Moonshine Pivot Trend MT4
- インディケータ
- バージョン: 1.1
- アクティベーション: 11
Moonshine Pivot Trend is a trend indicator for MT4 and MT5 whose trailing stop sits still for days at a time. On gold H1 it produced 153 signals in four and a half years, about 33 a year. If that sounds low, that's the argument I want to make before you look at a screenshot and decide something is broken.
Most ATR trend tools centre their band on the current candle, so every new bar moves the anchor and a quiet stretch nudges you out of a trend that never broke. This one anchors to confirmed swing pivots, which is why the line climbs in flat steps instead of a slope.
What the anchor actually is
| What a trailing stop needs | What most ATR trend tools do | What Moonshine Pivot Trend does |
|---|---|---|
| An anchor that ignores ordinary candles | Centre the band on the current bar's midpoint, so every candle drags the line | The centre moves only when a swing pivot is confirmed, as (centre x 2 + new pivot) / 3 |
| A swing point that won't be revised | Highest high of the last N bars, which moves the moment a higher high prints | A pivot needs N bars strictly lower on both sides, is booked N bars late, and never rewritten |
| An entry that isn't already extended | An arrow on the flip, wherever price happens to be | The flip counts only if price sat on the other side of the 50 EMA six bars earlier |
| A read from more than one chart | Nothing, or a second copy on another timeframe | Four timeframes in one panel, each from the same pivot-break rule |
Row two is worth another minute. "Highest high of the last N bars" sounds like a swing high and behaves nothing like one, because it moves whenever price does. A pivot here is a commitment: once N bars on the right close lower, the high is booked and no later candle takes it back. That's the whole non-repainting story, mechanical rather than a badge.
What it draws
The pivot trailing stop. Teal under price in an uptrend, salmon above it in a downtrend, flat between pivots. It doubles as the obvious stop level, which is why Buy and Sell print on the line rather than above the candles.
Buy and Sell signals. An arrow and a word on the bar where the trend flips and every filter agrees, each in its own indicator buffer so an EA can read them through iCustom.
A Heikin-Ashi trend cloud. Heikin-Ashi candles smoothed by a 21-period EMA, then by a 7. Context, not a signal. MT5 tints the candles to match.
A four-timeframe panel. H4, D1, W1 and MN1 by default, each reading UP or DOWN from the same pivot-break rule, plus a combined view and an ADX strength cell. A timeframe below your chart's is greyed out instead of showing a number that can't mean anything.
Optional layers, all off by default. Premium and discount bands from a rolling median of the price-to-average ratio, a squeeze ribbon, and extremum dots. Off, because everything at once buries the signals.
No stop or target levels, no position sizing, no zones, no scanner. The trailing stop is a line, not a trade plan.
Reading the chart
| On the chart | Read it as |
|---|---|
| Stop flat for many bars | The last swing still holds. Nothing to do |
| Stop steps toward price | A new pivot was confirmed; the trend has a tighter floor |
| Flip with no arrow | A filter said no, usually because price was already extended |
Long: the stop flips up, price is above the 50 EMA now but was at or below it six bars ago, RSI is above 50, and price is above the 120 EMA. Drop to your execution chart, take your own confirmation, and use the stop line as invalidation. Short is the same thing inverted.
Skip it when the panel splits two timeframes against two, or ADX reads Weak and the stop has flipped twice in a handful of bars. The engine is built for trends and will hand you flips in a range just as confidently.
