VALYNOX
VALYNOXSTRATEGY RESEARCH TECHNOLOGY
AlphaPrimeAlphaPrime
ACADEMY / Indicators and signals

Vortex indicator: crossover trend signals and an NQ backtest

The Vortex indicator signals an uptrend when VI+ is above VI- and a downtrend when VI- is above VI+, so the crossover is the trade signal. The signal only pays when a trend follows the cross, which a backtest after costs must show; in a sideways range the lines cross every few bars and each cross is a small loss.

By · Published · updated · Editorial policy

Contents
  1. How does the Vortex indicator signal a trend?
  2. How are VI+ and VI- calculated?
  3. What are the crossover rules for a Vortex strategy?
  4. Vortex vs ADX: what is the difference?
  5. What does a backtest of the Vortex crossover on NQ show?
  6. Why does the Vortex crossover lose in choppy markets?
  7. How do you test whether the Vortex signal holds up?
  8. Frequently asked questions
  9. References

How does the Vortex indicator signal a trend?

The Vortex indicator signals an uptrend when the VI+ line sits above the VI- line and a downtrend when VI- sits above VI+. The crossover of the two lines is the trade signal. The signal is only useful when a trend follows the cross, and only a backtest after costs can show how often that happens.

The Vortex indicator is a trend indicator built from two lines, VI+ and VI-, that measure upward and downward movement between consecutive bars over a lookback period. Think of two teams in a tug of war. VI+ is the pull of the buyers, VI- is the pull of the sellers, and a cross is the moment one team starts to win. The indicator does not know how long the winning team will keep pulling. That is the whole argument of this article: the cross tells you direction has changed, and the backtest tells you whether that change lasted long enough to pay for the trade.

This puts the Vortex indicator in the same family as other directional tools covered in our guide to which technical analysis tools and indicators you can test: it is a rule you can write down, run on old prices and judge by trades after costs, not a picture you interpret.

How are VI+ and VI- calculated?

VI+ and VI- come from two distances between consecutive bars, each summed over the lookback and divided by the summed true range, which is why both lines usually sit near 1.0 and cross when direction changes. The lookback is the number of bars the sums cover; 14 is the common default.

Two movements are measured on every bar:

  • VM+ (upward movement) is the distance from yesterday's low to today's high.
  • VM- (downward movement) is the distance from yesterday's high to today's low.

A bar that pushes higher than the previous bar's range stretches VM+; a bar that pushes lower stretches VM-. True range is the largest of three distances: today's high to today's low, today's high to yesterday's close, and today's low to yesterday's close (Wilder, 1978). It is the same measure the ATR uses. The one formula to know:

VI+ = sum of VM+ over N bars ÷ sum of true range over N bars

Here N is the lookback, sum of VM+ adds the upward movements of the last N bars, and sum of true range adds the true ranges of the same N bars. VI- is calculated the same way with VM-. Dividing by true range scales both lines by how much the market moved in total, so a reading of 1.20 means roughly the same thing on a calm week and a wild one.

The two Vortex movements on two consecutive bars: VM+ runs from yesterday's low up to today's high, VM- runs from yesterday's high down to today's low, and VI+ is the sum of VM+ over N bars divided by the sum of true range, with VI- calculated the same way from VM- Vortex measures two moves From yesterday's bar to today's Yesterday Today VM+ up VM− down VI+ = sum of VM+ ÷ sum of true range Both sums over N bars; VI− uses VM−
The two Vortex movements on two consecutive bars. The longer arrow over the lookback decides which line sits on top.

Hypothetical example

Yesterday's bar: high 20,100, low 20,000, close 20,080. Today's bar: high 20,150, low 20,050. VM+ is 20,150 − 20,000 = 150 points. VM- is 20,100 − 20,050 = 50 points. True range is the largest of 100 (today's range), 70 (high to yesterday's close) and 30 (low to yesterday's close), so 100. Over a 14-bar lookback, suppose the sums are 1,800 for VM+, 1,200 for VM- and 1,500 for true range. VI+ is 1,800 ÷ 1,500 = 1.20 and VI- is 1,200 ÷ 1,500 = 0.80. VI+ is on top, so the indicator reads uptrend.

Because the sums roll forward one bar at a time, a single large bar changes both lines for the next N bars and then drops out. This is why the lines move smoothly and why a shorter lookback crosses more often than a longer one.

What are the crossover rules for a Vortex strategy?

