Menu
Sister Sites

Options

Bookmark Thread

First Page First Unread Last Page Last Post

Print Thread

Similar Threads

Your and My Lessons Learned Losing and Gaining in Trading 5 replies

things you learned from forex and improved your life 12 replies

The most important lesson’s I’ve learned trading Forex, so far... 15 replies

Attachments: What I learned after losing $300,000 trading forex
Exit Attachments

What I learned after losing $300,000 trading forex

Quoting DrDave
{quote} FINALLY, I see someone on FF besides me that understands that R:R is a highly overrated trade management technique!
Thanks..very true
The impossible is possible
Quoting ursinho4711
{quote} Very good summary and conclusion. I agree, except for: {quote} I use some indicators that help me to get a picture of the situation. Indicators that work on different time frames and that I understand! I do not follow any "arrow up=buy" or "sell alert!". But it took me a loooooong time to find the indicators that suits my trading style. Same with expert advisors: I do not let my advisors run totally without control (anymore). Not anymore, because of some situation that suddenly appeared and blew my account. Not only once... I then started...
As long as it works for you. Indicators are not bad it's just not my thing. I just prefer price action because of demand and supply laws and all that.
The impossible is possible
Quoting tonny47
{quote} there are successful traders/investors using Kelly. what would you say is better way to manage risk? arbitrary %?

I blocked this guy. He clearly is just a Numpty. Let me be perfectly clear. All institutions use the kelly criterion. It is so important it is programmed directly into their platforms.
You are doing well but please keep in mind the levels are not support and resistance. Trying to use them as such might work for a while but eventually, you will get burned badly.

The levels are just a frame to give you the context of "possibilities" vs "probabilities".
Look at this H1 chart that fallinforex posted.
This should be triggering in your mind what you can do to develop a strategy.

Look at a day on that chart.
Check to see if price closed above 0.25%. If yes, did it reach 0.5% sometime after that and before the close of the day?
Make a count of the number of days that the H1 closes a bar above 0.25% and also a count for the days where price also reached 0.5% before the day closed.

After you have at least 100 days, you will have a reasonable indication of the probability of a winning trade if you simply put a pending order slightly above 0.25% with TP just below 0.5%. Note this does not look at drawdown or SL placement at this stage.

So, you would be wise to also do a similar study to determine the probability of price going a certain amount against you after the H1 closed above 0.25% BEFORE it reaches TP so you can have statistics help you place a SL and to give some statistics about how deep the drawdowns go.

Attached Image (click to enlarge)
Click to Enlarge Name: Screenshot1.png Size: 113 KB
To free Gazans of Hamas, flatten it--if that is what it takes.
Quoting CeeSo
You are doing well but please keep in mind the levels are not support and resistance. Trying to use them as such might work for a while but eventually, you will get burned badly. The levels are just a frame to give you the context of "possibilities" vs "probabilities".
Very true. I absolutely don't rely on support and resistance only. And I use very close stop loss. Plus it's important to understand the whole market. For example USD has been dropping...good profit if you bought xauusd yesterday. Canadian dollar has been doing well also for a while now. Plus I use other technical analysis which is really what my strategy is built around. The rest is just understanding the market to better my chances.

Trying to learn Gann theory. He was a genius. I think that would even further improve anyone's chances.
The impossible is possible
Quoting Teerexteerex
{quote} Very true. I absolutely don't rely on support and resistance only. And I use very close stop loss. Plus it's important to understand the whole market. For example USD has been dropping...good profit if you bought xauusd yesterday. Canadian dollar has been doing well also for a while now. Plus I use other technical analysis which is really what my strategy is built around. The rest is just understanding the market to better my chances. Trying to learn Gann theory. He was a genius. I think that would even further improve anyone's chances....
Be sure that you understand how to construct the square of 9 and how to apply values to a chart!

Also, check on the instrument that you trade the number of pips in 45 degrees of movement, then divide by 4 and get the typical swing length before a correction is seen.

Also, do not get confused by "squaring price and time" as meaning to mathematically calculate the square. Many people try to do that. It has nothing to do with such a calculation.
To free Gazans of Hamas, flatten it--if that is what it takes.
Quoting DrDave
{quote} Be sure that you understand how to construct the square of 9 and how to apply values to a chart! Also, check on the instrument that you trade the number of pips in 45 degrees of movement, then divide by 4 and get the typical swing length before a correction is seen. Also, do not get confused by "squaring price and time" as meaning to mathematically calculate the square. Many people try to do that. It has nothing to do with such a calculation.
Wow thanks man...Will keep that in mind. I haven't really started yet.
The impossible is possible
Hi,

Interesting thread, thank you for starting.

