The Myth of Averaging | Page 4 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

1% Money Management is a myth, fib is a myth and so on.. 45 replies

'Averaging Down Is for Losers' - valid or not? 27 replies

Fast Averaging versus Slow Averaging 1 reply

Averaging 7 replies

Martigale and Averaging Down 4 replies

Attachments: The Myth of Averaging
Exit Attachments

The Myth of Averaging

  • #61
  • Nov 27, 2023 6:55am Nov 27, 2023 6:55am
Yoriz. Could you do a test with just one position 1:1 TP/SL RR ratio but with a trailing stop. And then another test with no TP, but only SL trailing stop?
  • #62
  • Nov 27, 2023 12:37pm Nov 27, 2023 12:37pm
Quoting YoMoney
Could you do a test with just one position 1:1 TP/SL RR ratio but with a trailing stop. And then another test with no TP, but only SL trailing stop?
Interesting question. Please give it a try and let us know what you find? You can download the averaging up and down EAs from this thread.
  • #63
  • Nov 30, 2023 10:08pm Nov 30, 2023 10:08pm
Quoting YoMoney
Yoriz. Could you do a test with just one position 1:1 TP/SL RR ratio but with a trailing stop. And then another test with no TP, but only SL trailing stop?
But where to take profit? Manually or it's a kind of position that is held indefinitely?
  • #64
  • Dec 1, 2023 12:20pm Dec 1, 2023 12:20pm
Quoting yoriz
{quote} Interesting question. Please give it a try and let us know what you find? You can download the averaging up and down EAs from this thread.
I would if I had the mt and the data.
  • #65
  • Dec 1, 2023 12:25pm Dec 1, 2023 12:25pm
Quoting pizd0s
{quote} But where to take profit? Manually or it's a kind of position that is held indefinitely?
That depends on where the target is set. If it's 10 pips then it's 10 pips profit or loss for a 1:1 ratio. All it needs is trailing stop with it.
The other test is for no TP, just trailing stop. You let it run untill it get's stopped out. Can try diffrent amount of pips. Not too many since it would take a long time to get enough trades.
  • #66
  • Edited 7:33am Dec 14, 2023 7:17am | Edited 7:33am
Quoting yoriz
{quote} What I tried to check in this thread is to see whether averaging actually does anything useful. Of course a profitable strategy with an edge would be better. However, given that even the fully random strategy I tested can be profitable, this seems to suggested averaging does indeed improve your odds. So averaging is clearly a useful component to include to your strategies. See also post #52.
I think you've done some good research and found out what doesn't work. Sometimes coding an EA can remove you from the core of what you are trying to find out.
Manually back testing and collating stats works for me, it takes longer but just like handwriting I absorb more knowledge and observations, not sure this works with EA backtesting?

There are many ways to average, most of them with appalling results. I won't divulge what I do but your edge lies in finding an outcome with a high probability and a binary result. Hope that helps and good luck
  • #67
  • Dec 18, 2023 10:15am Dec 18, 2023 10:15am
Quoting Fagin
I think you've done some good research and found out what doesn't work.
And what does work, as you can see.

Quoting Fagin
Sometimes coding an EA can remove you from the core of what you are trying to find out.
Actually, exploring averaging was exactly what I wanted to find out in this thread, hence the title :-)

Quoting Fagin
Manually back testing and collating stats works for me, it takes longer but just like handwriting I absorb more knowledge and observations, not sure this works with EA backtesting?
Everyone has their own approach. Personally, what I like about EA backtesting is that you can try out many different strategies with relatively low effort (if you know how to code, obviously). In an automated backtest you can quickly scroll to those trades that were very bad and see if/how you can filter them and run the test again. Testing a strategy manually over hundreds of trades takes a lot of time and is hard to do consistently.

Quoting Fagin
your edge lies in finding an outcome with a high probability and a binary result
What do you mean by "binary result"?
  • #68
  • Jan 14, 2024 2:28am Jan 14, 2024 2:28am
From what I've skimmed it looks like all you've done is the classic curve fit to a specific slice of time. It almost certainly won't carry forward. So while you intend to prove the mathematicians wrong (or however you put it) it's all but certain you've proved them right.

It's also incredibly underwhelming in the amount of profits made (in hindsight) over the time it took to go from start to end. Time is money.

If you have a time machine, yes, go back and trade using the curved parameters and go hard.
Congratulations ScalaFX for coming out of the closet!
  • #69
  • Jan 15, 2024 12:49pm Jan 15, 2024 12:49pm
nubcake, you can't make very critical statements if you just "skimmed" the thread, as you call it.

Quoting nubcake
It almost certainly won't carry forward.
Please read the footnote on post #2. Then try it out yourself.