A testable Vortex strategy has four written rules in order, and the lookback is fixed before the test starts. Written rules matter because a cross can be seen only at the close of a bar; a rule that fills on the same bar it signals on uses a price nobody could have traded.

  1. Long entry: when VI+ crosses above VI- at the close of a bar, buy at the open of the next bar.
  2. Short entry: when VI- crosses above VI+ at the close of a bar, sell short at the open of the next bar.
  3. Exit: the opposite cross closes the open position. In the plain version it also opens the new position, so the strategy is always in the market.
  4. Lookback: choose it before the test (14 is the common default) and do not change it after seeing a result. A changed lookback is a new candidate.

Next-bar execution is the practice of taking the signal at the bar close and filling at the next bar's open, and it is the honest default for any close-based rule. On daily bars this means the fill is the next day's opening price, which can gap away from the close. That gap is part of the real cost of the signal and should stay in the test.

Timeline of one Vortex trade: at the bar 1 close VI+ crosses above VI-, at the bar 2 open the strategy buys, it holds long through bars 3 to N, at the bar N close VI- crosses above VI+, and at the bar N+1 open it exits the long and goes short Read at close, fill next open One trade, bar by bar Bar 1 close Cross: VI+ on top Bar 2 open Buy, go long Bars 3 to N Hold long Bar N close Cross: VI− on top Bar N+1 open Sell, go short The cross is known only at the close. So: fill every signal at the next open.
From cross to fill to exit. The signal is read at the close; the trade happens at the next open.

Many traders add a stop, a target or a trend filter at this point. Leave them out of the first test. The first test has one job: to show whether the bare cross, after costs, is worth anything. If it is not, a filter that rescues it has to be judged as a new idea, not as an improvement.

Vortex vs ADX: what is the difference?

ADX measures trend strength without direction and needs the DMI lines for direction, while the Vortex indicator gives direction and a rough strength reading in two lines. The result is that Vortex crosses sooner and more often, and ADX turns later and less often.

ADX is the average directional index, a single line between 0 and 100 that rises when a market trends in either direction and falls when it does not (Wilder, 1978). DMI is the directional movement system that produces the +DI and −DI lines; ADX is built from them, and traders read direction from which DI line is on top. Vortex uses a similar raw idea, movement beyond the previous bar's range, but keeps the two sides as separate lines scaled near 1.0 instead of folding them into one strength number. The gap between VI+ and VI- is the rough strength reading: lines far apart mean one side has been pulling hard for most of the lookback.

FeatureVortex indicatorADX
What it measuresDirection and movementTrend strength only
Gives directionYesNo without DMI
Output rangeAround 1.00 to 100
Number of linesTwo lines, VI+ and VI-One line plus DI lines
Typical signalCrossoverRising above 20 or 25
Speed of turnEarly and frequentLate and infrequent
Vortex compared with ADX: Vortex measures direction, ADX only strength with direction from DMI; Vortex sits around 1.0 with two lines VI+ and VI-, ADX runs 0 to 100 with one line plus DI lines; Vortex signals on a crossover, ADX on rising above 20 or 25; Vortex turns early and often, ADX late and rarely Vortex adds direction Vortex ADX Measures Direction Strength only Direction? Yes Only via DMI Range Around 1.0 0 to 100 Lines VI+ and VI− 1 + DI lines Signal Crossover Above 20–25 Turns Early, often Late, rarely Vortex turns early and often. So: count the false starts too.
Vortex answers "which way, and roughly how hard"; ADX answers "how hard" and leaves direction to DMI.

Early and frequent is a trade-off, not an advantage. Being early means catching more of each trend; being frequent means paying for more false starts. The same trade-off sits inside every crossover tool, including the MACD indicator's signal-line crosses, and the only way to know which side of the trade-off wins on your market is to count the trades.

What does a backtest of the Vortex crossover on NQ show?

The one candidate kept out of 1,024 tried on NQ daily bars from 2019-01-02 to 2025-06-30 stayed positive after $30 of costs per trade, with 68 trades and a drawdown of $38,185. Because it was picked after the fact for its return against its drawdown, the result describes what happened on that data, not what to expect.

What the run tested: NQ (E-mini Nasdaq-100 futures) daily bars over about six and a half years, long and short, one contract, with a Vortex crossover rule as the only entry and exit, no stop, no target and no filter. The search tried 1,024 variants of the rule's settings and kept the one with the best return against its drawdown. AlphaPrime (Valynox) ran every one of those candidates through its backtesting with the $20 commission, $10 slippage and next-bar execution already applied, so the ranking that chose the kept candidate was a ranking of net results, not gross ones.

