Grand Trend Forecasting - A Simple And Original Approach Today we'll link time series forecasting with signal processing in order to provide an original and funny trend forecasting method, the post share lot of information, if you just want to see how to use the indicator then go to the section "Using The Indicator".
Time series forecasting is an area dealing with the prediction of future values of a series by using a specific model, the model is the main tool that is used for forecasting, and is often an expression based on a set of predictor terms and parameters, for example the linear regression (model) is a 1st order polynomial (expression) using 2 parameters and a predictor variable ax + b . Today we won't be using the linear regression nor the LSMA.
In time series analysis we can describe the time series with a model, in the case of the closing price a simple model could be as follows :
Price = Trend + Cycles + Noise
The variables of the model are the components, such model is additive since we add the component with each others, we should be familiar with each components of the model, the trend represent a simple long term variation of high amplitude, the cycles are periodic fluctuations centered around 0 of varying period and amplitude, the noise component represent shorter term irregular variations with mean 0.
As a trader we are mostly interested by the cycles and the trend, altho the cycles are relatively more technical to trade and can constitute parasitic fluctuations (think about retracements in a trend affecting your trend indicator, causing potential false signals).
If you are curious, in signal processing combining components has a specific name, "synthesis" , here we are dealing with additive synthesis, other type of synthesis are more specific to audio processing and are relatively more complex, but could be used in technical analysis.
So what to do with our components ? If we want to trade the trend, we should estimate right ? Estimating the trend component involve removing the cycle and noise component from the price, if you have read stuff about filters you should know where i'am going, yep, we should use filters, in the case of keeping the trend we can use a simple moving average of relatively high period, and here we go.
However the lag problem, which is recurrent, come back again, we end up with information easier to interpret (here the trend, which is a simple fluctuation such as a line or other smooth curve) at the cost of decision timing, that is unfortunate but as i said the information, here the moving average output, is relatively simple, and could be easily forecasted right ? If you plot a moving average of high period it would be easier for you to forecast its future values. And thats what we aim to do today, provide an estimate of the trend that should be easy to forecast, and should fit to the price relatively well in order to produce forecast that could determine the position of future closing prices observations.
Estimating And Forecasting The Trend
The parameter of the indicator dealing with the estimation of the trend is length , with higher values of length attenuating the cycle and noise component in the price, note however that high values of length can return a really long term trend unlike a simple moving average, so a small value of length, 14 for example can still produce relatively correct estimate of trend.
here length = 14.
The rough estimate of the trend is t in the code, and is an IIR filter, that is, it is based on recursion. Now i'll pass on the filter design explanation but in short, weights are constants, with higher weights allocated to the previous length values of the filter, you can see on the code that the first part of t is similar to an exponential moving average with :
t(n) = 0.9t(n-length) + 0.1*Price
However while the EMA only use the precedent value for the recursion, here we use the precedent length value, this would just output a noisy and really slow output, therefore in order to create a better fit we add : 0.9*(t(n-length) - t(n-2length)) , and this create the rough trend estimate that you can see in blue. On the parameters, 0.9 is used since it gives the best estimate in my opinion, higher values would create more periodic output and lower values would just create a rougher output.
The blue line still contain a residual of the cycle/noise component, this is why it is smoothed with a simple moving average of period length. If you are curious, a filter estimating the trend but still containing noisy fluctuations is called "Notch" filter, such filter would depending on the cutoff remove/attenuate mid term cyclic fluctuations while preserving the trend and the noise, its the opposite of a bandpass filter.
In order to forecast values, we simply sum our trend estimate with the trend estimate change with period equal to the forecasting horizon period, this is a really really simple forecasting method, but it can produce decent results, it can also allows the forecast to start from the last point of the trend estimate.
Using The Indicator
We explained the length parameter in the precedent section, src is the input series which the trend is estimated, forecast determine the forecasting horizon, recommend values for forecast should be equal to length, length/2 or length*2, altho i strongly recommend length.
here length and forecast are both equal to 14 .
The corrective parameter affect the trend estimate, it reduce the overshoot and can led to a curve that might fit better to the price.
The indicator with the non corrective version above, and the corrective one below.
The source parameter determine the source of the forecast, when "Noisy" is selected the source is the blue line, and produce a noisy forecast, when "Smooth" is selected the source is the moving average of t , this create a smoother forecast.
The width interval control...the width of the intervals, they can be seen above and under the forecast plot, they are constructed by adding/subtracting the forecast with the forecast moving average absolute error with respect to the price. Prediction intervals are often associated with a probability (determining the probability of future values being between the interval) here we can't determine such probability with accuracy, this require (i think) an analysis of the forecasting distribution as well as assumptions on the distribution of the forecasting error.