Settings
Six inputs change signals. Everything else changes what you look at.
| Setting | Default | What it changes | When to move it |
|---|---|---|---|
| QA Period [1] | 2 | Bars required each side of a swing before it counts as a pivot | Raise to 4 on M15 and below, where two-bar pivots track noise. Gold H1: 129 at 4 against 153 at 2 |
| QA Multiplier | 3.0 | ATR multiples between pivot centre and stop. The real sensitivity control | Drop toward 1.9 for a tighter stop and far more signals. Gold H1: 251 at 1.9, against 153 at 3.0 |
| QA Factor [1] | 10 | ATR length feeding that distance | Leave it. Gold H1 went 153 to 144 moving it from 10 to 16, noise next to the multiplier |
| RSI Middle | 50 | The momentum line a signal must be on the right side of | Leave it. At 45 nothing changed on either symbol I measured |
| EMA Filter | on | Requires price beyond the 120 EMA in the signal's direction | Off nearly doubles the count, 153 to 289 on gold H1, and the extra mostly fires against the trend |
| QA SQZ Filter | off | Allows signals only when Bollinger Bands sit outside the Keltner channel | On when you want fewer, cleaner setups. Gold H1 drops from 153 to 80 |
Plenty of write-ups treat the multiplier and the ATR length as one control. The counts say otherwise: the multiplier does nearly all the work. Want more signals? Move it and leave the length alone.
Three configurations I've actually watched on a chart:
Default, what the screenshots use Period 2 - Multiplier 3.0 - ATR 10 ~33 signals/year, gold H1 Fewer and later Period 4 - Multiplier 3.0 - ATR 10 ~27 signals/year More, and noisier Period 2 - Multiplier 1.9 - ATR 10 ~53 signals/year
Those are counts, not results. I haven't swept the parameter space and won't pretend three names make a tuned system.
The display block is separate and safe: cloud, bands, ribbon, dots, tinting, panel corner, colours. None of them touch a signal. If you change three things, change the period, the multiplier, and the timeframe you run it on.
What this version can't do
It can't tell you a trend is ending before a pivot confirms it. That's the price of the anchor: N bars have to close on the other side before the line moves, so the flip arrives after the high is in. Want the turn called early? Wrong tool, and no setting fixes it.
The long and short filters aren't symmetrical. Longs are gated by a 120 EMA, shorts by a smoothed 120-period average that runs slower. I kept it rather than rewrite every historical signal, but the two sides aren't mirror images and you should look at both before deciding it suits you.
And it's selective to the point of being boring. Thirty-odd signals a year on H1 means weeks with nothing, and widening the multiplier until the chart looks busy defeats the point of running it.
You won't find a win rate table on this page, and I'd be careful with any indicator page that has one. Signals are entries, not trades, and the number changes completely depending on where the exit goes.
Honestly, the part that ate the most time wasn't the pivot engine: one of the two fill colours on the discount bands stayed a default dark red no matter what I set it to, because a fill reads a second colour slot that doesn't exist unless the plot declares two colours up front. Compiled clean the entire time.
Questions you'll have
Does Moonshine Pivot Trend repaint? Closed bars, no. A pivot needs N bars closed on both sides before it's booked, and it's never revised after that, so the line behind you is fixed. The forming bar is live, so a signal can come and go until that bar closes. That's why the alert defaults to firing after the close.
Which timeframe should I run it on? H1 and H4 are where I've spent the most time. Below M15 the two-bar pivots track noise, so put the period at 4 or more.
Why is my trailing stop line flat for days? Design, not a fault. It moves only when a new swing pivot is confirmed, so a steady trend with no new swing leaves it where it is. A line that moves every bar is what this was built to avoid.
MT4 or MT5? Both, same engine and same defaults bar for bar, each sold here as its own product. This page is the MetaTrader 4 build. MT4 allows 8 indicator buffers in total, so the cloud is two lines there and the candle tinting and extremum dots are MT5 only. MT5 alerts add Telegram with a chart screenshot; MT4 has alert, email and push.
Can my EA read the signals? Yes. Buy and Sell each have a buffer holding the stop price on a signal bar, empty otherwise. No chart objects to parse.
Support
Send me a message with a screenshot showing the panel, plus symbol and timeframe. That's normally enough for me to work out what happened. Updates are free.
Analysis tool. It places no trades and makes no profitability claim. Past price behaviour is not a guide to future price behaviour.