Quoting nubcake
It's also incredibly underwhelming in the amount of profits made
Please read post #52.
  • #70
  • Jan 15, 2024 1:08pm Jan 15, 2024 1:08pm
Quoting yoriz
Introduction Mathematicians will tell you that a strategy without an edge can not be made profitable with "smart" money management. Grid, averaging down, Martingale, etc. will all fail in the long run when an unfortunate series of loosing trades drains your account. In this thread I would like to do some silly experiments with averaging down to verify that idea. Simple Expert Advisor I created a simple expert advisor for Metatrader 5 to do some tests. You can download the EA below, if you want to play along. Basically: Each day we make a random...
Hi yoriz
I read your topic and I found it amazing. Thanks for sharing it. Is it possible to provide us with the source code in .mq5 format?
  • #71
  • Jan 15, 2024 1:21pm Jan 15, 2024 1:21pm
Quoting my7amin
I read your topic and I found it amazing. Thanks for sharing it.
Thanks, glad you liked it!

Quoting my7amin
Is it possible to provide us with the source code in .mq5 format?
No, sorry. It uses several proprietary libraries I can not share. However, the logic is obviously very simple, so I am sure you can have a coder write it with little effort for a low fee.
  • #72
  • Edited 1:30am Jan 16, 2024 1:16am | Edited 1:30am
Quoting yoriz
nubcake, you can't make very critical statements if you just "skimmed" the thread, as you call it. {quote} Please read the footnote on post #2. Then try it out yourself. {quote} Please read post #52.
Actually, I can. I've been around long enough to know this thread is all about curve fitting and nothing more. You've mined historical values for magical numbers that magically worked.

You haven't addressed this fact, nor the fact that it seems like the profits shown on the backtest were a waste of effort.

Throwing shit at the all and seeing what sticks doesn't disprove anything. As I said, if anything you've actually proven the "mathematicians" correct, you just don't realize it. I've seen nothing that isn't what thousands of other newbies haven't done thousands of times before.

If you've done something astonishingly novel and new you should have mentioned what that is. Otherwise, it's curve fitting and nothing more.

I'm annoyed at your arrogance for implying there's anything here, and for not addressing that this is the definition of curve fitting! How dare you suggest someone read all of the posts in this thread when it's a waste of time and you don't even know that you've done nothing remarkable here whatsoever! Shame on you.
Congratulations ScalaFX for coming out of the closet!
  • #73
  • Jan 16, 2024 2:22am Jan 16, 2024 2:22am
Quoting nubcake
I'm annoyed at your arrogance for implying there's anything here, and for not addressing that this is the definition of curve fitting! How dare you suggest someone read all of the posts in this thread when it's a waste of time and you don't even know that you've done nothing remarkable here whatsoever! Shame on you.
Everything ok at home?

This is not curve fitting but measuring the statistical properties of a symbol. What are typical ranges prices move before they revert? That behavior seems to be fairly consistent throughout the years. So yes, the results do carry forward, which you could have easily tried out yourself with the EA I provide here if you took the effort. But why do I even bother replying to your post...
  • #74
  • Jan 16, 2024 2:29am Jan 16, 2024 2:29am
Quoting yoriz
{quote} Everything ok at home? This is not curve fitting but measuring the statistical properties of a symbol. What are typical ranges prices move before they revert? That behavior seems to be fairly consistent throughout the years. So yes, the results do carry forward. But why do I even bother replying to your post...
Because you know I'm right and don't want to address it properly, so by replying you feel like you've at least shown to be trying to save face. You haven't.

What I see is someone who doesn't know what they are doing. You will probably find a few interesting bits of data, maybe, but this thread is cooked... the premise is bullshit, the conclusions are specious, and your replies so far address nothing relevant.

You haven't found an edge, and you haven't proven you don't need an edge. I'm out. Try to save face, or whatever, but this thread isn't what you dress it up to be. Those who know, know already.
Congratulations ScalaFX for coming out of the closet!
  • #75
  • Jan 16, 2024 4:00am Jan 16, 2024 4:00am
Quoting nubcake
{quote} Because you know I'm right and don't want to address it properly, so by replying you feel like you've at least shown to be trying to save face. You haven't. What I see is someone who doesn't know what they are doing. You will probably find a few interesting bits of data, maybe, but this thread is cooked... the premise is bullshit, the conclusions are specious, and your replies so far address nothing relevant. You haven't found an edge, and you haven't proven you don't need an edge. I'm out. Try to save face, or whatever, but this thread...
Why all the imperious grandstanding lol. It is just a bit of theorizing and experimenting so what is the harm? Even if this experiment seemed rudimentary or specious to you, could it not be part of someone else's empirical learning process to perhaps to lend perspective? You seem to be concerned with being right and the premise of the conclusions being wrong. However conclusions in the experimental shouldn't always be right. You need to be right and wrong sometimes in order to arrive at the truth or a better understanding of the situation. A true empiricist loves being wrong because it grants deeper understanding. Just let the process of rights and wrongs in this experiment play out without the snarky criticism. Don't be afraid to be wrong my friend. You'll find you do the most growth in those critical moments.
  • #76
  • Jan 16, 2024 4:10am Jan 16, 2024 4:10am
Quoting sharingan9
{quote} Why all the imperious grandstanding lol. It is just a bit of theorizing and experimenting so what is the harm? Even if this experiment seemed rudimentary or specious to you, could it not be part of someone else's empirical learning process to perhaps to lend perspective? You seem to be concerned with being right and the premise of the conclusions being wrong. However conclusions in the experimental shouldn't always be right. You need to be right and wrong sometimes in order to arrive at the truth or a better understanding of the situation....
What garbage. OP has a thread "proving" that it's possible to average down/up without an edge and be profitable. It's a curve fit, and nonsense. End of story.

