Higher Edge within a Single Candlestick | Page 124 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

An Edge for an Edge (System Development) 116 replies

Beating the Odds - Edge Within a Single Candlestick (Explained) 71 replies

Looking for higher timeframe candlestick overlay indicator 10 replies

top 2 single candlestick pattern 7 replies

There is No Edge in a Single Candlestick 64 replies

Attachments: Higher Edge within a Single Candlestick
Exit Attachments

Higher Edge within a Single Candlestick

  • #2,461
  • Mar 4, 2025 9:44pm Mar 4, 2025 9:44pm
Quoting W0lfram
{quote} I didn't express a single opinion, that were facts {quote} psss.. no one give a shit about it. Best of luck edit: Actually, you sound like it's your first week in business and demo acc you have inherited from someone. - My first opinion here
I already said "Best of luck, but you just can't seem to STFU. The truth is, the only reason you're saying you don’t care about it is because you don’t have it. Just like you have nothing in trading other than words, you have no results to back them up. It takes just a moment to create a demo and show everyone, even a few days demo from you would be enough. But you won’t do it, because you can’t.
  • #2,462
  • Mar 4, 2025 10:04pm Mar 4, 2025 10:04pm
Quoting W0lfram
{quote} I didn't express a single opinion, that were facts {quote} psss.. no one give a shit about it. Best of luck edit: Actually, you sound like it's your first week in business and demo acc you have inherited from someone. - My first opinion here
I just noticed this part, and this will be my last reply to you. If this is truly what you believe, then you're actually dumb, and I owe you an apology for ever engaging in this conversation. I assumed you had at least some reasoning ability, but I gave you too much credit. If you seriously think that’s a reasonable conclusion, you’ve proven yourself to be completely out of your depth. Only a fool would come to that assessment. I apologize for the words battle, please go back to whatever you were doing before this. I apologize again, I did not mean to go back and forth with someone at this level. Peace.
  • #2,463
  • Mar 4, 2025 10:05pm Mar 4, 2025 10:05pm
Quoting Th3W1z4rd
{quote} I already said "Best of luck, but you just can't seem to STFU. The truth is, the only reason you're saying you don’t care about it is because you don’t have it. Just like you have nothing in trading other than words, you have no results to back them up. It takes just a moment to create a demo and show everyone, even a few days demo from you would be enough. But you won’t do it, because you can’t.
I just warned the others not to argue with proud demo acc owner.
U can't STFU and keep clucking about it.
Btw, where is the older ones? (rhetoric question)
Attached Image (click to enlarge)
Click to Enlarge Name: chrome_8taAyryqAs.png Size: 6 KB



Quote
It takes just a moment to create a demo and show everyone, even a few days demo from you would be enough.
:facepalm:
I told you about math, patterns in real and synth data. Asked you to back up you statistical claims. You just fussing around with your non-representative account.

You are right: I don't have demo and can't set it up, bc why the hell?
Air force
  • #2,464
  • Mar 4, 2025 10:11pm Mar 4, 2025 10:11pm
Quoting W0lfram
{quote} I just warned the other not to argue with proud demo acc owner. U can't STFU and keep clucking about it. Btw, where is the older ones? (rhetoric question){image} {quote} :facepalm: I told you about math, patterns in real and synth data. Asked you to back up you statistical claims. You just fussing around with your non-representative account.
It seems you may have missed when I mentioned that I'm trading live. This demo is just something I shared to test a few things on the side. And, I’m not getting into another words battle with you. That would be like playing chess against someone who thinks the pieces are just decorations. Completely unfair, and honestly, pointless. Have a great day Peace.
  • #2,465
  • Mar 4, 2025 11:32pm Mar 4, 2025 11:32pm
Quoting Th3W1z4rd
I took a quick look at the post you mentioned, and I saw the condition that "The data must be stationary and symmetrically distributed around the E(X)." This condition has a big flaw. Stationary data means that its key properties, like the average and how much it fluctuates, don't change over time.
Correct. It of course won't be applicable to trading. But I just want to show you that random data can be beaten, too!