Finally it is possible to see historical forecasts, that is, forecasts previously generated by checking the "Show Historical Forecasts" option.
Examples
Good forecasts mostly occur when the price is close to the trend estimate, this include the following highlighted periods on AMD 15TF with default settings :
We can see the same thing at the end of EURUSD :
However we can't always obtain suitable fits, here it is isn't sufficient on BTCUSD :
We can see wide intervals, we could change length or use the corrective option to get better results, another option is to use a log scale.
We will end the examples with the log SPX, who posses a linear trend, so for example a linear model such as a linear regression would be really adapted, lets see how the indicator perform :
Not a great fit, we could try to use an higher length value and use "Smooth" :
Most recent fits are quite decent.
Conclusions
A forecasting indicator has been presented in this post. The indicator use an original approach toward estimating the trend component in the closing price. Of course i should have given statistics related to the forecasting error, however such analysis is worth doing with better methods and in more advanced environment allowing for optimization.
But we have learned some stuff related to signal processing as well as time series analysis, seeing a time series as the sum of various components is really helpful when it comes to make sense of chaotic and noisy series and is a basic topic in time series analysis.
You can see that in this new year i work harder on the visual of my indicators without trying to fall in the label addict trap, something that i wasn't really doing before, let me know what do you think of it.
Thanks for reading !
Indicator
Voss Predictive Filter█ OVERVIEW
The Voss Predictive Filter (VPF) is a negative group delay (NGD) filter that anticipates cyclical price movement through phase compensation. The VPF isolates band-limited cyclical components via a bandpass filter, then applies negative group delay to shift the signal's phase forward, causing the output to lead the input by a fraction of the cycle period.
Based on Dr. John F. Ehlers' "Voss Predictive Filter" article in Technical Analysis of Stocks & Commodities (TASC) magazine, the VPF displays a predictive oscillator with optional dynamic threshold bands for identifying significant cycle behavior. The indicator is timeframe-agnostic - the mathematics work identically from tick charts to monthly bars, though shorter timeframes require more careful parameter selection due to noise.
█ CONCEPTS
Bandpass Filtering
A bandpass filter isolates price activity within a specific frequency range, removing both high-frequency noise and low-frequency trend drift. The VPF uses a second-order IIR (Infinite Impulse Response) bandpass filter characterized by the center frequency (the Bandpass Period input) and bandwidth. The center frequency determines which cycle period the filter emphasizes, while bandwidth controls the damping coefficient - how tightly the filter focuses around that frequency. Before filtering, the source is debiased via 2-bar momentum to remove DC offset, ensuring the filter operates around a true zero centerline.
Negative Group Delay Filtering
The predictive capability stems from negative group delay (NGD) - a filter characteristic where output appears to "lead" the input. Most causal filters introduce lag (positive group delay), but by combining the bandpass filter output with appropriately weighted past values, the VPF achieves negative group delay characteristics.
This is a universal NGD filter application for band-limited signals: the bandpass filter isolates the cyclical component of interest, then the NGD stage advances the phase within this limited frequency range to create an anticipatory output. This isn't statistical forecasting; it's phase compensation that shifts the signal's timing forward, causing peaks and troughs to appear before they occur in the bandpass output.
Negative Group Delay Stage
The NGD stage combines the current bandpass output with weighted historical values to produce an output that leads the input. By subtracting a weighted average of past deviations from a scaled version of the current filter value, the algorithm advances the signal's phase: peaks and zero-crossings in the voss output appear before the corresponding events in the bandpass filter.
The prediction order (`3 * Prediction Multiplier`) controls how many past values contribute to the phase advance. Higher orders provide smoother output but reduce the leading effect; lower orders maximize anticipation at the cost of stability.
█ INTERPRETATION
Zero-Line Crossovers
Crossings above zero suggest bullish momentum in the filtered cycle; below zero suggests bearish momentum. Crossings from near-zero regions are most reliable, as extreme excursions need time to return to equilibrium.
Threshold Bands
Threshold bands define "significant" deviation. Breaches indicate unusually strong behavior and can serve as:
• Trend confirmation when aligned with price direction
• Overbought/oversold warnings at extremes
• Trade entry filters (requiring threshold breach in the intended direction)
Threshold Mode affects sensitivity: MAD (outlier-resistant), Standard Deviation (volatility-sensitive), Percentile Rank (fixed probability bands).
