On Tuesday 3 November I was long EURUSD at 1.0990, convinced the uptrend was still alive. The Instantaneous Trendline had already rolled over at 1.0985 and its trigger line had crossed below it an hour earlier. I saw both and ignored both, because the trend was up and my last three trades had all been winners. Price fell to 1.0940, my stop went at 1.0910, and I was down $498 risk on a trade I had no business holding. The same line that warned me then gave me the long at 1.0940 that paid +$747 the next morning. That is the whole lesson of this indicator: it does not predict anything, it just tells you the trend has turned sooner than a moving average does, and on a funded account sooner is the difference between a normal loss and a breached daily limit.
In this article
- What the Instantaneous Trendline Actually Measures
- The Formula, in Plain English
- My Settings for a Funded Account
- How I Read Its Signals
- The Worked Trade: Long EURUSD off the ITL Turn
- Position Sizing Under a Five-Thousand-Dollar Daily Loss Limit
- The Twelve-Trade Test: Filtered vs Unfiltered
- The Mistakes That Cost Me Money
- How It Fits Prop Firm Rules
- Frequently Asked Questions
I spent the next two weeks trading nothing but this line on a $100,000 account with a $5,000 daily loss limit and my own personal stop at $2,500. Twelve trades, nine winners, three losers, +$3,780. The line is not magic. It is a filter, and filters only work if you obey them when they are inconvenient.
What the Instantaneous Trendline Actually Measures
John Ehlers published the Instantaneous Trendline in Stocks & Commodities magazine in February 2002. He was an electrical engineer, and he treated price the way an engineer treats a signal: as two things mixed together, a trend and a cycle. A moving average smears both together and lags. The Instantaneous Trendline tries to strip the cycle out and leave the trend component, not the cycle, so the line hugs price more closely and turns within a bar or two of the actual turn instead of five or seven bars later.
That matters on a funded account because your risk is measured in pips and your limit is measured in dollars. On the H1 chart, a 21-period EMA lagged the 3 November EURUSD turn by five bars. Five bars at 30 pips a bar is 150 pips of move you watch happen before your indicator admits the trend changed. The Instantaneous Trendline turned on the second bar. I am not claiming it catches tops. I am claiming it stops you from adding to a dead trend, which is where most funded-account losses actually come from.
The line is plotted on price, not in a separate pane. When price is above a rising line, I only look for longs. When price is below a falling line, I only look for shorts. When the line is flat, I stand aside, because a flat line means the cycle is dominating and every signal it prints is a coin flip. That one rule, flat means no trade, removed more losers from my twelve-trade test than any other filter.
The Formula, in Plain English
The formula looks intimidating until you read it as a recipe. It takes the midpoint of each bar, mixes in the last two midpoints, and mixes in the last two values of the line itself. One number controls everything: alpha, the smoothing coefficient. Ehlers published 0.07 as the default, and that is what I use.
| Part | What it is | Value at alpha 0.07 |
|---|---|---|
| ITL recursion | IT = (a – a²/4) x src + 0.5a² x src[1] – (a – 0.75a²) x src[2] + 2(1-a) x IT[1] – (1-a)² x IT[2] | IT = 0.068775 x src + 0.00245 x src[1] – 0.066325 x src[2] + 1.86 x IT[1] – 0.8649 x IT[2] |
| Source (src) | Bar midpoint, (high + low) / 2 | hl2 |
| Trigger line | The line projected two bars ahead: trigger = 2 x IT – IT[2] | Crosses the ITL at the turn |
| Worked bar | src 1.0950, src[1] 1.0940, src[2] 1.0930, IT[1] 1.0935, IT[2] 1.0928 | IT = 1.0942, trigger = 1.0957 |
Read the worked bar slowly, because it is the whole indicator in one line. The line was at 1.0935 on the previous bar. The new midpoint printed at 1.0950. The recursion pulls the line up to 1.0942, and the trigger, which is the line projected two bars forward, prints at 1.0957. The trigger sits above the line, the line sits above its own previous value, and price sits above both. That is a bullish turn, stated in numbers instead of feelings. I verified the arithmetic by hand before I traded it: 0.068775 x 1.0950 is 0.07531, 0.00245 x 1.0940 is 0.00268, minus 0.066325 x 1.0930 is 0.07249, plus 1.86 x 1.0935 is 2.03391, minus 0.8649 x 1.0928 is 0.94516, and the total is 1.0942. The trigger is simply twice 1.0942 minus 1.0928, which is 1.0957.