Quoting Th3W1z4rd
But real financial markets don't work this way. Markets constantly change. Trends shift, volatility goes up and down, and major events like policy changes or economic crises shake things up.
Maybe, I really don't know. I would say no. However, it's of course fine if you disagree with me. I just like to encourage you to prove your believes right or wrong. Collect data from the market, think about how you can prove your hypotheses.

I mean, I somewhat can agree with you. I personally speak about the market as a whole, an entirety. Therefore, I talk about the average behavior when it comes to probabilities. However, if you are only looking at particular time spans, the sample size gets smaller an smaller and your estimate of the probability, the proportion, varies more and more. Yes, in short time spans you can thus of course see many trends and many states that can either support people believing in trend trading as well as those believing in mean reversion or any other kind. You'll find beautiful trends in roulette spins, too if you by chance get to the right wheel at the right time and left the table before you see randomness kicking in again.

Quoting Th3W1z4rd
Stock markets tend to rise over time because of growth and inflation. Currencies follow economic policies and interest rate differences, leading to long trends instead of balanced movements. Commodities go through supply and demand cycles, which also create trends rather than symmetric price moves. Momentum is another issue. Many assets keep moving in the same direction for a while because of investor behavior, institutional trading, and large scale trends. If returns were truly balanced around the average, momentum strategies would not work, but...
All correct. At least in the FX, you'll mostly only find very restricted (very short-term) effects of news like policies and interest rates. Thus, no deviation from randomness. And there are some seasonal trends in some commodities and stocks related to agriculture. But again: try to prove that these are indeed present. Believe me, it's incredibly hard to truely find any repeatable patterns in the market. Just try it yourself - you'll be impressed!

Another gem of CP: For every line break you show me, I show you another that doesn't work.
  • #2,466
  • Mar 5, 2025 12:07am Mar 5, 2025 12:07am
Quoting AntiCre
{quote} Correct. It of course won't be applicable to trading. But I just want to show you that random data can be beaten, too! {quote} Maybe, I really don't know. I would say no. However, it's of course fine if you disagree with me. I just like to encourage you to prove your believes right or wrong. Collect data from the market, think about how you can prove your hypotheses. I mean, I somewhat can agree with you. I personally speak about the market as a whole, an entirety. Therefore, I talk about the average behavior when it comes to probabilities....
You cannot consistently beat truly random price data over time without introducing some kind of external bias.

Short term profitability on random data is possible due to variance (luck). If you flip a fair coin 10 times, you might get 7 heads, but that doesn’t mean you have an edge in predicting heads. In the same way, a trading strategy might show profit over a small sample of random price data, but that doesn't prove a real edge.

Profitability requires a statistical edge, which cannot exist in purely random data. If prices are truly random, meaning they follow a Brownian motion or a perfectly efficient market, there is no exploitable structure. Any trading strategy applied to pure randomness would eventually break even after costs or lose money due to spreads, commissions, and slippage.

If a strategy consistently "beats" random data, then the data is not truly random. If someone claims they are profitable on synthetic random data, then either their strategy isn’t actually trading randomness and is instead exploiting some bias in how the data was generated, or the results are due to chance and would fail over a long enough period.

A way to test this would be to generate truly random price data, like Brownian motion or normally distributed returns, then apply a trading strategy that supposedly works. If profitability holds up over a large enough sample, meaning hundreds or thousands of trades, and results deviate from what would be expected in a purely random environment, then there may be some bias in the data itself. Comparing these results against expected outcomes from pure randomness, such as through Monte Carlo simulations, would tell if any edge actually exists.

Any serious person on this topic will tell you that without structure in the data, profitability cannot hold up. That’s why traders look for patterns, market inefficiencies, and institutional behavior, because price isn't completely random in reality.

