Questions for real pros: How to trade with reference points only? | Page 5 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Price action trading - Trade like the pros! 964 replies

Retail Trade Gains vs Pros 6 replies

newbie questions on daily pivot points 18 replies

Retail Sales? Trade Balance? What it means? Newbies ask here. Pros answer here. 9 replies

Attachments: Questions for real pros: How to trade with reference points only?
Exit Attachments

Questions for real pros: How to trade with reference points only?

  • #81
  • Jan 6, 2024 6:55am Jan 6, 2024 6:55am
Quoting oochiwaawaa
{image} found this could be helpful. i make no comment about others on these forums anymore. not worth it. equations in screenshot have worked consistently since rice trading used candlesticks. a person creates their own trading wisdom. its always about why you do it. have fun. O.
Thanks for sharing. It's a pity that you don't make comments anymore. If nobody would make comments, it would be very silent and no new ideas would pop up. Discussing and having different arguments is ok as long as everybody defend their arguments in a polite way and accept that others can have other perceptions. Thus, I'd never say it is like this or that. It's always how I see things at this very moment (maybe I'll change my mind later on). Hopefully this may motivate you to continue posting/discussing!
  • #82
  • Jan 6, 2024 6:55am Jan 6, 2024 6:55am
Quoting ryuryu
Hi AntiCre! {quote} Price moves because of market orders executions. Market is not random. It moves because of money. Money makes money. If there will be no market orders there will be no movement and limit orders in the orderbook will stay there forever. Also stops are market orders too (always), so triggering stops also cause market order executions. This is the basics. There is no sense to dream or create ideas how it works. It works this way only and always. {quote} That's why this statements (or idea) make no sense at all. There are no reference...
people then in turn markets. money is just an idea. people create / make / agree to ideas. then reinforce them with choices. maybe more experience will help.

have fun.

O.
protein for breakfast !
  • #83
  • Jan 6, 2024 7:00am Jan 6, 2024 7:00am
Quoting AntiCre
{quote} Thanks for sharing. It's a pity that you don't make comments anymore. If nobody would make comments, it would be very silent and no new ideas would pop up. Discussing and having different arguments is ok as long as everybody defend their arguments in a polite way and accept that others can have other perceptions. Thus, I'd never say it is like this or that. It's always how I see things at this very moment (maybe I'll change my mind later on). Hopefully this may motivate you to continue posting/discussing!
a persons life is full of contradictory moments but to exist is different to living and learn. once u live and learn well, more is seen and understood. if people spent more time on this than making money then they would be better for it. hint=peers are peers coz they dont know better.

O.
protein for breakfast !
  • #84
  • Jan 6, 2024 10:08am Jan 6, 2024 10:08am
Quoting AntiCre
{quote} I agree again. You don't know how many "price changes" you had during the lifetime of a candlestick at just looking at price (well, you could use tick volume). However, to be consistent, I suggested in #post 69 to use 1000 price changes (ticks in the meaning of price updates, not 1/10th of 1 pip). That way one can exclude inconsistent measures. And in this case, if I'd apply your MA, I'd always get an "equal" estimate. But even using candlesticks may give an idea on how much price moves on the average tick number in the FX markets since...
1000 price change makes no big sense with additional context. What are the sizes? 1000 * 1$? or 1000 * 1m$?
Anyway, it is impossible to discuss everything here. I just wanted to point you. All I want to say that price moves not because of candle patterns it moves because of market order execution. Because of stops triggered. But definitely not because of some people see "cats" or "bart head" on the chart.
Observer effect
  • #85
  • Jan 6, 2024 10:28am Jan 6, 2024 10:28am
Quoting ryuryu
{quote} All I want to say that price moves not because of candle patterns
I don't think OP ever talked about candlesticks patterns or any kind of pattern, indeed the thread is basically advocating a whole different approach than the one you're describing.