I am sure it is something I am looking at incorrectly, but from what I see, the data for the E/U is only 80.63% within 0.50% to -.50%, and the data for A/U is only 65.85% within 0.50% to - .50% levels (open to close). Please see your spread sheets attached, specifically the green boxes I added. What am I interpreting wrong, no way this has been missed this whole thread, has it?

Cheers.
Attached Image(s) (click to enlarge)
Click to Enlarge Name: EUR.USD.jpg Size: 178 KB
Click to Enlarge Name: AUD.USD.jpg Size: 180 KB
This is what I got. Also did some smaller sample sizes (over a couple random months) and some went pretty low - in the 65% area. As far as I can tell, the statistics don't show 85%, over multiple years seems like 77% would be the more accurate figure, or am I missing something? Cheers
Attached Image
What am I doing wrong?
Attached Image
Quoting Phil_GMT
$300K Loss I couldn't do it, I'd probably never trade again. Then again you must have deep pockets to lose that much. I'm pleased you are turning it around. Point 2, Agreed Demo trading successfully for 5 months is a real test, I wish I only demo traded until I was successful. In day trading X5 your account isn't anything special, so many people x2-x4 their demo and start Live trading, but newbies don't know that's not enough. Once you have X50-X100 on Demo you may have the skills to do it live. Also any new setups or ideas need to be executed on...
dont give up
Quoting SunSaltwater
This is what I got. Also did some smaller sample sizes (over a couple random months) and some went pretty low - in the 65% area. As far as I can tell, the statistics don't show 85%, over multiple years seems like 77% would be the more accurate figure, or am I missing something? Cheers {image}
Well, when you take a small sample size, the smaller you go, the larger the expected deviations can be. I think CeeSo did mention the levy flights, where your distribution of data can have small bouts of large movement here and there. However over a large enough span, they all normalize.

Hence, your EURUSD data for 0.5 to -0.5 shows 77% probability over 2016-2020. It might not be 85%, but debating whether it's 80% or 85% or 77% seems to be missing the reason Ceeso brings up the percentages, in my opinion.
Quoting Brandon65
{quote} Well, when you take a small sample size, the smaller you go, the larger the expected deviations can be.
Completely wrong. You cannot know what the deviation/dispersion is in advance. Similarly, when taking samples, the expectation is that the variance is the same as the population variance, within the limits. Go study Analysis of Variance (ANOVA) and how it is used in a variety of contexts.

You should also study Analysis of means. But ANOVA tends to be more robust because the distribution curve of Variances of the samples from the same population is entirely different from the distribution curve of the means of the samples.
Quoting Brandon65
I think CeeSo did mention the levy flights, where your distribution of data can have small bouts of large movement here and there. However over a large enough span, they all normalize.
No, they do not normalize. You mean something else.
Quoting Brandon65
Hence, your EURUSD data for 0.5 to -0.5 shows 77% probability over 2016-2020. It might not be 85%, but debating whether it's 80% or 85% or 77% seems to be missing the reason Ceeso brings up the percentages, in my opinion.
I posted earlier that after you collect your larger dataset that it would be useful to split it into groups and run statistical tests to determine if any group does not belong. Likewise, as more data are collected, you can take this new data and test statistically if it belongs to the prior data. If not, that is a signal that a significant change has occurred and your model may no longer be valid.
To free Gazans of Hamas, flatten it--if that is what it takes.
Quoting SunSaltwater
This is what I got. Also did some smaller sample sizes (over a couple random months) and some went pretty low - in the 65% area. As far as I can tell, the statistics don't show 85%, over multiple years seems like 77% would be the more accurate figure, or am I missing something? Cheers {image}
I think your numbers are correct, but you're only verifying if the price closes within +0.5 -0.5%
now if you check if the price went above +0.5 then closed under +0.5, then the stats are not the same. I got around 30% hitrate then. Meaning you would be better off trading a breakout than a reversal. I think this whole thread is based on bogus statistics

here is my code:
Inserted Code
double top = o0 + percentage * o0 / 100.0;
double bot = o0 - percentage * o0 / 100.0;
        
if ((c0 <= top && high0 > top) || (c0 >= bot && low0 < bot))
{
  proba++;
  trades++;
}
else
  trades++;
 