If markets were truly random, trading would be nothing more than a coin flip and unprofitable over time. Since some traders do consistently profit, the market is not fully random. But claiming you can "beat" completely random data in the long run shows a misunderstanding of randomness.
  • #2,467
  • Mar 5, 2025 5:09am Mar 5, 2025 5:09am
Quoting Th3W1z4rd
{quote} You cannot consistently beat truly random price data over time without introducing some kind of external bias. Short term profitability on random data is possible due to variance (luck). If you flip a fair coin 10 times, you might get 7 heads, but that doesn’t mean you have an edge in predicting heads. In the same way, a trading strategy might show profit over a small sample of random price data, but that doesn't prove a real edge. Profitability requires a statistical edge, which cannot exist in purely random data. If prices are truly random,...
Have you done any statistics and you get data that there are non random moves in the market?
From my knowledge and statistics about market, I think none has true edge with positive pips. Everyone has 50:50 probability, but plays with money/risk management with price movement, because knows that price has to move some distance through day/week/month. And you can be succesful several years. But in the longterm, more than 10 years, there will be market conditions, where you lose your account. But you can trade this way too, maybe you take more profit than you loose and it is worth for you in the end.
  • #2,468
  • Mar 5, 2025 5:59am Mar 5, 2025 5:59am
Quoting Th3W1z4rd
If markets were truly random, trading would be nothing more than a coin flip and unprofitable over time. Since some traders do consistently profit, the market is not fully random. But claiming you can "beat" completely random data in the long run shows a misunderstanding of randomness.
It may be a coincidence that nobel prizes have been awarded for both the theory of random markets and the theory of behavioral markets.

Many people seem to have a hard time accepting that in the short term markets can be driven (and it has been proven that they were), and in the long term they inevitably show random characteristics. How could it be otherwise, after all? in the short term speculators invest big to make profits, in the long term the positions and the different time-frame lead to a behavior that is no longer predictable. It seems pure logic to me
  • #2,469
  • Mar 5, 2025 5:59am Mar 5, 2025 5:59am
Quoting IenDzi
{quote} Have you done any statistics and you get data that there are non random moves in the market? From my knowledge and statistics about market, I think none has true edge with positive pips. Everyone has 50:50 probability, but plays with money/risk management with price movement, because knows that price has to move some distance through day/week/month. And you can be succesful several years. But in the longterm, more than 10 years, there will be market conditions, where you lose your account. But you can trade this way too, maybe you take more...
Markets aren't fully random. Non random moves exist, driven by liquidity shifts, institutional order flow, economic cycles, and market structure. If price movement were truly 50:50, momentum strategies wouldn’t work, mean reversion wouldn’t work, and institutions wouldn’t have a consistent trading edge. But they do.

Saying that every trader will eventually lose their account over 10+ years also doesn’t make sense. If that were true, hedge funds and institutional traders wouldn’t be able to stay profitable for decades. Markets evolve, but so do strategies, and traders who adapt can continue to succeed.

The fact that traders can be profitable for several years, as you admitted, already proves that markets are not purely random. If they were, no one would last beyond short term variance (luck). The real question isn’t whether an edge exists, but whether someone can identify an edge and adjust to it over time.
  • #2,470
  • Mar 5, 2025 6:10am Mar 5, 2025 6:10am
Quoting Fiox89
{quote} It may be a coincidence that nobel prizes have been awarded for both the theory of random markets and the theory of behavioral markets. Many people seem to have a hard time accepting that in the short term markets can be driven (and it has been proven that they were), and in the long term they inevitably show random characteristics. How could it be otherwise, after all? in the short term speculators invest big to make profits, in the long term the positions and the different time-frame lead to a behavior that is no longer predictable. It...
I'm not surprised that Nobel Prizes have been given for both the theory of random markets and the theory of behavioral markets because markets aren’t completely one or the other. They aren’t fully random, and they aren’t fully predictable. Instead, they’re influenced by liquidity, institutional positioning, and trader behavior.

Short term price movement is clearly driven by things like order flow and liquidity hunts, and other institutional strategies. That’s why short term edges like momentum and mean reversion exist. Just look at my TradeExplorer for real examples of pairs you can analyze. The idea that markets automatically become random in the long run assumes that all structure disappears over time, but that’s not how markets work. Long term trends in stocks, commodities, and currencies happen because of macroeconomic cycles, central bank policies, and capital flows.

If markets were truly random over the long term, we wouldn’t see multi year trends in stocks, commodities, or forex that line up with fundamental changes. Institutions wouldn’t be able to profit for decades. Just because the timeframe expands doesn’t mean inefficiencies stop existing. New ones appear, and traders who adapt keep finding setups.