Source: AlphaPrime product engine (CPU), run 2026-10-31. NQ daily bars, 2019-01-02 to 2025-06-30, long and short, 1 contract, $20 commission and $10 slippage per trade.

NumberResult
Net profit (closed trades)$61,655
Max drawdown$38,185
Trades68
Win rate63.24%
Profit factor1.597
Average trade$906.69
Time in market7.45%
Vortex versus ADX: the vortex shows direction and flips fast, with whipsaws as its risk; ADX shows strength only and flips later, with late entries as its risk Vortex flips faster than ADX Earlier entries, more false ones Vortex ADX Shows direction Flips fast Risk: whipsaws Strength only Flips later Risk: late entry Speed buys entries and whipsaws. So: try ADX as a filter.
Look at how much of the climb comes from a few long stretches, how long the flat and losing periods lasted, and whether the out-of-sample stretch looks like the rest of the curve. NQ · daily · 2019-01-02 to 2025-06-30 · CPU engine backtest

Three things in the table deserve a closer look. First, 68 trades over six and a half years is a small sample; a per-trade average built from 68 trades moves a long way if a handful of trades are replaced. Second, the max drawdown of $38,185 is more than half the net profit of $61,655, so the account that traded this would have spent a long time under water for the result it got. Third, time in market of 7.45% means this candidate was flat most days, so it is not the always-in version from the rules section; the settings the search kept left it out of the market far more often than in it. On NQ the point value is $20 and the minimum tick is 0.25 points, or $5.00, by the exchange's current contract specifications, so the $30 of costs charged per trade is 1.5 index points; exchanges can change these values, and the CME contract page is the place to confirm them.

Bailey, Borwein, López de Prado and Zhu (2014) show why the kept candidate's numbers are flattered: when many variants are tried on one history and the best is kept, part of the best result is noise that happened to line up with that history. All the candidates the search tried and did not keep are the context for the one that it did. Nothing in this table says that the Vortex indicator earned $61,655 or that a reader can expect it; it says that one setting of one rule, chosen after the fact, did this on that data.

Why does the Vortex crossover lose in choppy markets?

In a sideways range VI+ and VI- cross each other every few bars, so each cross opens a position just before price turns back. A string of small losses plus costs is the price paid for being early on the trends that do run.

The mechanism follows from the calculation. In a range, bars push above the previous high and below the previous low in turn, so VM+ and VM- take turns being the larger movement. Both sums hover around the same value, both lines sit close to 1.0, and a small push either way flips which line is on top. Every flip is a signal. Every signal is a fill at the next open, two costs, and a position that faces the next flip a few bars later. The same thing happens to the stochastic oscillator's %K and %D crosses in a range: a crossover tool cannot tell a small push from the start of a trend until after the fact.

Illustration: in a sideways range VI+ and VI- cross five times and each of the five trades closes a small loss; in the trend on the right the lines cross once and one long trade wins, so many small losses pay for one trend Small losses, one big trend Illustration: a range, then a trend Range: 5 losses Trend: 1 win VI+ VI− Many small losses pay for one trend. So: judge trend size, not win rate.
In the range the lines cross five times and every cross is a small loss. The trend on the right is one cross and one trade. The strategy's result is the second minus the first.

Hypothetical example

Suppose a sideways month produces five crosses. Each trade loses 20 NQ points, or $400 at $20 per point, plus $30 of costs, so the month costs 5 × $430 = $2,150. The next month trends, produces one cross, and the single trade gains 180 points, or $3,600, less $30. The two months together net $1,420. Change the trend to 100 points and the two months together lose $180. The result of a Vortex strategy is decided by how long the trends run, not by how often the cross is right.

This is also why the win rate in the NQ table should not be read on its own. A crossover strategy's result is a few large trades against many small ones, so the size of the winners decides more than their count. The profit factor and the average trade carry more of the story than the win rate does.

How do you test whether the Vortex signal holds up?

Run the fixed rules on data the search never used, check that neighbouring lookbacks behave alike, double the slippage, and count trades by market condition, all before adding any filter. The order matters because each step answers a question the later steps assume.