Alert Conditions
Four built-in alerts trigger on bar close (no repainting): Above +Threshold (strong bullish cycle), Below -Threshold (strong bearish cycle), Above Zero (bullish phase shift), Below Zero (bearish phase shift).
█ SETTINGS & PARAMETER TUNING
Voss Predictive Filter
• Source : Price series to filter.
• Bandpass Period (1-100): Primary tuning parameter determining which cycle length the filter emphasizes. Short periods (8-15) are more responsive but noisier; medium periods (16-30) balance responsiveness and smoothness; long periods (31-100) focus on longer cycles with more smoothing.
• Bandwidth (0.01-0.45): Controls filter selectivity. Narrow bandwidths (0.01-0.15) isolate specific cycle periods precisely; medium (0.16-0.30) tolerate cycle irregularity; wide (0.31-0.45) capture broader cycle ranges. Shorter periods pair well with narrower bandwidths.
• Prediction Multiplier (2-10): Controls how many past values contribute to the phase advance. Higher values provide smoother output but reduce the leading effect; lower values maximize anticipation at the cost of stability.
Display Settings
Control visibility and colors of the Voss output, bandpass filter, and zero reference lines.
Diagnostics - Dynamic Thresholds
Three methods identify significant signal deviation:
• MAD (Median Absolute Deviation) : Robust, outlier-resistant measure using `k * MAD` where `MAD ≈ 0.6745 * stdev`.
• Standard Deviation : Volatility-sensitive, calculated as `k * stdev` of Voss over the lookback period.
• Percentile Rank : Fixed probability bands using the percentile of |Voss| (e.g., 90% means only 10% of values exceed threshold).
Settings:
• Dynamic Threshold : Toggle threshold bands and set colors.
• Threshold Mode : Select MAD, Standard Deviation, or Percentile Rank.
• Period (2-200): Lookback for threshold calculations. Default 50.
• Multiplier (k) : Scaling for MAD/Standard Deviation modes. Default 1.5.
• Percentile (%) (0-100): For Percentile Rank mode only. Default 90%.
█ LIMITATIONS
Inherent Characteristics
• Residual lag : Despite negative group delay design, some lag remains relative to price action.
• Cyclical markets required : Performs best on instruments with clear cyclical components. Strongly trending markets with little cyclicality produce less useful signals.
• Signal interpretation : Absolute Voss values are instrument-specific. Always interpret relative to adaptive threshold bands, not fixed levels.
Market Conditions to Avoid
• Sudden news events/gaps : Major discontinuities disrupt cycle continuity, causing erratic signals. Requires 1-2 full cycle periods to re-stabilize.
• Low volume/illiquid markets : Sporadic trading produces false cycles from liquidity artifacts. Use only on actively traded instruments during liquid hours.
• Regime changes : During cyclical ↔ trending transitions, watch for persistent extremes without mean reversion, increasing price/indicator divergence, or unresolved threshold breaches.
Parameter Selection Pitfalls
• Mismatched period : If Bandpass Period doesn't match actual market cycles, the filter produces weak signals. Use cycle measurement tools (FFT, autocorrelation, Dominant Cycle) to identify appropriate periods first.
• Overoptimization : Perfect historical fits typically fail forward. Choose robust parameters that work across multiple instruments and timeframes.
█ NOTES
Credits
This indicator is based on concepts from Dr. John F. Ehlers' work on predictive filters and bandpass techniques for technical analysis. Dr. Ehlers has published extensively on applying digital signal processing methods to financial markets in Technical Analysis of Stocks & Commodities (TASC) magazine. His articles on bandpass filters and predictive techniques, particularly the Voss Predictive Filter concept, provided the theoretical foundation for this implementation.
For those interested in the underlying mathematics and DSP concepts:
• Ehlers, J.F. (2001). Rocket Science for Traders: Digital Signal Processing Applications . John Wiley & Sons.
• Various TASC articles by John Ehlers on bandpass filters, cycle analysis, and predictive filtering techniques.
• Ehlers, J.F. "Voss Predictive Filter" - Technical Analysis of Stocks & Commodities magazine.
by ♚@e2e4 Indicator
Indicator
Linear Quadratic Convergence Divergence OscillatorIntroduction
I inspired myself from the MACD to present a different oscillator aiming to show more reactive/predictive information. The MACD originally show the relationship between two moving averages by subtracting one of fast period and another one of slow period. In my indicator i will use a similar concept, i will subtract a quadratic least squares moving average with a linear least squares moving average of same period, since the quadratic least squares moving average is faster than the linear one and both methods have low-lag this will result in a reactive oscillator.