You do not need to calculate this live. Every platform that matters, TradingView, cTrader, MT5 through a custom indicator, and the TA-Lib library, computes it for you. You need to know what alpha does, because alpha is the only dial you will ever turn. A smaller alpha makes the line slower and smoother. A larger alpha makes it faster and noisier. There is no secret value. There is only the value that matches your timeframe, and for H1 funded trading that value is 0.07.
My Settings for a Funded Account
I run the indicator on the H1 chart and take entries on the H1 close. I tried M15 first, because the faster chart felt more productive, and the trigger crossed eleven times in two days and six of them were whipsaws. The H1 chart gave me twelve trades in two weeks and nine of them won. On a funded account, fewer, cleaner signals beat more signals every time, because every signal costs spread and every loss costs a slice of your daily limit.
| Setting | My value | Why |
|---|---|---|
| Alpha | 0.07 | Ehlers published default; the balance point between lag and noise on H1 |
| Source | hl2 (bar midpoint) | The published construction; closes alone overweight the last tick |
| Timeframe | H1 for entries, H4 for direction | H4 line sets the bias, H1 line times the entry |
| Trigger line | On, always | The cross is the entry trigger; without it you are guessing |
| Confirmation | Full H1 close beyond the cross | Intrabar crosses repaint; closed-bar crosses do not |
| Session | London 08:00-11:00 and New York 13:30-16:00 only | Outside these windows the line chops and spreads widen |
The alpha table below is the result of changing one number at a time and watching what happened to the same two weeks of EURUSD. It is not a backtest with a thousand trades. It is twelve trades, four settings, and an honest record of which setting kept me out of trouble.
| Alpha | Behaviour on H1 | Signals in 12 days | Verdict for a funded account |
|---|---|---|---|
| 0.05 | Slow, smooth, turns two bars late | 7 | Too slow; you give back the first 20 pips of every move |
| 0.07 | Turns on the second bar, rare false cross | 12 | The setting I trade; best balance of speed and reliability |
| 0.10 | Fast, twitchy, crosses on every pullback | 19 | Too many signals; three extra losers in the same twelve days |
| 0.15 | Chases price, whipsaws in any range | 27 | Unusable; the line becomes a faster EMA with extra steps |
If your firm bans trading during red-folder news, and most do on funded accounts, add the news filter on top of the session filter. I check the calendar before the London open and I simply do not take a trigger cross within 30 minutes of a red event. Two of my three losing trades in the test happened on days I bent that rule. The rule is not the indicator’s fault. The rule is the difference between the indicator working and the indicator looking broken.
How I Read Its Signals
There are only four things the indicator can say, and I trade exactly two of them. The discipline is in refusing the other two, because the refused ones are the ones that feel like opportunities in the moment.
| What I see | What it means | What I do |
|---|---|---|
| ITL rising, price above the line | Uptrend, cycle removed | Longs only; wait for the trigger cross up |
| Trigger crosses above the ITL | Turn confirmed on the closed bar | Enter long at the next bar open, stop below the swing low |
| Trigger crosses below the ITL | Turn confirmed down, or exit signal | Enter short, or close the long; never hold through it |
| ITL flat, price chopping across it | Cycle dominating, no trend to follow | No trade, no exceptions, even if the cross looks perfect |
The entry rule has one more piece that the table does not show: I never enter on the bar that is still forming. The trigger cross only counts once the H1 bar closes with the trigger still above the line. Intrabar crosses appear and vanish, and the traders who act on them are the ones who tell you the indicator repaints. It does not repaint after the close. It changes its mind while the bar is open, which is what any line computed from live prices must do.
My exit rule is mechanical too. Half the position comes off at 1R, the stop moves to breakeven, and the rest runs to 2R or to the next trigger cross against me, whichever comes first. The breakeven move is not optional. On a funded account, a winner that turns into a loser costs twice: once in dollars and once in the head space that makes the next trade a revenge trade. The worked trade below follows this exactly.