It might sound logical to assume markets become random over time, but I don't see this idea in real world price behavior. The truth is more complicated than just simply calling markets "random" or "predictable."
  • #2,471
  • Mar 6, 2025 8:44am Mar 6, 2025 8:44am
You guys miss a reference point!

- "Dude, don't you know how to read, I have backtested any possible distance in pips to any tick/time length combination"

Attached Image (click to enlarge)
Click to Enlarge Name: image1.png Size: 108 KB

- Ok man peace! Theres only one way to trim your buds right, its only natural!
  • #2,472
  • Mar 6, 2025 8:57am Mar 6, 2025 8:57am
Quoting W0lfram
{quote} {quote} Well, that's completely false. truly/falsely random Price has stochastic nature. I don't want to dive too deep and write novels here, briefly: If you generate synthetic tick-data, then compose it into candles (https://www.forexfactory.com/thread/...9#post15111019), you'll find there every single pattern, known and unknown. And I'm quite sure, real and synth will have ~same periodicity of those patterns (https://www.forexfactory.com/thread/...3#post15126933) {quote} I would like to see statistical...
Wolfram, with a username like yours, I'm sure you you're just taking the piss and I'm 99% certain you already know. To create random data that resembles asset prices you need to use a garch model because volatility clusters. And the mere existence of vol clusters proves a price dependency on what's just elapsed. But even that will be different from real world asset prices. You need to add intraday seasonality, predictable volatility events, structural events like the cut and roll, delivery times for fx. And then you're getting close, but that's still simply incorporating all the fixed differences between price and random movement that are widely known. Shortcut = use real data.

On another note - CP - these convos always get tetchy because if you've been exposed to even a glimmer of quant financial modelling, then by comparison CP looks like amateur voodoo. His grammar is all over the shop, he comes out with things completely refuted by current trad finance thinking and makes wild claims to boot. So much so, I can easily he how he's mistaken for some of real shills who have grifted on this site over the years.

But I found his posts invaluable. Some are direct hints as to what to look at, others are to show different ways of thinking. To get ahead you need to do what everyone else is doing but better, or do it differently. He simply chose different and asked for nothing in return.
  • #2,473
  • Mar 6, 2025 9:33am Mar 6, 2025 9:33am
Quoting triphop
{quote} Wolfram, with a username like yours, I'm sure you you're just taking the piss and I'm 99% certain you already know. To create random data that resembles asset prices you need to use a garch model because volatility clusters. And the mere existence of vol clusters proves a price dependency on what's just elapsed. But even that will be different from real world asset prices. You need to add intraday seasonality, predictable volatility events, structural events like the cut and roll, delivery times for fx. And then you're getting close, but...
Quote
Wolfram, with a username like yours, I'm sure you you're just taking the piss and I'm 99% certain you already know.
Maybe
Quote
To create random data that resembles asset prices you need to use a garch model because volatility clusters.
I've mentioned it and others in the next post with some sarcastic tone.
Quote
Of course it's impossible to incorporate spreads, market depth, slippage, volatility clusters, etc. into model.
There are advanced models around (even w/ synt order book); GANs, GARCH, Fama-French, etc
As the saying goes, all models are wrong, but some are useful.
In this post, quoted earlier (https://www.forexfactory.com/thread/...9#post15111019)
Inserted Code
{probability} -> {tick size}
{0.0005, 0.0245, 0.075, 0.8, 0.075, 0.0245, 0.0005} -> {-0.001, -0.0002, -0.0001, 0, 0.0001, 0.0002, 0.001}
simple logic worked quite well.
idk, if you followed the links from my post, here is a result.
I was testing some reversion 'strategy' on it (hourly data for GU).
Then I found twice the volume of historical data -- percentages are the same. +/- of course. Could be some fundamental property of given 'real data' and corresponding synth random walk. Not sure
Attached Image (click to enlarge)
Click to Enlarge Name: WolframNB_oIi8CtPEEI.png Size: 5 KB
Air force
  • #2,474
  • Mar 6, 2025 10:21am Mar 6, 2025 10:21am
- Dear God, I've lost faith, please send me a sign that you exist
- Imbecile! May you forever wander in random walks!