Regards.
  • #86
  • Jan 6, 2024 10:53am Jan 6, 2024 10:53am
Quoting ryuryu
{quote} 1000 price change makes no big sense with additional context. What are the sizes? 1000 * 1$? or 1000 * 1m$? Anyway, it is impossible to discuss everything here. I just wanted to point you. All I want to say that price moves not because of candle patterns it moves because of market order execution. Because of stops triggered. But definitely not because of some people see "cats" or "bart head" on the chart.
As apprendista pointed out, I haven't mentioned candlestick patterns at all. And I was talking about 1000 ticks, i.e. the 1st candlestick is from now 999 ticks back. Then another tick comes in and I repeat that process. Why am I doing this? I'm interested to see how price moves from open to high/low, to low/high and finally to the close. This gives me an average, i.e. a number of what I can expect during a particular period. Why 1000? It's just an arbitrary number. And no, it doesn't matter to me whether this number is arbitrary or not. Why? Because regardless of the period I'm looking at, price movement shows some invariabilities. For instance, the body-to-range or wick-to-range ratios stay constant on average (for the average candlestick). Same applies for the "time" at which either the high or low (depending on whether the candlestick closes red or green) will show up (e.g. regardless of the tick range used to create a new candlestick, in bullish candlesticks the low is usually build around 25 % of the high at 75 % of the candlestick's lifetime; for an AVERAGE candlestick!).
  • #87
  • Jan 6, 2024 1:53pm Jan 6, 2024 1:53pm
Quoting AntiCre
{quote} As apprendista pointed out, I haven't mentioned candlestick patterns at all. And I was talking about 1000 ticks, i.e. the 1st candlestick is from now 999 ticks back. Then another tick comes in and I repeat that process. Why am I doing this? I'm interested to see how price moves from open to high/low, to low/high and finally to the close. This gives me an average, i.e. a number of what I can expect during a particular period. Why 1000? It's just an arbitrary number. And no, it doesn't matter to me whether this number is arbitrary or not....
Hi AntiCre

This data is still historical, you can't work out the low of the next candle till that candle is complete. If that next candle was actually a Bearish candle, where are you now?.
There are Quant firms using very sophisticated mathematics always trying to work out similar information you are also seeking, and they couldn't do that. That's the best minds in the game came out blank trying to work out systems based on previous history like you are trying to achieve. The profit results are just too low.

So what did they do?

Worked out that it's better to control future price by dominating the top of the order book where they sit in front of the Queue. If price starts moving down the Queue against them, they just sell out to the guys at the rear for no loss. This is how a few Quants firms have never had a losing day for years.

They worked out that risking their equity based on strategies that use historical data is just plain dumb when you can control the future order price AHEAD OF TIME.

Cheers
Trading thin liquidity at the boundary of the charts
  • #88
  • Edited 3:18pm Jan 6, 2024 2:47pm | Edited 3:18pm
Quoting RickM
Hi AntiCre This data is still historical, you can't work out the low of the next candle till that candle is complete. If that next candle was actually a Bearish candle, where are you now?
That's what I planned to talk about if anybody could get that stats. The idea is not to use past performance to predict next occurences, but to use it to measure where price is right now (in relation to these reference points). Yes, price moves according to orders. Ergo: Orders and price moves go hand in hand. Thus, if I don't have access to trustworthy order book/level 2 data, I could measure how price changes each and every moment, couldn't I (at least in theory)? And even if you'd have a trustworthy order feed, every second a new big order block could appear and price will change its direction.

Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 4 KB


This is for example how an average bearish candlestick looks like. In post #69 you can read that I've also surveyed how the bearish candlestick's appearances can change for different lengths of its body. It's similar to the wicks. If you fix the length of one wick, it will impact the wick of the other side AND the body.

Yes, with just looking at the open-close range you're doomed. Why? because you may see that the candlestick is still bearish. However, price could have retraced a lot from the low of the candlestick. Thus, incorporating a 2nd measure which can tell you whether you can trust your open-close range measure is needed. And you can find a good candidate for such a measure by knowing how the average candlestick behaves, i. e. when during the lifetime of an average bullish/bearish candlestick the highs and lows occur, and how these measures as well as the body/wick sizes changes if the high is hit before the low or vice versa. It's like a measure of the wick that's associated to the close. But it's not close - low when candles are bearish or high - close if candles are bullish! That way you'd have the problem that ryuryu pointed out in post #78 (see my reply in #80).

It's like looking at a trend. The trend (analogy to the open-close range) can be weak or strong. And there are retracements and pullbacks (analogy to the 2nd measure) in a trend that can be weak or strong, too. If price retraces too much during your trend, you'll be scared to lose (gained) money and you'll exit. It's the same with just 1 candlestick. Of course, you don't need to stick to just one candlestick, you can generalize the candlestick analogy, e.g. by using rolling candlesticks with a fixed tick number as suggested in post #69.