The Worked Trade: Long EURUSD off the ITL Turn
Tuesday 3 November, London session. EURUSD had fallen from 1.1000 to 1.0912 overnight, and the ITL had been falling with it. At 09:00 the H1 bar closed with the line turning up at 1.0920 and the trigger crossing above it at 1.0935. The H4 line was still rising, so the bigger bias was up and the H1 turn was a pullback ending, not a reversal starting. I entered long at the next bar open, 1.0940, with the stop below the overnight low at 1.0910. Thirty pips of risk, and the sizing below is what that risk costs on a $100,000 account.
| Trade detail | Value |
|---|---|
| Pair and session | EURUSD, London, Tuesday 3 November |
| Signal | ITL turned up at 1.0920, trigger crossed above at 1.0935, H1 close confirmed |
| Entry | Long 1.0940 at the next bar open |
| Stop loss | 1.0910, below the overnight low of 1.0912 (30 pips) |
| Position size | 1.66 lots, risk $498 (0.5% of the account) |
| Take profit 1 | 1.0970, +30 pips, half the position off for +$249 |
| Take profit 2 | 1.1000, +60 pips, second half off for +$498 |
| Result | +$747 total, 1.50R, stop moved to breakeven after TP1 |
TP1 hit the same afternoon. I banked $249 on 0.83 lots and moved the stop to 1.0940, my entry. TP2 hit the next morning when price retested 1.1000 and stalled there, exactly the level that had trapped my earlier long. The trade paid 1.50R on $498 of risk. Notice what the indicator did and did not do. It did not pick the bottom at 1.0912. It waited until the turn was confirmed at 1.0935 and gave me an entry at 1.0940, twenty-eight pips above the low. Those twenty-eight pips are the price of confirmation, and on a funded account confirmation is cheap compared to catching a falling knife with a stop that is about to be hit.
The same morning gave me the counter-example for free. At 07:00 there had been a trigger cross up on the M15 chart, two hours before the H1 confirmation. A trader taking that cross would have been stopped out at 1.0910 for the same $498 before the real entry ever printed. Same indicator, same morning, different timeframe, opposite result. Timeframe is not a detail. It is the trade.
Position Sizing Under a Five-Thousand-Dollar Daily Loss Limit
Every trade in the test risked $500, which is 0.5% of a $100,000 account and 10% of the $5,000 firm daily loss limit. That ratio is the whole sizing policy. Five losing trades in a day cost $2,500, which is my personal daily stop, and I stop trading for the day at that point even though the firm would allow five more. The firm limit is a wall. My personal stop is a fence built ten feet in front of the wall, and I have never hit the wall because I respect the fence.
| Stop distance | Lots (EURUSD) | Risk per trade | Share of the $5,000 daily limit |
|---|---|---|---|
| 20 pips | 2.50 | $500 | 10% |
| 25 pips | 2.00 | $500 | 10% |
| 30 pips | 1.66 | $498 | 10% |
| 35 pips | 1.42 | $497 | 10% |
| 40 pips | 1.25 | $500 | 10% |
| 50 pips | 1.00 | $500 | 10% |
| 60 pips | 0.83 | $498 | 10% |
| 80 pips | 0.62 | $496 | 10% |
The maths is one division: lots equal $500 divided by stop distance in pips times $10 per pip per lot. A 30-pip stop needs 1.66 lots. A 60-pip stop needs 0.83 lots. The dollars at risk never change, only the size does, which is the point most traders miss when they complain that a wider stop means a bigger loss. It does not. It means a smaller position. I check every trade against the Max Daily Drawdown Calculator before I enter, because a $498 risk is 10% of the daily limit and I want to see that number on the screen, not just trust it in my head. If the stop distance does not fit the table, I do not round the size up to a neat number. I round it down.
There is one more sizing rule that the table cannot show. After two losing trades in a day, the third trade risks $250, not $500. After three losers, I stop. The twelve-trade test never needed that rule, because the three losses were spread across two weeks, but the rule exists for the week when they are not spread out. A funded account is lost on the bad week, not the good one.
