in it, and whether the author added any smoothing on top of the raw sums. Open the source and check both before you trust the line. On forex specifically the arithmetic works fine, because the calculation only needs a high, a low and a previous close, and every FX bar has all three. The genuine complication is the bar boundary. A 24-hour market has no single agreed daily close, so where your broker cuts the day determines which bar is the prior bar, which determines every +VM and -VM term on a daily chart. Two brokers with different server times produce visibly different daily Vortex readings on the same pair. Intraday charts are much less exposed to this, since the boundary only shifts by the same offset everywhere.
Why is 14 the default if the creators themselves recommend 21, 34 and 55?
Because 14 was chosen as a compatibility decision rather than as a performance one. Botes and Siepman set their baseline at 14 specifically to match the parameter Wilder used for the Directional Movement Index, so that anyone comparing the two studies side by side would be comparing them over the same window rather than over two different windows. It is a like-for-like default, and the charting platforms inherited it. In the same paper the creators then say plainly that longer parameters are more robust and more accurate, and name 21, 34 and 55 as the options they favour for longer timeframes or for filtering out false signals. Those are Fibonacci numbers, which is an aesthetic choice on their part rather than a mathematical necessity; nothing in the formula cares whether the period is 21 or 22. The practical read is that 14 is the comparison baseline, the longer settings are the creators' own recommendation for a quieter line, and the cost of moving up is the same cost as every smoothing decision, which is that every signal arrives later.
Should you wait for +VI to clear 1.10 instead of taking the bare crossover?
It is one documented way to filter, and it costs you exactly what every filter costs. The StockCharts reference treatment describes requiring the crossing line to clear a level such as 1.10, or requiring the losing line to fall under roughly 0.90, before the cross counts as confirmed. Mechanically that demands the winning side accumulate a clear majority of the period's movement rather than simply edging ahead by a hair, which is the state that produces most of the crossings that reverse two bars later. The trade-off is unavoidable. Some genuine moves never produce a reading above 1.10 at all, particularly on slow, orderly trends where no single bar stretches far past the prior bar's extreme, and you will sit those out entirely. Others clear the threshold late enough that the first leg is gone. The original paper solves the same problem from a different direction, leaving a resting order at the extreme of the crossing bar so that price itself has to confirm rather than a second indicator level. Pick one. Running both is just a stricter filter with two ways to miss.
This article is for educational and informational purposes only and is not investment, financial or trading advice. The Vortex Indicator formula expressed as +VM equal to the absolute difference between the current high and the prior low, -VM equal to the absolute difference between the current low and the prior high, true range as the greatest of the current high minus the current low, the absolute difference between the current high and the prior close, and the absolute difference between the current low and the prior close, each summed over the chosen period and the two VM sums then divided by the true range sum to give +VI and -VI, together with the crossover interpretation in which +VI crossing above -VI is read as bullish and -VI crossing above +VI as bearish, the normalised ratio scale oscillating around 1.0, the 14-period default and the stated caution that the indicator is not designed to be used standalone and produces false signals in sideways or trendless markets, are the conventional published accounts reproduced by charting platforms and reference sources including StockCharts ChartSchool and TradingView's support documentation. The attribution to Etienne Botes and Douglas Siepman and the January 2010 publication in Technical Analysis of Stocks & Commodities, Vol. 28:1, pp.20-30, the Viktor Schauberger vortex-flow inspiration, the explanation that the 14-period default was chosen to match Wilder's original Directional Movement Index parameter, the alternative 21, 34 and 55 period guidance for longer timeframes, the good-till-canceled resting-order entry at the high or low of the crossing bar, the preference for ATR-derived trailing stops over fixed-dollar stop-losses, and the statement that the Vortex Indicator never substitutes a zero in its calculation where Wilder's DMI zeroes out the smaller of a bar's up-move and down-move, are taken from that original paper and are the creators' own stated positions rather than independently verified findings. The five-bar worked example is arithmetic derived directly from the stated formula using invented prices at a 5-period setting chosen for legibility; it is not a real security, a real trading session or a market observation, and real charts use a 14-period default. The comparative backtest figures in the limitations section, attributing to Richard Denning in the Traders' Tips section of the January 2010 Stocks & Commoditiesissue a result in which the Vortex Indicator outperformed the Directional Movement Index on a portfolio of 38 futures contracts between January 3 1978 and November 6 2009 while the Directional Movement Index outperformed the Vortex Indicator on 101 NASDAQ stocks between January 2 1992 and August 14 2009, both at a 14-day parameter, are that single named study's reported output as catalogued by a third-party reference; they have not been independently reproduced here, and they are presented as a genuinely mixed, asset-class-dependent result rather than as evidence that either indicator is generally superior. The systematic rule-set pairing a Vortex crossover with a 50-day moving average filter and a positive MACD reading comes from a single published source and is one example approach, not a consensus method. No backtest here is my own, no period, threshold, timeframe, filter or indicator pairing described is claimed to be profitable, and no edge is claimed or implied. Indicator readings describe what price has already done and do not predict what it will do next. Day trading and active trading carry a substantial risk of loss and are not suitable for every investor. SnapPChart grades a static chart screenshot you upload and returns a setup grade, entry, stop, targets and reasoning for that single image; it does not calculate a +VI value, a -VI value, a +VM or -VM term, a summed true range, an ADX value, a DMI value or any other indicator value itself, does not track indicator states it has not been shown, does not scan the market, and does not track your account, positions or P&L. It can only account for indicators that are visibly drawn on the image you upload. Do your own analysis, size positions so that being wrong is survivable, and consider speaking to a licensed financial professional about your own circumstances before trading.
Writes about AI-assisted day trading, technical analysis, and the systems traders actually use to stay disciplined.
+VI just crossed above -VI. In a range those two lines will do that again in forty minutes, and again after that.
The crossover fires in chop as readily as it does at the start of a real move, because the formula has no term that can tell the two apart. Upload the screenshot and SnapPChart reads that one image against a fixed rubric, then returns a setup grade, an entry, a stop with the reasoning behind the level, targets, and the reward-to-risk those levels imply. One skipped bad trade covers it.