Sorry "red diamond" members. But if you'd have read what I wrote earlier, your replies should look different. Nevertheless, thanks for passing by! Everyone's input to the discussion is welcome!
  • #89
  • Jan 6, 2024 3:18pm Jan 6, 2024 3:18pm
Quoting AntiCre
{quote} That's what I planned to talk about if anybody could get that stats.
Why I love FF?
Because of that.

1. Guy asked for help creating new (never-ever-seen-brand-new-thread)
2. At least 2 real traders told him what to do
...
STILL SAME BULLSHIT

Am I tired of this shit? No. I'm happy. Because, because I'm trader
Observer effect
  • #90
  • Jan 6, 2024 3:21pm Jan 6, 2024 3:21pm
Quoting ryuryu
{quote} Why I love FF? Because of that. 1. Guy asked for help creating new (never-ever-seen-brand-new-thread) 2. At least 2 real traders told him what to do ... STILL SAME BULLSHIT Am I tired of this shit? No. I'm happy. Because, because I'm trader
That's not the point - and you know it. And no, I haven't asked for help to create something. Point me to that post! I already have the stats and was going to discuss it. That's why this thread is in the discussion section.
  • #91
  • Jan 7, 2024 2:25pm Jan 7, 2024 2:25pm
Quoting AntiCre
{quote} I think what he meant is that the high/low of a candlestick will be made with high probability short after the opening time and short before the close time of that candlestick. From my studies I know that in a bullish candlestick the low will be made before the high 88.5 % of the time. Here, the low is approximately made at the time point 0.191 x timeframe, while the high is made at 0.807 x TF on average. Only in 11.5 % of all bullish candlesticks the high is hit before the low. In these cases, the average times at which highs and low are...
Hi AntiCre, I follow your posts with interest as they often make me think, so please take the following thoughts as a brainstorm. After reading these data I thought they were too good to be true.. by combining these information, knowing that the price tends to close above/below the opening price 95% of the time, and using SL and TP based on time points data, it doesn't seem unrealistic to me to create a strategy with a good statistical edge, maybe I'm just too optimistic... For example, what happens if I decide to stay in the buy hemisphere, without knowing for sure what the outcome of the next candle will be, and setting TL and SL only based on the minutes that have passed since the candle opening, perhaps buying only when the open line is exceeded, wouldn't this already be a statistically promising edge? cheers
  • #92
  • Edited 12:52am Jan 8, 2024 12:32am | Edited 12:52am
Quoting Fiox89
{quote} Hi AntiCre, I follow your posts with interest as they often make me think, so please take the following thoughts as a brainstorm. After reading these data I thought they were too good to be true.. by combining these information, knowing that the price tends to close above/below the opening price 95% of the time, and using SL and TP based on time points data, it doesn't seem unrealistic to me to create a strategy with a good statistical edge, maybe I'm just too optimistic... For example, what happens if I decide to stay in the buy hemisphere,...
Dear Fiox89,

thanks a lot for your input! Yes, had the same idea. However, (and I hope I won't disappoint you) the brutal truth is that regardless of your entry criteria, you'll always encounter a 50/50 chance to get a target or take a loss first if you're using a 1:1 RR. If you like to, you can have a look at this thread where the thread starter collected some stats with the hope to predict a good entry point into a candlestick. I added some measures to his original stats and came up with these results for a 500-tick candlestick:

https://www.forexfactory.com/attachm...4?d=1519643945

One measure is called "open line cross". You can see that during that 500 ticks you often have a multitude of open line crosses, sometimes up to 60 or 70! I guess you know what that means. And unfortunately, even waiting until you've reached the average wick size until price crosses the open line again, won't help (see the stats collected here). In that post's image (2nd row) you can see that I've used a 1:1 RR, i.e. I was interested in getting X pips long while the price first retraced X pips. If I don't know in advance how the candle will end, I'd have had at best a 50 % win ratio (black points). However, if I'd have known in advance that the candle would end bullish (and I only traded long), I'd have had a a winning percentage of 60 - 65 % in a range for a retracement of 20 up to 40 pips first. The reason is, that the average candlestick I've looked at only has an average high-to-open range of 20 - 40 pips.