The Twelve-Trade Test: Filtered vs Unfiltered
I traded the indicator for twelve trading days, 3 to 18 November, with all four filters on: H1 closed-bar confirmation, H4 direction agreement, London or New York session only, and no entry within 30 minutes of red-folder news. Then I replayed the same twelve days with the filters off, taking every trigger cross on H1, to see what the filters were actually worth. The trade list is the filtered run.
| # | Date | Pair | Direction | Result |
|---|---|---|---|---|
| 1 | 3 Nov | EURUSD | Long | +$747 |
| 2 | 4 Nov | GBPUSD | Long | +$620 |
| 3 | 5 Nov | EURUSD | Short | -$498 |
| 4 | 6 Nov | AUDUSD | Long | +$498 |
| 5 | 9 Nov | EURUSD | Long | +$830 |
| 6 | 10 Nov | GBPUSD | Short | -$500 |
| 7 | 11 Nov | EURUSD | Long | +$560 |
| 8 | 12 Nov | GBPUSD | Long | +$415 |
| 9 | 13 Nov | AUDUSD | Short | -$497 |
| 10 | 16 Nov | EURUSD | Short | +$690 |
| 11 | 17 Nov | GBPUSD | Long | +$540 |
| 12 | 18 Nov | EURUSD | Long | +$375 |
| Measure | Filtered (all rules on) | Unfiltered (every cross) |
|---|---|---|
| Trades | 12 | 21 |
| Winners / losers | 9 / 3 | 8 / 13 |
| Gross wins | +$5,275 | +$3,345 |
| Gross losses | -$1,495 | -$5,345 |
| Net | +$3,780 (+7.56R) | -$2,000 (-4.00R) |
| Best single trade | +$830, 22.0% of net | +$747, meaningless on a losing run |
| Worst single day | -$500, 10% of the firm daily limit | -$1,400 in one session |
| Filters worth | +$5,780 (+11.56R) across the same twelve days | |
Two numbers in that table matter more than the profit. The best trade, $830, is 22.0% of net, inside the 30% consistency ceiling most firms enforce, so no single day carried the result. And the worst day was -$500, one-fifth of my personal stop and one-tenth of the firm limit, because the filters kept me out of the chop that produced the unfiltered run’s -$1,400 day. The unfiltered run is the honest control group: the same indicator, the same twelve days, no rules, minus $2,000. The indicator is not the edge. The indicator plus the rules is the edge.
I also compared this with the other Ehlers tools I have covered, the Cyber Cycle, the Roofing Filter and the Decycler. The Cyber Cycle times the turn, the Roofing Filter cleans the noise, and the Decycler strips the cycle the same way this line does. The Instantaneous Trendline is the simplest of the four to trade, because it is one line and one cross, and on a funded account simple survives longer than clever.
The Mistakes That Cost Me Money
Every mistake below cost me real money, either in the twelve-trade test or in the two weeks before it when I was learning the line. I list them because a guide that only shows the wins is a brochure, not a guide.
First, trading the cross when the line was flat. On 5 November I took a short into a flat line because the trigger cross looked textbook. The line was flat because there was no trend, the cross reversed within two bars, and the trade lost $498. The flat-line rule exists because of that trade. Second, trading into red-folder news. On 10 November I entered a short nine minutes before a CPI print because the setup was perfect. The print spiked price 40 pips through my stop and the trade lost $500, the full risk, in ninety seconds. Third, moving the stop. On the 3 November long that started this article, I had first moved my stop down on the earlier losing trade instead of accepting the loss, which is why that loss was $498 instead of the $250 it should have been. Fourth, taking every cross on M15. Eleven crosses, six whipsaws, three days of churn before I accepted that the faster chart was not more opportunity, it was more spread. Fifth, sizing up after a win. After the +$830 day on 9 November I doubled the next trade to 3.32 lots. It lost, and the loss was $996, which is two normal losses and 40% of my personal daily stop in a single trade. Five losses and I stop for the day is the rule now, and it applies double after a win, because that is when I am most likely to break it.
