D Money – EMA/TEMA Touch Strategy (Distance)

What it’s trying to capture

You want mean-reversion “tags” back to a moving average after price has stretched away and momentum flips:

Bearish setup (short): price has been above EMA(9) for a few bars, then MACD turns bearish, and price is far enough above the EMA (by an adaptive threshold). Exit when price tags the EMA.

Bullish setup (long): price has been below your chosen TEMA rail (actually an EMA of 50/100/200 you pick) for a few bars, then MACD turns bullish, and price is far enough below that TEMA. Exit when price tags that TEMA.

The moving averages it uses

EMA(9) — your fast “tag” for short take-profits.

“TEMA line” input = one of EMA(50) / EMA(100) / EMA(200). (Labelled “Chosen TEMA” in the plot; it’s an EMA rail you pick.)

When it will enter trades

It requires four things per side:

Short (EMA-Touch Short)

MACD bearish cross on the signal bar

If “Require NO MA touch on cross bar” = true, the bar’s low must be above EMA(9), so it didn’t touch EMA on the cross bar (fake-out guard).

Extension/Context: you’ve had at least barsAbove consecutive closes above EMA(9) (default 3), so it’s truly stretched.

Distance test: absolute % distance from price to EMA(9) must be ≥ minDistEMA_eff (an adaptive threshold; details below).

Bounce filter: there was no bullish bounce off the EMA in the last bounceLookback bars (excluding the current one).

If all pass and you’re inside the backtest window → strategy.entry short.

Long (TEMA-Touch Long)

MACD bullish cross on the signal bar

With the same fake-out guard: the bar’s high must be below the chosen TEMA if the guard is on.

Extension/Context: at least barsAbove consecutive closes below the chosen TEMA.

Distance test: absolute % distance from price to TEMA must be ≥ minDistTEMA_eff (adaptive).

Bounce filter: there was no bearish bounce off the TEMA in the last bounceLookback bars.

If all pass and you’re in the window → strategy.entry long.

MACD timing option:

If Pure MACD Timing = ON, it only checks for the cross.

If OFF (default), it also enforces “no touch on the cross bar” if that checkbox is true. That’s your “fake-out” filter.

The adaptive distance threshold (the “secret sauce”)

You can choose how “far enough away” is determined—per side:

Fixed %

Short uses Fixed: Min distance ABOVE EMA (%)

Long uses Fixed: Min distance BELOW TEMA (%)

Auto (ATR%) (default)

Short threshold = max(floorEMA, kAtrShort × ATR%)

Long threshold = max(floorTEMA, kAtrLong × ATR%)
This scales distance by recent volatility, with a floor.

Auto (AvgDist%)

Short threshold = max(floorEMA, kAvgShort × average(|Dist to EMA|) over avgLen)

Long threshold = max(floorTEMA, kAvgLong × average(|Dist to TEMA|) over avgLen)
This adapts to the instrument’s typical stretch away from the rails.

These become minDistEMA_eff and minDistTEMA_eff and are re-computed each bar.

Fake-out / bounce logic (the “don’t get tricked” part)

A touch means the bar’s high/low overlapped the MA ± a small buffer % (touchBufPct).

A bounce is a touch plus a close on the “wrong” side (e.g., touch EMA and close above it on shorts = bullish bounce).

The script blocks entries if a bounce happened within bounceLookback bars (excluding the current signal bar).

Exits & risk

Take profit: when price touches the target MA:

Short TP = touch EMA(9)

Long TP = touch chosen TEMA

Stop loss: either

ATR stop: entry ± (atrMultStop × ATR) (default ON), or

Percent stop: entry × (1±stopPct%)

Time stop: if timeExitBars > 0, close after that many bars if still open.

Quality-of-life features

Backtest window (btFrom, btTo) so you can limit evaluation.

Labels on signal bars that show:

MACD bucket (Small/Moderate/HUGE/Violent — based on % separation on the bar),

the current absolute distance to the target MA,

and the effective minimum the engine used (plus which engine mode).

Data Window fields so you can audit:

abs distance to EMA/TEMA,

the effective min distance used on each side,

ATR%,

average absolute distances (for the AvgDist mode).

Alerts fire when a short/long signal is confirmed.

Optional debug panel to see the exact booleans & thresholds the bar had.

Quick mental model

Are we properly stretched away from the rail (by an adaptive threshold) and held on that side for a few bars?

Did MACD flip the way we want without price already tagging the rail that bar?

Have we avoided recent bounces off that rail (no fake-out)?
→ If yes, enter and aim for a tag back to the rail, with ATR/% stop and optional time stop.

If you want, I can add a simple on-chart “rating” (0–100) similar to your Python scorer (distance beyond min, MACD bucket, extension streak) so you can visually rank signals in TradingView too.