6 Less Lagging Moving Averages + 2 New MA Indicators

Hey Team,

We all know and love the popular SMA and EMA, but the problem with these moving averages is that they are slow to react to price changes leading to late entries and exits.

Another FF member Alastair and I have teamed up as he reviews 6 BETTER MOVING AVERAGES that aim to reduce the lag to price movement.

He walks through a simulated price chart and shows how the EMA/SMA and the 6 different types of MAs fair against price movement. This was a real eye opener to me of how you can reduce lag and use MAs for more accurate entries/exits.

Of the 6 moving averages, Alistair has created 2 of them as MT4 indicators. Comment on what is your favorite Moving Average!





Thanks,
Zoheb.

zoheb01 Joined Aug 2016

It only took you 5 years to found ALMA indicator that was developed by Arnaud Legous and Dimitrios Kouzis Loukas in 2009. My congratulations!

ALMA always was the best but sadly it's unknown and widely ignored. I posted it here for MT4. Just checked my Git history and found that I added it to the MQL4 framework on 25.02.2011. The version in the linked FF thread is already 3 years old, so if there's interest I will upload a current version (less bugs).

Another very good MA is the original Jurik MA which unfortunately almost disappeared. The versions floating around here are *NOT* the real one. Especially there is a well-known thread here just for Jurik indicators but ironically the thread lacks the most basic one, the Moving Average. There are countless so-called "Smoothers" and "Filters" but the algorithm and the resulting MA are quite different from the original (in between the real author stopped calling that algorithm "JMA"). Again the real JMA is available for MT4 in my framework. On interest I would prepare a build and upload it to this thread.

Here are a paper for the ALMA and an Excel sheet to play with the parameters:

According to explanation in mql5 site ALMA ''uses curve of the Normal (Gauss) distribution which can be placed by Offset parameter from 0 to 1. This parameter allows regulating the smoothness and high sensitivity of the Moving Average.''

So this guy improved the ALMA further by replacing the normal distribution with the Ricker wavelet (Mexican hat wavelet) https://en.wikipedia.org/wiki/Ricker_wavelet .Here is the Right Sided Ricker Moving Average https://www.tradingview.com/script/K...n-Derivatives/

Here is the double jurik moving average that is faster and smoother than the original jurik according to Mladen the author of the indicator.

That's not a JMA. In between Mladen stopped calling this algorithm "Jurik" and just calls it "Smoother" or "Double Smoother". Nevertheless it's a very good indicator.

Emmzett,

As far as I know this is the one: '' simple jurik filter '' https://www.mql5.com/en/forum/180002#comment_4431854 :

That again looks like some derivative of the Smoother. The original JMA is here: https://www.mql5.com/en/articles/1450. Unfortunately the first versions posted there had some bugs from conversion TradeStation -> MQL, and also repainted. At the moment I don't remember a working link to a corrected and non-repainting version other than mine. That's because until now nobody took the effort to convert the original code from old MQL4 to new MQL4/MQL5 - and the code is heavy.

Anyway, here's a screenshot of ALMA and JMA together in one chart. It's a matter of taste what a trader prefers:



edit: Here is a JMA version with the original algorithm for MT5 but on first look it seems to include the repaint bug: https://www.mql5.com/en/code/34224

My last post regarding the jurik:

In my previos post Mladen said about Nikolay Kositsin's JMA '' made after wealth lab decompiled code'' where as some Russians and Igorad revealed the "Jurik filter algorithm ", and it was not an indicator based on a decompiled code from Wealth lab.

Anyway here is the full reverse engineered Jurik algorithm calculations from Igorad and Alexander Smirnov from the Russian magazine 'Spekulant':

Thank you for uploading the PDF. As Igor Durkin wrote it's a Jurik Smoothing Filter, not the real JMA. See the second sentence. Just put both the original and that filter on your screen and compare for yourself.

Let me tell you one story. Few years ago I have created very complicated formula for my indicator. Later I have created light one it just calculate this way: if X candle back lower or higher. For example 14 candles back. And you know what? The results were the same. Exactly the same.

Now about ALMA. Here we have exactly the same thing. I don't know why and how. Don't ask me. But. You can use any moving average you like and just caclulate high/low of some candles back. Default settings for mladen ALMA is 6. Now, let's take a closer look.



Arrows are exactly at the same places where ALMA change color. But ALMA use huge and complicated formula. Arrows on the chart is just simple comparing close value of X candles back.

What I'm talking about. Searching for non-laggin MA is... strange. The KEY feature of MA is to show average price of selected period. If you want non-lag data just use candles.

Now let's think how can that "non-lagging" MA help us in tradind? It can't. There is no positive expectation in long-run. It is about 47% as everything we have. So my advice: don't waste your time searching for non-lag MA. You can use simple MA and the result will be the same.

In trading everything is connected somehow. For example Asias session range is always 1/2 of ATR. If we are taking ADR it is (with some master pair calulations) always the same as ATR. Etc etc etc. Examples are everywhere.

P.s. I'm not a guru and I'm still searching for the edge. So may be I'm wrong, but I think all we see everyday is a big algo driven. Seems to me the gravity and distance (speed) rule the game.

Thank you. Much better expressed than I ever could...because I'm a techie. But I cannot agree more. A Moving Average can guide us, for so much time. Until it stops working. A MACD of two Moving Averages can guide us better but only for so long...until sentiment changes. A channel is the next step and is better to another degree. Until it stops working. So our task is to chose the right tools for the right moment. If markets are crazy like in 2008 a simple moving average (SMA) makes you shitloads of money. If markets range like crazy (look at the Swissie in the last six years) no moving average in the world will save you from slow death.