What can you learn from these findings? My conclusions are:

  1. Only trading once price crosses any price level (if candlesticks are arbitrary cut-off points in time, then you can create a candlestick anywhere) you're 50/50 best.
  2. The time horizont you're going to trade is very important for what you can expect.
  3. You won't find an edge in price alone.

Yes, 3. is really disappointing. However, you can use price to guide you. If you see a trend atm, you have a higher chance to make some pips long. However, this trend can end every moment. Thus, if you track the price development every tick, you maybe will identify with minimal lag when that trend ends and can exit. However, 1. will still prevent you to make enough money to end positive. Thus, where can the edge be found? To me it's to adjust to the current market situation (i.e. moving your SL/exit according to whether market is trending weak/strong, ranging, whether you're in a strong/weak retracement or pullback, etc.). But that's not a trivial task.

Just my 50 cents ...

  • #93
  • Jan 8, 2024 2:27am Jan 8, 2024 2:27am
Quoting AntiCre
{quote}...One measure is called "open line cross". You can see that during that 500 ticks you often have a multitude of open line crosses, sometimes up to 60 or 70! I guess you know what that means. And unfortunately, even waiting until you've reached the average wick size until price crosses the open line again, won't help (see the stats collected here). In that post's image (2nd row) you can see that I've used a 1:1 RR, i.e. I was interested in getting X pips long while...
Hi AntiCre,

I believe the solution is in your own words, and you are probably aware of it. Regardless of the attempts, the only way to have a probability greater than 50% is to be able to follow a trend, to have a bias. In this regard, an interesting post that you may already know here

You see a limit in point 1, however, this point seems irrelevant to me when you have decided on the trend (and this is where the real challenge lies). It doesn't matter how many times the price crosses the opening level, as long as it doesn't hit the SL, you go in your chosen direction once the reference point is hit.

Some questions that seem relevant to me (based on the assumption that your tests are run on random data)

a. These tests are typically performed on a sample of data, large enough to be considered representative. But we know well that in the long term the market is in equilibrium, and therefore my question: does it make sense to use this type of sample? Wouldn't it be more appropriate to use a specific sub-set?

b. We know very well that over any period the market may not have a trend. Assuming that we have a decent ability to define a trend (short as desired), can we achieve more than 50% chance in these circumstances? To be clear, I'm suggesting that the tests should be performed once we have our definition of trend, that is to say only in certain time intervals. How would the results vary? are these results repeatable, given a rule to define a trend?

c. Regarding the example from my previous post, I imagine you have considered that SL and TP are fixed, but is this necessary? In a rolling candle the information keep coming, and SL and TP could move accordingly, if once the expected time for SL and TP has passed and they have not been reached. What do you think, does this makes sense?


Cheers
  • #94
  • Jan 8, 2024 3:08am Jan 8, 2024 3:08am
Quoting Fiox89
{quote} Hi AntiCre, I believe the solution is in your own words, and you are probably aware of it. Regardless of the attempts, the only way to have a probability greater than 50% is to be able to follow a trend, to have a bias. In this regard, an interesting post that you may already know here You see a limit in point 1, however, this point seems irrelevant to me when you have decided on the trend (and this is where the real challenge lies). It doesn't matter how many times...
Just a brief reply since I'm busy right now with other stuff ...

a. I've done data on random walks and on EU and GU tick data. The average candlesticks perform similar, i.e. the wick-to-range and body-to-range ratios as well as the frequencies of bull/bear candles with high time>low time and vice versa are identical, as well as the average relative high/low times. Thus, I'd consider my sample size (for my current tests 30 weeks of tick data, i.e. > 27.2M data points) valid.
b. Yes, probably. Well, in trends there are as many bullish than bearish candelsticks (green vs. red), almost 50/50. However, on average red ones are smaller in up trends than the green candles. Nevertheless, you'll also encounter retracements in up trends that reverse the candles' strength. But in general, I think it's doable. Won't go into your specific questions yet.
c. Yes, you're right. But the way how you adjust it is tricky to come up with. And I'm still try to refine my decision tree.
  • #95
  • Jan 8, 2024 3:19am Jan 8, 2024 3:19am
Morning AntiCre

Honest question with no sarcasm or ill will intentioned: Do you want to trade and earn money or do academic analysis?