LQCD In Details
A quadratic least squares moving average try to fit a quadratic function (parabola) to the price by using the method of least squares, the linear least squares moving average try to fit a line. Non-linear fit tend to minimize the sum of squares in non-linear data, this is why a quadratic method is more reactive. The difference of both filters give us an oscillator, then we apply a simple moving average to this oscillator to provide the signal line, subtracting the oscillator and its signal line give us the histogram, those two last steps are the same used in the MACD.
Length control the period of the quadratic/linear moving average. While the MACD use a signal line for plotting the histogram i also added the option to plot the momentum of the quadratic moving average instead, the result is smoother and reduce irregularities, in order to do so just check the differential option in the parameter box.
The period of the signal line and the momentum are both controlled by the signal parameter.
A predictive approach can be made by subtracting the histogram with the signal line, this process make the histogram way more predictive, in order to do so just check the predictive histogram option in the parameter box.
Predictive histogram with simple histogram option. The differential mode can also be used with the predictive parameter, this result in a smoother but less reactive prediction.
Information Interpretation
The amount of information the MACD can give us is high. We can use the histogram as signal generator, or the if the oscillator is over/under 0, combine the oscillator/signal line with histogram, combinations can provide various systems. Some traders use the histogram as signal generator and use the cross between the histogram and the signal line as a stop signal, this method can avoid some whipsaw trades. The study of divergences with the price is also another method.
Conclusion
This oscillator aim to show the same amount of information as the MACD with a similar calculation method but using different kind of filters as well as eliminating the need to use two separates periods for the moving averages calculation, its still possible to use different periods for the quadratic/linear moving average but the results can be less accurate. This indicator can be used like the MACD.
Indicator
Indicator
Dominant Cycle Tuned RsiIntroduction
Adaptive technical indicators are importants in a non stationary market, the ability to adapt to a situation can boost the efficiency of your strategy. A lot of methods have been proposed to make technical indicators "smarters" , from the use of variable smoothing constant for exponential smoothing to artificial intelligence.
The dominant cycle tuned rsi depend on the dominant cycle period of the market, such method allow the rsi to return accurate peaks and valleys levels. This indicator is an estimation of the cycle finder tuned rsi proposed by Lars von Thienen published in Decoding the Hidden Market Rhythm/Fine-tuning technical indicators using the dominant market vibration/2010 using the cycle measurement method described by John F.Ehlers in Cybernetic Analysis for Stocks and Futures .
The following section is for information purpose only, it can be technical so you can skip directly to the The Indicator section.
Frequency Estimation and Maximum Entropy Spectral Analysis
“Looks like rain,” said Tom precipitously.
Tom would have been a great weather forecaster, but market patterns are more complex than weather ones. The ability to measure dominant cycles in a complex signal is hard, also a method able to estimate it really fast add even more challenge to the task. First lets talk about the term dominant cycle , signals can be decomposed in a sum of various sine waves of different frequencies and amplitudes, the dominant cycle is considered to be the frequency of the sine wave with the highest amplitude. In general the highest frequencies are those who form the trend (often called fundamentals) , so detrending is used to eliminate those frequencies in order to keep only mid/mid - highs ones.
A lot of methods have been introduced but not that many target market price, Lars von Thienen proposed a method relying on the following processing chain :
Lars von Thienen Method = Input -> Filtering and Detrending -> Discrete Fourier Transform of the result -> Selection using Bartels statistical test -> Output
Thienen said that his method is better than the one proposed by Elhers. The method from Elhers called MESA was originally developed to interpret seismographic information. This method in short involve the estimation of the phase using low amount of information which divided by 360 return the frequency. At first sight there are no relations with the Maximum entropy spectral estimation proposed by Burg J.P. (1967). Maximum Entropy Spectral Analysis. Proceedings of 37th Meeting, Society of Exploration Geophysics, Oklahoma City.
You may also notice that these methods are plotted in the time domain where more classic method such as : power spectrum, spectrogram or FFT are not. The method from Elhers is the one used to tune our rsi.
The Indicator
Our indicator use the dominant cycle frequency to calculate the period of the rsi thus producing an adaptive rsi . When our adaptive rsi cross under 70, price might start a downtrend, else when our adaptive rsi crossover 30, price might start an uptrend. The alpha parameter is a parameter set to be always lower than 1 and greater than 0. Lower values of alpha minimize the number of detected peaks/valleys while higher ones increase the number of those. 0.07 for alpha seems like a great parameter but it can sometimes need to be changed.
The adaptive indicator can also detect small top/bottoms of small periods
Of course the indicator is subject to failures
At the end it is totally dependent of the dominant cycle estimation, which is still a rough method subject to uncertainty.
