Discussing the article: "Gain An Edge Over Any Market (Part II): Forecasting Technical Indicators"
Given that all technical indicators are lagging or redrawing, their predictions will also be lagging or changing.
this statement is not true -- there are many examples -- for example, Zig-Zag -- it does not redraw or lag -- trend reversal shows point to point -- local extrema shows -- current edge is forming, it is not redrawing.
one should proceed from what and how the indicator determines.
this statement is not true -- there are many examples -- for example, Zig-Zag -- it does not redraw or lag -- trend reversal shows point to point -- local extrema shows -- current edge is forming, it is not redrawing.
one should proceed from what and how the indicator determines.
You are misinformed. The ZigZag is the worse of the lot—it both lags and it redraws.
@Stanislav Korotky is correct—ALL technical indicators either lag or redraw or both. This is a mathematical fact, because it is based on past data. There is no future data.
Given that all technical indicators are lagging or redrawing indicators, their predictions will also lag or change.
I also disagree with such a categorical statement. There is a simple divergence, for example. It doesn't lag. And it can trumpet several times about the imminent reversal or price correction.... Of course, the prediction will have some degree of probability, but this is another conversation....
For example, there is an indicator for clustering volatility of returns (ReturnsIndicator). It has the simplest logic: if there was a small volatility today, it will be the same tomorrow. What is not a prediction?
- www.mql5.com
The evidence is the mathematics and code implementation of the ZigZag. Look at it yourself. This is common knowledge. I'm surprised you are not aware.
I even published a version a decade ago in the CodeBase to explain this ...
ZigZag Indicator with Extra Features
Fernando Carreiro, 2014.01.15 11:26
Taking a Closer Look at the Workings of the ZigZag Indicator.
- www.mql5.com
The proof is in the maths and the implementation of the ZigZag code. Look at it for yourself. It's common knowledge. I'm surprised you're not aware of it.
I even published a version from a decade ago in CodeBase to explain it...
At the link is your key description of Zig-Zag in that quote:
However, nothing could really be further from the truth, mainly because it does what is called"redrawing " . In other words, during a live sequence of price action changes, the indicator changes the most recent peak or trough to reflect the new price data. By the time a ZigZag top or trough settles and becomes fixed on the indicator, the current market situation has long since changed and no longer corresponds to the point that was originally indicated as the price top or trough.
-- here you are only talking about the current ZigZag edge.
Talking about an indicator -- you have to proceed from:
-- that the indicator defines (the purpose of the indicator)
-- how the indicator defines it (algorithm)
As for Zig-Zag -- what Zig-Zag defines:
1) local extremum formed -- this extremum Zig-Zag shows on history -- it cannot be lagging or redrawing by definition.
In your quote you write,"the current market situation has long since changed and no longer corresponds to the point that was originally indicated as a price peak or trough".
So Zig-Zag is not arguing against the current situation -- it is showing a previously formed extreme. The current situation can be analysed in relation to that extremum. For example, extrema can be understood as support/resistance levels (which, by the way, work).
2) the moment of formation of a local extremum occurs on the current flow of quotes. Zig-Zag this moment shows point to point. It is possible to highlight the tick on which the extremum was formed. There is no and cannot be overdrawing or lagging here.
3) the moment of trend change -- coincides with the moment of local extremum formation -- the exact tick at which the trend changed is known.
There may be misunderstandings about the current Zig-Zag edge. This edge shows extrema that have not formed. Again, a tick on which this extremum failed can be shown. The cancellation of such extrema is interpreted as a continuation of the trend, the edge moves. Such moving of the edge is not a redrawing -- it is a search for a local extremum.
So there is no lagging or redrawing in Zig-Zag. How to use Zig-Zag in trading is a question for the trading strategy. It is important that the trading strategy understands Zig-Zag correctly.In the first case they average prices, and averaging itself implies a distance from the price.
This very "lagging" is compensated by a forecast with 50% working out of the price return to the average. By this logic, a "lagging" indicator turns into a "leading" one. Which is a contradiction in itself.
There is no lag, there is working with the existing prices, i.e. the indicator is "present", not past or future.
As for redrawing: it is childish naivety to consider leg lengthening as redrawing .
Appearance of a leg IMMEDIATELY cannot become fixed at once, because ZigZag has no Vanga code for this purpose. It has an ordinary executable algorithm of following the price in the true sense: first price - then ZigZag.
The notion of "lag" and "redrawing" in the context of evaluative judgements is a kindergarten of the level of "I will add a secondmoving average and my Ilan 1.6 will never fall into a counter-trend"
All the more so from venerable programmers.
I also disagree with such a categorical statement. For example, there is a simple divergence. It doesn't lag. And it can trumpet several times about an imminent reversal or price correction..... Of course, the prediction will have some degree of probability, but this is another conversation....
For example, there is an indicator for clustering volatility of returns (ReturnsIndicator). It has the simplest logic: if there was a small volatility today, it will be the same tomorrow. What is not a prediction?
Divergence requires finding some already formed "figures" - extrema, fractals, etc. in order to make a decision/prediction. - to make a decision/prediction. The time for formation is given by the lag, because of which the signal is delayed.
Predicting the future value of the value equal to the current one is the most trivial technique, which does not count as a prediction and does not give a gain if it is followed systematically.
All the more so from venerable programmers .
Some rubbish. Averaging is digital or analogue signal processing that builds up by delaying it - purely by formulae. To deny it is to show illiteracy.
And there is no need to juggle with terms - of course every indicator has a current value, but it does not become more accurately predictive from this name. And to dispute the fact that the last segment of a zigzag is redrawn is ridiculous. You can't tell where the current leg of a zigzag will end until the next leg appears. This means that the prediction that can be made based on the statistics for the current end of the zigzag configuration will change dynamically.
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
You agree to website policy and terms of use
Check out the new article: Gain An Edge Over Any Market (Part II): Forecasting Technical Indicators.
Did you know that we can gain more accuracy forecasting certain technical indicators than predicting the underlying price of a traded symbol? Join us to explore how to leverage this insight for better trading strategies.
Investors seeking to apply machine learning in electronic trading environments face numerous challenges, and the reality is that many do not achieve their desired outcomes. This article aims to highlight some reasons why, in my opinion, an aspiring algorithmic trader may fail to achieve satisfactory returns relative to the complexity of their strategies. I will demonstrate why forecasting the price of a financial security often struggles to exceed 50% accuracy and how focusing on predicting technical indicator values, instead, can improve accuracy to around 70%. This guide will provide step-by-step instructions on best practices for time series analysis.
By the end of this article, you will have a solid understanding of how to enhance the accuracy of your machine learning models and discover leading indicators of market changes more effectively than other participants using Python and MQL5.
Author: Gamuchirai Zororo Ndawana