TradingView Bayesian Trend Indicator [ChartPrime] Conversion to NinjaTrader 8

Most trend indicators give you a binary answer: up or down, with maybe a third ‘neutral’ state. That’s useful — but it throws away the most actionable piece of information, which is how confident the indicator is in that call. ChartPrime’s Bayesian Trend Indicator reframes the question as a probability: given the last N bars of fast-MA-vs-slow-MA spread, what’s the posterior probability that we’re currently in an uptrend?

The output is a number between 0 and 1 that drives gradient bar coloring (blue at 1.0, red at 0.0, with a smooth fade through the midpoint), per-bar percentage labels, and a corner-anchored SharpDX HUD that reports the current probability, slope, and regime. It reads more like a model output than a traditional indicator — because it is.

This post is the NinjaTrader 8 conversion. The full source compiles standalone in indTradingView/, exposes the posterior probability as a Series<double> for strategy use, and is yours to download as a NinjaScript Archive.

Original Pine Script: Bayesian Trend Indicator by ChartPrime
License: Mozilla Public License 2.0 (MPL 2.0)

🆚 What Makes the Bayesian Trend Different

Traditional MA-cross indicators commit to a regime change the moment fast crosses slow. Bayes doesn’t commit — it updates a belief. Each new bar feeds the prior probability through a likelihood function (computed from the current MA spread), producing a posterior that smoothly tracks the strength of the trend rather than its presence or absence.

Gradient bar coloring. Instead of a solid green or red bar, each candle is tinted by interpolating between the up-trend color and down-trend color in proportion to the current posterior. A 95% bullish reading paints a saturated up-trend candle; a 55% reading paints a much fainter one. The gradient encodes confidence directly into the chart.

Per-bar percentage labels and HUD. Every bar can carry its own integer-percent label (e.g., ‘P=87%’) showing the live posterior, with configurable size and tick offset. A SharpDX HUD anchored to one of the four chart corners shows the current probability, slope, and regime classification — so you can read the model state without staring at a sub-panel.

⚙️ Settings

Bayesian Trend exposes its parameters in four property groups: the calculation inputs that drive the posterior math, the labels group for per-bar probability text, the table group for the SharpDX HUD configuration, and the display colors that interpolate the gradient bar tint.

Indicator Setup

Labels

Table

🧠 How It Works

Each bar updates the running posterior probability that the trend is up, then paints the bar and HUD accordingly.

1. Fast and slow MAs. Compute SMAs of the source on two windows — slow at MA's Length, fast at (MA's Length − Gap Length). The signed spread between them is the raw signal that drives the likelihood.
2. Likelihood computation. Map the MA spread through a logistic function so it becomes a probability in [0, 1]. Large positive spread → likelihood near 1; large negative → near 0; near-zero spread → 0.5.
3. Bayesian update. Combine the prior (last bar’s posterior) with the new likelihood to produce the current bar’s posterior probability of an up-trend. The posterior smooths abrupt likelihood changes — it doesn’t jump from 0.2 to 0.9 in one bar even if the spread reverses.
4. Gradient bar tint. Interpolate the bar’s brush between Down Trend Color and Up Trend Color in proportion to the posterior. A posterior of 0.8 paints the bar 80% of the way toward the up-trend color. BarBrushes[0] and CandleOutlineBrushes[0] get the same blended color.
5. Per-bar label. Format the posterior as an integer percent and render via SharpDX above the high (when bullish) or below the low (when bearish), at a configurable tick offset.
6. HUD render. Compose a small panel with rows for current probability, slope (delta vs prior bar), and regime classification. Render in OnRender at the configured corner of the chart panel.

The indicator is non-repainting: every posterior is computed on the closed bar’s data and never updates after the bar confirms.

🛠️ Using It in a Strategy

Bayesian Trend exposes its posterior probability as a single Series<double> output. The most natural strategy use is to gate entries by a probability threshold: enter long when the posterior crosses above 0.7, short when it crosses below 0.3, and stay flat in the 0.3–0.7 noise zone.

Below is the lifecycle: instantiate the indicator in State.DataLoaded, then in OnBarUpdate read the posterior and act on threshold crosses.

private BayesianTrend bayes;

protected override void OnStateChange()
{
    if (State == State.SetDefaults)
    {
        Name = "BayesianTrendStrategyExample";
    }
    else if (State == State.DataLoaded)
    {
        bayes = BayesianTrend(
            Close,                                     // Source
            50,                                        // MA's Length
            20,                                        // Gap Length
            0.30,                                      // Gap Signals
            true,                                      // Show HUD Table
            true,                                      // Show Probability Labels
            BayesianTrend_TablePosition.TopRight,      // Position
            18,                                        // Title Font Size
            22,                                        // Value Font Size
            14,                                        // Header Font Size
            16,                                        // Body Font Size
            70,                                        // Background Opacity
            12,                                        // Label Font Size
            8                                          // Label Tick Offset
        );
    }
}

protected override void OnBarUpdate()
{
    if (CurrentBar < 51) return;

    double prob     = bayes.PosteriorUp[0];
    double probPrev = bayes.PosteriorUp[1];

