Simple Mean Reversion | Page 31 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

A Simple Mean Reversion Strategy 392 replies

Pairs Trading: Reversion to the Mean 22 replies

Forward test of my new mean reversion strategy 11 replies

Synthetic hedges, cointegration, mean reversion and similar stuff 419 replies

Cointegration, Synthetic hedges, mean reversion in R, Tech Thread 0 replies

Attachments: Simple Mean Reversion
Exit Attachments

Simple Mean Reversion

  • #601
  • May 14, 2023 6:42am May 14, 2023 6:42am
Quoting osfx
Hi marengo, I coudn't dercypt alphaomega's messages in every detail either, but as an algo-trader, they gave me some valuable hints for my search for statistical edges. I would like to comment two of your statements: {quote} At least for Forex, my statistical evaluations and tests tell me, that mean reversion strategies mostly have a better edge than trend following ones. It seems, that the old economic rule "low prices attract buyers and high prices attract sellers" is still valid in the long run. {quote} I coudn't find any edge until...
There's no edge selling/buying in the first x hours of the day. But you can find the edge when you can identify the value (undervalued/overvalued) of certain instrument, but that under/over valuation returns to the fair value/mean within short-term. Sometimes the value offer very good opportunity (like 100 pips), and sometimes few pips.

As a proof of my word, look at USDCAD. The fair value is 1.34, but according to volatility and time fair value can change a little up/down. At least here 100 pips can be grabbed.
Cheers

And this leads to that if someone can't identify the worth/value, how someone know where the mean is?!
  • #602
  • May 15, 2023 1:24am May 15, 2023 1:24am
Quoting Ihlas
{quote} There's no edge selling/buying in the first x hours of the day. But you can find the edge when you can identify the value (undervalued/overvalued) of certain instrument, but that under/over valuation returns to the fair value/mean within short-term. Sometimes the value offer very good opportunity (like 100 pips), and sometimes few pips. As a proof of my word, look at USDCAD. The fair value is 1.34, but according to volatility and time fair value can change a little up/down. At least here 100 pips can be grabbed. Cheers And this leads to...

The simplest solution is to calculate mean of N last observations and use it in mean reversion strategy but this is MA-divergence strategy at its core, maybe playing with weights or window size will help but overfitting is likely at some stage.
  • #603
  • May 15, 2023 5:17am May 15, 2023 5:17am
Quoting pizd0s
{quote} The simplest solution is to calculate mean of N last observations and use it in mean reversion strategy but this is MA-divergence strategy at its core, maybe playing with weights or window size will help but overfitting is likely at some stage.
I have already solution for this

Nothing to do with MA
  • #604
  • May 23, 2023 2:25am May 23, 2023 2:25am
Quoting Ihlas
{quote} I have already solution for this Nothing to do with MA
Care to elaborate? I am confused…
  • #605
  • May 23, 2023 2:31am May 23, 2023 2:31am
Quoting marengo10021
{quote} Care to elaborate? I am confused…
I mean it has nothing to do with averaging
  • #606
  • May 23, 2023 6:02am May 23, 2023 6:02am
Quoting Ihlas
{quote} I mean it has nothing to do with averaging
Also do you pay attention other pairs are trending while USDCAD is ranging? Because it should hit the mean

And pay attention if UCAD breaks today's low first, be ready to see 1.34

If it breaks first today's high, be ready to see 1.36332

UCAD
  • #607
  • May 25, 2023 6:45am May 25, 2023 6:45am
Quoting Ihlas
{quote} Also do you pay attention other pairs are trending while USDCAD is ranging? Because it should hit the mean And pay attention if UCAD breaks today's low first, be ready to see 1.34 If it breaks first today's high, be ready to see 1.36332 UCAD
BAM!
  • #608
  • May 25, 2023 2:33pm May 25, 2023 2:33pm
Apparently there is some sort of counting candle system using timeframes to match, it would be interesting if someone knew about this.
  • #609
  • May 26, 2023 4:04am May 26, 2023 4:04am
Forget candle
  • #610
  • May 27, 2023 11:54pm May 27, 2023 11:54pm
One more example: AUDUSD should break the low first, target 0.64479, max DD zone 0.65858
  • #611
  • May 28, 2023 10:04pm May 28, 2023 10:04pm
Quoting Ihlas
One more example: AUDUSD should break the low first, target 0.64479, max DD zone 0.65858
It has broken H, so not valid anymore
  • #612
  • May 29, 2023 7:18am May 29, 2023 7:18am
USDCAD: Our new target is May 23 Low
  • #613
  • May 30, 2023 3:31am May 30, 2023 3:31am
Quoting Ihlas
USDCAD: Our new target is May 23 Low
Cancel it; EURCAD acted quickly