Conclusion
Tuning your indicator is a great way to make it adapt to the market, but its also a complex way to do so and i'm not that convinced about the complexity/result ratio. The version using chart background will be published separately.
Feel free to tune your indicators with the estimator from elhers and see if it provide a great enhancement :)
Thanks for reading !
References
for the calculation of the dominant cycle estimator originally from www.davenewberg.com
Decoding the Hidden Market Rhythm (2010) Lars von Thienen
Ehlers , J. F. 2004 . Cybernetic Analysis for Stocks and Futures: Cutting-Edge DSP Technology to Improve Your Trading . Wiley
Indicator
Alpha-Sutte ModelThe Alpha-Sutte model is an ongoing project run by Ansari Saleh Ahmar, a lecturer and researcher at Universitas Negeri Makassar in Indonesia, that attempts to make forecasts for time series like how Arima and Holt-Winters models do. Currently Ahmar and his team have conducted research and published papers comparing the efficacy of the Alpha-Sutte and other models, such as Arima and Holt-Winters, on topics ranging from forecasting Turkey's CPI data, Bitcoin prices, Apple's stock prices, primary energy supply of Indonesia, to infant mortality rates in China.
The Alpha-Sutte model in comparison to the other two models listed above shows promise in providing a more accurate forecast, and the project has been able to receive some of its funding from organizations such as the US Agency for International Development, which is a part of the US Federal Government, so maybe the project has some actual merit.
How it works:
In this model there are four values presented at the top of the window.
1) The first value in blue is the value of the Alpha-Sutte model whose purpose is to forecast the price of the current bar.
2) The second value in yellow is an adaptive version of the Alpha-Sutte model that I made. The purpose of the adaptive Alpha-Sutte model is to expand upon the Alpha-Sutte by allowing new information to be introduced, causing the value to change during the current period, hence the adaptiveness of it.
3) The third value in aqua is the moving average of the low% Sutte line which is a predictive line that is based off of the close and low of the current and previous periods.
4) The fourth value in red is the moving average of the high% Sutte line which is a predictive line that is based off of the close and high of the current and previous periods.
Trend signals:
If low% Sutte (aqua value/line) is greater than high% Sutte (red value/line) then this is a buy signal.
If high% Sutte (red value/line) is greater than low% Sutte (aqua value/line) then this is a sell signal.
Caveat:
Even though this model's purpose is to forecast the future, will it be able to predict periods of large movements? No, of course not, but it will adjust quickly to try to make more accurate forecasts for the next period. This was also a reason why I made an adaptive version of this model to try to reduce some of the discrepancies between the Alpha Sutte and price when there is a large unexpected move.
*WARNING before using this I would highly recommend that you look up "Sutte Indicator" online and read some of the papers about this model before you use this , even though this model has shown merit when compared to Arima and Holt-Winter models this is still an ongoing project.*
Hopefully this project will actually come to something in the near future as the calculation for this time series predictive model is much easier to calculate and program in pine editor than something like an Arima model.
*Also, if you know how to use R language there is a package for the "Alpha-Sutte model".* Indicator
Indicator
Godmode 3.1.4 - SNOW_CITY SCALPER EDITION - 7/2017 Updates to xSilas Godmode Oscillator published December 19th, 2014
Thanks to LEGION, LAZYBEAR, Ni6HTH4wK, xSilas
Updates:
Changed default sources to include BITSTAMP and REMOVED BTC-e
Changed default lengths for SCALPING SUPREMACY - See Instructions
Changed "Caution dots" to RED because YELLOW was hard to see.
I mostly Ignore the oscillators and only use the caution dots with this configuration:
MOST EFFECTIVE USE: BITSTAMP:BTCUSD BITSTAMP:BTCUSD
- Use on 1m charts of your BITMEX swap, yes 1m. I know, but it works better this way. 5m and 15m work best when using on an exchange index.
- IF overall 2H trend is DOWN, the 1m godmode on swap CAUTION DOTS appear on the UPPER BOUNDS means good SHORT entry points, and the lower dots suggest a possible reversal and good exit opportunity (not always)
- If overall 2H trend is UP, the 1m godmode on swap CAUTION DOTS appear on the LOWER BOUNDS means good LONG entry points, and upper dots suggest possible oversold if they start to round over, again not always.
This is ultra simple, and very effective.
Default settings for VERY sensitive CAUTION DOT blop: 17,6,4
Use these settings for a slightly less sensitive CAUTION DOT blop: 14,12,9 Indicator
Indicator