    // Threshold cross from neutral into bullish
    if (prob > 0.70 && probPrev <= 0.70)
        EnterLong();

    // Threshold cross from neutral into bearish
    if (prob < 0.30 && probPrev >= 0.30)
        EnterShort();

    // Exit when posterior decays back through the midline
    if (Position.MarketPosition == MarketPosition.Long && prob < 0.50)
        ExitLong();
    if (Position.MarketPosition == MarketPosition.Short && prob > 0.50)
        ExitShort();
}

private BayesianTrend bayes;

protected override void OnStateChange()
{
    if (State == State.SetDefaults)
    {
        Name = "BayesianTrendStrategyExample";
    }
    else if (State == State.DataLoaded)
    {
        bayes = BayesianTrend(
            Close,                                     // Source
            50,                                        // MA's Length
            20,                                        // Gap Length
            0.30,                                      // Gap Signals
            true,                                      // Show HUD Table
            true,                                      // Show Probability Labels
            BayesianTrend_TablePosition.TopRight,      // Position
            18,                                        // Title Font Size
            22,                                        // Value Font Size
            14,                                        // Header Font Size
            16,                                        // Body Font Size
            70,                                        // Background Opacity
            12,                                        // Label Font Size
            8                                          // Label Tick Offset
        );
    }
}

protected override void OnBarUpdate()
{
    if (CurrentBar < 51) return;

    double prob     = bayes.PosteriorUp[0];
    double probPrev = bayes.PosteriorUp[1];

    // Threshold cross from neutral into bullish
    if (prob > 0.70 && probPrev <= 0.70)
        EnterLong();

    // Threshold cross from neutral into bearish
    if (prob < 0.30 && probPrev >= 0.30)
        EnterShort();

    // Exit when posterior decays back through the midline
    if (Position.MarketPosition == MarketPosition.Long && prob < 0.50)
        ExitLong();
    if (Position.MarketPosition == MarketPosition.Short && prob > 0.50)
        ExitShort();
}

PosteriorUp is a Series<double>, so you can index back in time freely — bayes.PosteriorUp[5] reads the posterior five bars ago. The Update() wrapper inside the getter handles sync.

Public Outputs

🔄 Conversion Notes

A few things changed in the translation from Pine Script to NinjaScript:

Gradient bar tint via BarBrushes + CandleOutlineBrushes. Pine’s barcolor() with a gradient fill is reproduced by computing a blended SolidColorBrush per bar and assigning it to both BarBrushes[0] and CandleOutlineBrushes[0]. The interpolation runs in ARGB space across the user’s Up Trend Color and Down Trend Color brushes.

Per-bar SharpDX labels. Pine’s label.new for the per-bar probability is replaced with SharpDX text rendering in OnRender. The label position uses the bar’s high/low plus a configurable tick offset, and the integer-percent format is computed in InvariantCulture to avoid locale-dependent decimals.

SharpDX HUD with configurable typography. The corner-anchored stats panel is fully rendered via SharpDX with four independent font sizes (title, value, header, body), explicit colors for text and muted-headers, and an opacity setting on the background panel that lets users dial it from solid backdrop to fully transparent.

Series-based public output. The posterior is exposed as a Series<double> via PosteriorUp — strategies can index back in time and the Update() wrapper inside the getter ensures sync.

📦 Download

The full source is available as a free NinjaScript Archive. To install:

1. Download the .zip file below.
2. In NinjaTrader 8, go to Tools → Import → NinjaScript Add-On.
3. Select the downloaded .zip file.
4. The indicator will appear under Indicators → indTradingView → Bayesian Trend Indicator [ChartPrime] on your chart.

📊 Chart Example

Bayesian Trend Indicator on a 1-minute chart with default settings. Bars are tinted by the live posterior probability — the deeper blue the bar, the higher the model’s confidence in an up-trend. The corner HUD shows the current numerical probability, slope, and regime classification. Notice how the gradient tracks the strength of moves rather than just their direction — early bars in a new trend show muted color, then saturate as the posterior converges.

🎉 Prop Trading Discounts

💥89% off at Bulenox.com with the code MDT89

The original Pine Script™ code is by ChartPrime and is licensed under the Mozilla Public License 2.0 (MPL 2.0). This NinjaTrader 8 adaptation is by MyDailyTake.com. The use of ChartPrime’s name or adapted code does not imply endorsement by the original author.