Instead EG from here, 300 pips up
  • #614
  • May 30, 2023 5:52am May 30, 2023 5:52am
Quoting alphaomega
Hi, and thanks for stopping by! Regardless if you're new or old member of the forum, keep reading here and I promise you will not be disappointed! From my point of view this thread is the natural continuation of the thread "Higher Edge within a Single Candlestick” and here I intend to share with you some of my discoveries regarding this concept. As I commented in the original thread, I believe that we can create and apply multiple strategies only by using the information which is...

Should we allow to pass some time to enter into sell after current price = cycle high condition is met? The price can retest cycle high in 30 seconds, 1 minute, 2 hrs etc. Not clear from your post how it should be done
  • #615
  • May 30, 2023 6:17am May 30, 2023 6:17am
Quoting Ihlas
{quote} Cancel it; EURCAD acted quickly Instead EG from here, 300 pips up
EG today's H is a piece of cake to reach
  • #616
  • May 31, 2023 5:56am May 31, 2023 5:56am
Quoting Ihlas
USDCAD: Our new target is May 23 Low
Now we go from here May 24 low
  • #617
  • Jun 1, 2023 7:59am Jun 1, 2023 7:59am
Quoting Ihlas
{quote} EG today's H is a piece of cake to reach
It wasn't that easy

Quoting Ihlas
{quote} Now we go from here May 24 low
This was easy
  • #618
  • Jun 24, 2023 8:39pm Jun 24, 2023 8:39pm
Quoting marengo10021
Hey all, Ever since I came across this thread I started thinking of ways to actually make profit in the market without using technical analysis--which, besides analyzing MAs speed/acceleration/convergence/divergence, feels like pure astrology to me-- and fundamental/sentiment analysis--which, at a retail level, feels very hard to get a complete grasp on. I have actively been trying to decrypt alphaomega's messages in this thread on mean reversion. I personally tried to understand what he meant through those messages, overall I had a few ideas but...
Disclosure: I don't trade with AO's method. But I've been a follower of his posts for years. I have tested, analyzed but haven't gone live with his methods before.

AO's thread has fascinated me for so long. His messages are quite cryptic indeed, but he's also shared plenty of interesting nuggets that may help you with your trading. He's pretty much laid out the entire system in the first few pages of the post, albeit with obvious flaws when a major trend hits. He's suggested filtering out major news events, and optimizing the mid cycle hour according to each pair. I'm personally not a fan of filtering out major news events and optimizing the hour, but there may be merit in that.

I have recently jumped back into analyzing AO's content, and here's what I will share:
1) The 3 market models that he proposed exist indeed. Find a way to mechanically classify each day into each market model. I have simplified them into just 2 models. Ranging days (Including flat ranging days and ranging days with a directional drift), and trending days (Trending days with a directional drift).
2) How the market moves determines when it is the best time for you to enter into a mean reversion strategy. (What kind of mean-reverting strategy doesn't seem to be material. The key is the filter, not the actual strategy itself).
3) AO's original trading rules can act as a decent mean-reverting strategy. You can use your own methods to test out the efficacy as well.


Combine 1, 2 and 3.
  • #619
  • Sep 6, 2023 9:45am Sep 6, 2023 9:45am
Hello @alphaomega

can i get the mql file of simple mean ea .. i think large take profit and stop loss will be good to go.

thank you
  • #620
  • Sep 16, 2023 5:44am Sep 16, 2023 5:44am
Quoting alphaomega
Initial testing from Jan 1, 2018, standard rules, without optimization: {image} {file}
Hello @alphaomega. Its very interesting strategy^ thank you for posting.

But, I have a request to you, like the previous user. Can you provide mql of your robot. I need this for two reasons - I want to create a bot using this strategy for the cTrader platform, I don’t quite understand and cannot clearly understand from your explanation its rules, such as where to enter the market and under what conditions
0 traders viewing now