Oh I meant to ignore this thread. That's my bad <- see, I can admit fault when I fault. Unlike the rest of you.
Congratulations ScalaFX for coming out of the closet!
  • #77
  • Jan 16, 2024 4:13am Jan 16, 2024 4:13am
I personally disagree with the mathematicians or at least the phrasing of their argument in the original post. "Mathematicians will tell you that a strategy without an edge can not be made profitable with "smart" money management".. This statement is rather contradicts itself as "smart" money management is an edge. In essence the statement reads, "A strategy without an edge can not be made profitable with an edge" which is to say that trading is not a profitable venture lol. I dont believe martingale is "smart money management" however if the mathematicians argue that an edge is required for indelibly profitable trading then smart money management can be an edge and there are many sophisticated ways to manage losses and wins to be consistently profitable. This is treading on the discretionary though as well as the psychological in that one would need to determine how much one is willing to lose and how comfortable are they with losses which depends on available capital and risk assessment.
  • #78
  • Jan 16, 2024 4:20am Jan 16, 2024 4:20am
Quoting nubcake
{quote} What garbage. OP has a thread "proving" that it's possible to average down/up without an edge and be profitable. It's a curve fit, and nonsense. End of story. Oh I meant to ignore this thread. That's my bad <- see, I can admit fault when I fault. Unlike the rest of you.
Right, his conclusion may be wrong or right and you may disagree. Either way it is all part of the learning process. Maybe his is right and he learned something. Maybe the conclusion is wrong and he won't realize it until much later in his trading. Who cares it's all part of the process of experimenting. There have been Nobel Prize winners that have been proven wrong but their work still lead to advancements despite wrong conclusions lol. I wish you green pips though my friend.
  • #79
  • Jan 17, 2024 10:49am Jan 17, 2024 10:49am
Quoting sharingan9
Who cares it's all part of the process of experimenting.
Thank you for your support, sharingan9, but just put him on the ignore-list as he is a troll here on FF. Just check the last two threads he started: one implying that real traders hedge and Martingale (yes, really!! And he is upset about my averaging down thread?!?!), and in another thread he is bashing other FF users in rhyme?!
  • #80
  • Jan 18, 2024 2:47am Jan 18, 2024 2:47am
Quoting yoriz
{quote} Thank you for your support, sharingan9, but just put him on the ignore-list as he is a troll here on FF. Just check the last two threads he started: one implying that real traders hedge and Martingale (yes, really!! And he is upset about my averaging down thread?!?!), and in another thread he is bashing other FF users in rhyme?!
Alright, I'll tag back in to clarify for the slow bus since I saw this pop-up.

It seems you can't tell when someone is making threads to shitpost about the state of the forum (lots of retarded martingale talk at the time I made the thread) versus when they are correctly telling you the things you claim are without merit. Looks like you are just another one who knows so very little that anyone who knows a lot more than you seems like a troll because you just can't keep up. You do you, boo.

As I said in the thread, as a shitpost, big brains discuss, Here you've done so then complained when nobody continued to post here. Bring something new and correct, not rookie nonsense, and there would actually be something further to discuss. There simply isn't. A whole bunch of posts about curve fitting for almost no profit. That's not a win for the concept of averaging down/up randomly, that's a loss.

You find a way to trade without an edge and you'll have done what seems impossible. Even those who claim to trade randomly and survive a long career really aren't trading randomly, whether they know it or not or can articulate the edge or not.

Again, what is there further to discuss? I've told you this is curve fitting and you've proved nothing. You've provided nothing but a little retort stating you are actually just measuring MFE and MAE. Ok. Great. That's... cool... but not much more to discuss still. Whether you are trying to prove, or disprove, the effects or utility or whatever regarding averaging down/up, you've shown that it almost always ends in disaster except a small range of values over a certain time period that produced hardly any profit.

You've posted your results, people have looked it at, and moved on. What more do you want? What more is there to discuss?

Don't ever learn anything. Put me on ignore because I'm just a dumb troll
Congratulations ScalaFX for coming out of the closet!
0 traders viewing now