  1. Write the rules and the lookback down before the first run: the four rules from the crossover section, the lookback, the costs and next-bar execution.
  2. Run the rules unchanged on out-of-sample data. Out-of-sample data is the part of the history that was not used to choose any setting. How to split it, and how many trades the out-of-sample part needs, is covered in in-sample vs out-of-sample: how to split and how much is enough. For a daily-bar crossover with few trades per year, this step alone can take most of the history.
  3. Run neighbouring lookbacks, for example 10, 12, 14, 16 and 18, and double the slippage. If 14 is good and 12 and 16 are poor, the result belongs to the noise of that history, not to the rule. If doubling slippage from $10 to $20 per trade removes most of the profit, the rule was depending on cheap fills.
  4. Count trades and results by trending and sideways stretches. Mark the stretches by eye or with a simple range measure, then tally the trades in each. The crossover should lose in the ranges and win in the trends; if it also loses in the trends, no filter will help.
Four checks in order: 1 write the rules and fix the lookback, 2 run on unseen out-of-sample data, 3 run nearby lookbacks and double the slippage, 4 count trades in trending and sideways stretches; a dashed arrow leads back to step 1 because a changed rule is a new candidate, not a repair Four checks, in this order 1 2 3 4 Write rules Fix the lookback Unseen data Out-of-sample run Near lookbacks And 2× slippage Trend vs range Trades in each Changed the rule? Back to step 1. So: a new candidate, not a repair.
The checks in order. A rule changed after step 4 goes back to step 1 as a new candidate, not as a repair.

Only after these four steps does it make sense to try a filter, such as a minimum gap between VI+ and VI- before a cross counts. The filter then faces the same four steps on its own. The same discipline applies to any packaged signal you did not build yourself; our test of whether the LuxAlgo indicator repaints and how to test its signals follows the same order. Written rules first, unseen data second, neighbours and doubled costs third, a count by market condition fourth.

The next step is one run: write the four crossover rules and a lookback of 14 on paper, pick a block of your own NQ daily bars that no earlier test of yours has touched, charge at least $30 per trade, and read the trade count and the average trade after costs before you look at anything else.

Frequently asked questions

How is the Vortex indicator calculated?

On each bar, measure the distance from yesterday's low to today's high (VM+) and from yesterday's high to today's low (VM-). Sum each over the lookback, commonly 14 bars, and divide each sum by the sum of true range over the same bars. The two results are VI+ and VI-. Both usually sit near 1.0, and the line on top shows the direction that has moved more over the lookback.

How do you use the Vortex crossover as a signal?

VI+ crossing above VI- at a bar close is the long signal, and VI- crossing above VI+ is the short signal. Fill at the open of the next bar, not at the close that showed the cross, because the cross is only known once the bar has closed. The opposite cross exits the position. Write these rules and the lookback down before testing, and judge the result by trades after commission and slippage.

What is the difference between Vortex and ADX?

ADX is one line from 0 to 100 that measures how strongly a market is trending without saying which way; direction comes from the +DI and −DI lines of DMI. The Vortex indicator keeps the two directions as separate lines, VI+ and VI-, scaled around 1.0, so it gives direction and a rough strength reading at once. Because of this, Vortex crosses earlier and more often than ADX turns, which means more trades, more false starts, and more of each trend caught.

What is the best lookback for the Vortex indicator?

There is no single best lookback, and a lookback that is good on one history can be poor on the next. Start with the common default of 14, then run neighbours such as 10, 12, 16 and 18 on the same data. If the results of neighbouring lookbacks are similar, the rule is reasonably stable; if only one lookback works, the result belongs to that history rather than to the rule.

Does the Vortex indicator repaint?

The completed values do not change once a bar has closed, because they depend only on finished bars. The value of the current bar does change while that bar is still forming, so a cross that appears during the bar can disappear before the close. A backtest that reads the cross at the close and fills at the next open avoids this problem.

References

  • Wilder, J. W. (1978). New Concepts in Technical Trading Systems. Trend Research.
  • Bailey, D. H., Borwein, J., López de Prado, M., & Zhu, Q. J. (2014). Pseudo-Mathematics and Financial Charlatanism: The Effects of Backtest Overfitting on Out-of-Sample Performance. Notices of the American Mathematical Society, 61(5), 458–471.
  • CME Group. E-mini Nasdaq-100 Futures contract specifications (contract size $20 × index, minimum tick 0.25 index points = $5.00). cmegroup.com/markets/equities/nasdaq/e-mini-nasdaq-100.contractSpecs.html