I ask because I think many struggle to break away from habits formed trying to find an edge, if I told you most days you may only manage one or two trades and then close your charts down maybe for a few days, what would happen?

I watch a runner lap and his average stride is 780 for that 800m lap. A friend also trades in the runners stride, he's trading using the heartbeat of the runner, he can predict the number of strides depending on how fast the heart is pumping, the higher it is the more strides he'll predict, etc.

Over time we compare notes, we're pretty close in results. Some days the runner trips but it's always pretty close. Then the race changes, it's now 400m, so my friend is now out buy 50% he needs to calibrate and uses some fancy math (/2) to make adjustments. I'm still pretty close.

The distance will always be made, even if it is cancelled and starts again the next day or the day after. Most days it's exceeded.

There are near certainties in the forex market, find them and then work out your money management before you trade.
  • #96
  • Jan 8, 2024 3:28am Jan 8, 2024 3:28am
Quoting Fagin
Morning AntiCre Honest question with no sarcasm or ill will intentioned: Do you want to trade and earn money or do academic analysis? I ask because I think many struggle to break away from habits formed trying to find an edge, if I told you most days you may only manage one or two trades and then close your charts down maybe for a few days, what would happen? I watch a runner lap and his average stride is 780 for that 800m lap. A friend also trades in the runners stride, he's trading using the heartbeat of the runner, he can predict the number of...
I'll stick to Einstein who said "Any fool can know. The point is to understand." Yes, the key imo lies in managing. But in a world with inifinite complexity, how do you know how do manage? You need to identify patterns. At least that's what I'm on ....
  • #97
  • Jan 8, 2024 3:48am Jan 8, 2024 3:48am
Quoting AntiCre
{quote} I'll stick to Einstein who said "Any fool can know. The point is to understand." Yes, the key imo lies in managing. But in a world with inifinite complexity, how do you know how do manage? You need to identify patterns. At least that's what I'm on ....
The only pattern here is of people sticking their face deep into their screen staring at candlesticks wondering why it all looks the same, coming to the same conclusions that you have that price goes up and down and more or less the same amount, and that it's all just a coin flip.

So maybe you aren't looking at the right stuff. And that's why there's nothing to discuss.
Congratulations ScalaFX for coming out of the closet!
  • #98
  • Jan 8, 2024 4:21am Jan 8, 2024 4:21am
Quoting nubcake
{quote} The only pattern here is of people sticking their face deep into their screen staring at candlesticks wondering why it all looks the same, coming to the same conclusions that you have that price goes up and down and more or less the same amount, and that it's all just a coin flip. So maybe you aren't looking at the right stuff. And that's why there's nothing to discuss.
Yeah, maybe. Or you don't see it because I haven't talked about it. So, what's the difference between your and this post (content-wise)? Nothing, I guess. So, let's start again and rephrase it:

"Hey Anti, well, I was on the same track like you years ago. And I found the same, but then I did XYZ and ZYX happened." That's how a discussion could have started.
  • #99
  • Jan 8, 2024 4:36am Jan 8, 2024 4:36am
Quoting AntiCre
{quote} Yeah, maybe. Or you don't see it because I haven't talked about it. So, what's the difference between your and this post (content-wise)? Nothing, I guess. So, let's start again and rephrase it: "Hey Anti, well, I was on the same track like you years ago. And I found the same, but then I did XYZ and ZYX happened." That's how a discussion could have started.
Nobody is going to hand you their homework. I know where you are wrong, but have no reason whatsoever to tell you where to look instead.

You are, after all, the one complaining that nobody is contributing... there's nothing to contribute. This topic isn't a winner, and what you want out of it isn't likely to happen. What you should take from this thread is that you are probably on the wrong path.
Congratulations ScalaFX for coming out of the closet!
  • #100
  • Jan 8, 2024 4:44am Jan 8, 2024 4:44am
Quoting nubcake
{quote} Nobody is going to hand you their homework. I know where you are wrong, but have no reason whatsoever to tell you where to look instead. You are, after all, the one complaining that nobody is contributing... there's nothing to contribute. This topic isn't a winner, and what you want out of it isn't likely to happen. What you should take from this thread is that you are probably on the wrong path.
Well, that's fine and I thank you for your thoughts!
0 traders viewing now