The Coral Filter is a linear combination of moving averages, all obtained by a triple or higher order of exponential smoothing. To learn more about the Coral Filter, a.k.a. Coral Trend indicator, watch the video or continue reading below:
As the name implies, it is a tool to determine trend bias, filtering out setups that go against the general market direction. Specifically, when prices move above the Coral Filter plot, short positions are blocked. On the other hand, when prices move below the Coral Filter, one should not open long positions.
A Normalized Slope
One may further refine the basic, price action above / below vs. the Coral Filter plot by considering a normalized slope. Determining market bias is then based on the Coral Trend slope. An upwards moving slope signifies a bullish trend whereas a down slope indicates a bearish scenario. Both long and short setups may be considered in cases when you see a flat slope, i.e. a neutral / sideways market condition.
The trend indication is exposed via paint bars, as based on the normalized slope of the plot:
- green for up-trends
- red for down-trends
- yellow for neutral trends
One may define the range of coral trend slopes for neutral trends by adjusting the threshold parameter. If set to 0, neutral states are eliminated and the indicator will only plot up and down trends.
However, a disadvantage of using a slope for determining trends (via coloring) is that it delivers the relevant information with a significant delay (lag). More timely trend readings are found by applying a moving average cross. A bullish trend may then be established when a fast moving average crosses above the slow moving average and vice versa for a bearish scenario.
Although this approach will reduce lag in a significant way, many noise signals will appear during ranging markets. Price action will then move back and forth, crossing the moving average several times from above and below.
The chart below applies a standard EMA, as available through the Library amaMultiMACross indicator. The indicator offers a selection of more than 30 different algorithms plotting fast and slow moving averages, disclosing the trend information via paint bars. The trend information is furthermore exposed as a public property data series, available for use with automated strategies and with other indicators.
A Volatility Channel with Neutral Territory
An approach for dealing with lag and at the same time avoiding whipsaws, is to add a volatility channel around the moving average. Below we’ve applied the amaMultiMAWave indicator, again using the EMA but instead conditioning trend readings according to price action vs. the channel:
- bar closes above the upper channel: uptrend
- bar closes inside the channel: neutral trend
- bar closes below the lower channel: downtrend
By adding a channel with “neutral territory”, most whipsaws are then eliminated without the lag seen with the Coral Trend slope approach. Effectively, this will allow you to hold positions as long as they’re not contradicted by an opposing trend reading.
As with the MA Cross, the Moving Average Wave indicator also features a selection of 30 different moving averages. You may also customize the Wave bands using the indicators multiplier function. The default value of 1 creates the EMA Wave using the High and Lows of the specified lookback period.
Other Indicators from the Filter and Multiple Series categories
The Coral Filter, a.k.a. the Coral trend indicator, is found in the Filters category of our NinjaTrader indicator Library. In this category, you’ll also find the Adaptive Laguerre Filter, ADXVMA, Butterworth Filter, Heikin Ashi and the Supersmoother Filter, just to name a few.
The amaMultiMACross and the amaMultiMAWave discussed above are available from the Multiple Series indicators category. Here you’ll also find the Fibonacci Bands, Multiple Bollinger Bands, Multiple Deviation Bands, Multiple Keltner Channels, Multiple MACD, Multiple Rainbow Filter, Multiple True Strength, Rainbow Oscillator, Residual Deviation Bands and the SupertTrend U11 indicator.