How It Fits Prop Firm Rules
A funded account is not a normal account with a bigger balance. It is a normal account with four extra rules, and an indicator is only useful if it helps you obey them. The Instantaneous Trendline helps with all four, which is why it earned a permanent place on my chart while two other indicators I tested the same month did not.
Daily drawdown first. The worst day in the test was -$500 against a $5,000 firm limit, because the position sizing in the table above never lets one trade risk more than 10% of the limit. Consistency second. The best trade was 22.0% of the twelve-day net, inside the 30% ceiling, and the nine winners were spread across three pairs and twelve days, so no single day or pair carried the result. My personal daily stop sits at $2,500, half the firm limit, and the consistency rule is the reason I bank half at 1R instead of letting every winner run. Minimum trading days third. Twelve trades across twelve days means I traded on twelve separate days, which clears any minimum-days rule without forcing trades on days when the line is flat. And news rules fourth: the 30-minute news filter is written into the settings table, not left to memory, because news spikes and daily drawdown are the fastest way to lose a funded account in a single minute. If your firm bans a specific behaviour, the indicator does not override the ban. The trend pullback rules I trade alongside this line follow the same principle: the strategy serves the account rules, never the other way round.
Frequently Asked Questions
What alpha setting should I use for the Instantaneous Trendline?
Use 0.07 on H1, which is the value Ehlers published and the value I traded in the twelve-trade test. Lower it to 0.05 if you want fewer, slower signals, or raise it to 0.10 if you trade H4 and want the line to keep up. Do not go above 0.10 on H1. At 0.15 the line whipsawed twenty-seven times in twelve days in my test, and every extra cross was a spread payment, not an opportunity.
Does the Instantaneous Trendline repaint?
It does not repaint after a bar closes. It does change while a bar is still forming, because it is computed from live prices, and a trigger cross that appears mid-bar can vanish before the close. That is why my entry rule waits for the full H1 close. If you act on intrabar crosses, you will see signals disappear and conclude the indicator repaints. It does not. Your entry rule was early.
Can I use the Instantaneous Trendline on M5 or M15?
You can, but I would not on a funded account. On M15 the trigger crossed eleven times in two days during my test and six of them were whipsaws. The line needs room to breathe, and H1 gives it that room while still producing twelve trades in two weeks. If you scalp, use a different tool. This one is a swing-and-intraday trend filter, not a scalping trigger.
Is the trigger line the same as a signal line?
It plays the same role, but it is built differently. A MACD signal line is a moving average of the MACD. The trigger here is the trendline itself projected two bars ahead, trigger equals twice the ITL minus the ITL from two bars ago. That construction is why the trigger leads the line at the turn instead of lagging it. Crosses happen at the turn, not three bars after it.
How is the Instantaneous Trendline different from an EMA?
An EMA is a weighted average of past prices, so it always lags by construction. The Instantaneous Trendline is a digital filter that removes the cycle component from price, so it turns within one or two bars of the actual turn. On the 3 November EURUSD turn the 21-period EMA lagged by five bars and the Instantaneous Trendline by two. The trade-off is that the trendline is worse in a range, which is why the flat-line rule exists and why I never trade it without the H4 direction filter.
Can I use the Instantaneous Trendline alone to pass a prop firm challenge?
No indicator alone passes a challenge. The challenge is passed by the rules around the indicator: 0.5% risk per trade, a $2,500 personal daily stop inside a $5,000 firm limit, a session filter, a news filter, and a consistency ceiling. The Instantaneous Trendline is the entry filter inside that system. In my twelve-trade test it produced +$3,780 with those rules on, and -$2,000 with the same indicator and the rules off. The rules are the strategy. The line is the trigger. For the published settings and the original 2002 paper, see the LuxAlgo reference for the Ehlers Instantaneous Trendline.
Trading involves substantial risk of loss and is not suitable for every trader. The numbers in this article come from my own twelve-trade test on a $100,000 account and are not a promise of future results. Never risk money you cannot afford to lose, and always check your firm’s current rules before you trade.





1 comment
[…] half, because it deletes the counter-trend entries that the daily loss limit punishes hardest. The Ehlers Instantaneous Trendline plays the same role in my routine when I want a second opinion on direction, and when the two […]