https://chatgpt.com/share/67c9c66e-e...d-7fcc033fcfd4
  • #2,475
  • Mar 6, 2025 1:25pm Mar 6, 2025 1:25pm
Quoting W0lfram
{quote} {quote} Maybe {quote} I've mentioned it and others in the next post with some sarcastic tone. {quote} As the saying goes, all models are wrong, but some are useful. In this post, quoted earlier (https://www.forexfactory.com/thread/...9#post15111019) {probability} -> {tick size} {0.0005, 0.0245, 0.075, 0.8, 0.075, 0.0245, 0.0005} -> {-0.001, -0.0002, -0.0001, 0, 0.0001, 0.0002, 0.001} simple logic worked quite well. idk, if you followed the links from my post, here is a result. I was testing some reversion...
you were easily unearthed
What you found may be useful but I don't use probabilistic models at all tbh. Not because they've no value; simply because I terrorised myself reading Prado & Chan and made a solemn vow to never, ever compete in fields I can't win. My maths is nowhere near strong enough, and from what I've seen most true quants - a few polymaths excepted - are woefully under-resourced compared to the competition.
It forced me to find a different path. Eventually a blessing. A blessing with a bloody long wait mind
  • #2,476
  • Mar 6, 2025 1:26pm Mar 6, 2025 1:26pm
Quoting chriseel
- Dear God, I've lost faith, please send me a sign that you exist - Imbecile! May you forever wander in random walks! https://chatgpt.com/share/67c9c66e-e...d-7fcc033fcfd4
On average you post once a year, you're the epitome of restraint. Nice gpt thread.
  • #2,477
  • Edited 4:14pm Mar 6, 2025 1:32pm | Edited 4:14pm
Data: random.org. Atmospheric noise as randomizer. Not some pathetic pseudo-random machine. Truly random.
I took a set of integers {1,2}, then replaced 2 with -1. Should be 50/50
Attached Image (click to enlarge)
Click to Enlarge Name: WolframNB_7QF6VgT2HP.png Size: 16 KB

Someone says that momentum/reversion wouldn't work.
They can't fail to work simultaneously. Something will.

Time unit is abstract. Take any scale.
Clearly visible the absence of a trend.
Attached Image (click to enlarge)
Click to Enlarge Name: WolframNB_PPdb9LUs0d.png Size: 35 KB

Renko brick: 5.
This charting style is cool w/ super-trend and some MM machinations...
Attached Image (click to enlarge)
Click to Enlarge Name: WolframNB_X5xJNpsQbc.png Size: 9 KB
Air force
  • #2,478
  • Mar 6, 2025 2:00pm Mar 6, 2025 2:00pm
Quoting AntiCre
{quote} Correct. It of course won't be applicable to trading. But I just want to show you that random data can be beaten, too! {quote} Maybe, I really don't know. I would say no. However, it's of course fine if you disagree with me. I just like to encourage you to prove your believes right or wrong. Collect data from the market, think about how you can prove your hypotheses. I mean, I somewhat can agree with you. I personally speak about the market as a whole, an entirety. Therefore, I talk about the average behavior when it comes to probabilities....
It's nice to see you back. Your posts were very insightful in my market stats exploration.
Air force
  • #2,479
  • Mar 6, 2025 4:17pm Mar 6, 2025 4:17pm
Quoting W0lfram
{quote} It's nice to see you back. Your posts were very insightful in my market stats exploration.
Thank you. Yours inspired me, too! A great pleasure to see you here!
  • #2,480
  • Mar 6, 2025 6:26pm Mar 6, 2025 6:26pm
Quoting triphop
... then by comparison CP looks like amateur voodoo.

... But I found his posts invaluable. Some are direct hints as to what to look at, others are to show different ways of thinking. To get ahead you need to do what everyone else is doing but better, or do it differently. He simply chose different and asked for nothing in return.
An invaluable amateur voodoo... a double pseudo-oxymoron

Like the smart and intelligent people posting here; knowing that tomato is a fruit (not a vegetable).
But they lack the Wisdom to not mix tomato into a fruit salad.

Like what's the point of being smart, but lacks applicable wisdom? (rhetorical)
1 trader viewing now