Print("proba closes between top and bot = ",proba*100/double(trades));

and if you use smaller numbers than 0.5 (like 0.1), then the hitrate goes up obvisouly (around 70%), but profit is then smaller (too small)...so that's, not a solution either
Quoting ionone
{quote} I think your numbers are correct, but you're only veryfying is the price closes within +0.5 -0.5% now if you check if the price went above +0.5 then closed under +0.5, then the stats are not the same. I got around 30% hitrate then. Meaning you would be better off trading a breakout than a reversal. I think this whole thread is based on bogus statistics here is my code: double top = o0 + percentage * o0 / 100.0; double bot = o0 - percentage * o0 / 100.0; if ((c0 <= top && high0 > top) || (c0 >= bot && low0 < bot)) { proba++; trades++;...
You think wrong. The statistics are not bogus. However, how they are being (mis)used in most of the postings and by many people that clearly know very little about the subject, which can certainly make it appear that the statistics are a problem when in reality, it is the presenter that is the problem.
To free Gazans of Hamas, flatten it--if that is what it takes.
Quoting ionone
{quote} I think your numbers are correct, but you're only veryfying is the price closes within +0.5 -0.5% now if you check if the price went above +0.5 then closed under +0.5, then the stats are not the same. I got around 30% hitrate then. Meaning you would be better off trading a breakout than a reversal. I think this whole thread is based on bogus statistics here is my code: double top = o0 + percentage * o0 / 100.0; double bot = o0 - percentage * o0 / 100.0; if ((c0 <= top && high0 > top) || (c0 >= bot && low0 < bot)) { proba++; trades++; } else...
Maybe you should read, or re-read it if that applies, this post and try doing a study similar to what is outlined there.
To free Gazans of Hamas, flatten it--if that is what it takes.
Quoting ionone
{quote} I think your numbers are correct, but you're only veryfying is the price closes within +0.5 -0.5% now if you check if the price went above +0.5 then closed under +0.5, then the stats are not the same. I got around 30% hitrate then. Meaning you would be better off trading a breakout than a reversal. I think this whole thread is based on bogus statistics here is my code: double top = o0 + percentage * o0 / 100.0; double bot = o0 - percentage * o0 / 100.0; if ((c0 <= top && high0 > top) || (c0 >= bot && low0 < bot)) { proba++; trades++; } else...
Thank you for your post. Yes, I was not asking how one would trade this, I merely wanted to confirm that my stats are correct.

CeeSo, any particular reason you say 85% when the stats from many different samples sizes indicate that it's closer to the 75% area over longer term stats to close within +.5 to -.5? From what I can see, you posted two spreadsheets showing 81% and 66%, but continually speak about 85%. Not trying to create a debate, just curious why you don't say 'somewhere between 65 - 85%', or 'in the 70's on average', or something similar. Again, I am open to the possibility that I could be looking at something differently to you.

Cheers.
Quoting DrDave
{quote} You think wrong. The statistics are not bogus. However, how they are being (mis)used in most of the postings and by many people that clearly know very little about the subject, which can certainly make it appear that the statistics are a problem when in reality, it is the presenter that is the problem.
well the "85%" stat is based on the assumption that 85% of the time price will close between +0.5 and -0.5
but if price doesn't cross +0.5 or -0.5 then what good does it make ? It means price will simply stay between -0.5 and 0.5 most of the time but you don't know how close from 0.5 or -0.5, how is it useful? Maybe it will stay around -0.4 for 85 out of 100 trades. then the "85%" stat is correct but how do you trade that successfully ?
Quoting SunSaltwater
{quote} Thank you for your post. Yes, I was not asking how one would trade this, I merely wanted to confirm that my stats are correct. CeeSo, any particular reason you say 85% when the stats from many different samples sizes indicate that it's closer to the 75% area over longer term stats to close within +.5 to -.5? From what I can see, you posted two spreadsheets showing 81% and 66%, but continually speak about 85%. Not trying to create a debate, just curious why you don't say 'somewhere between 65 - 85%', or 'in the 70's on average', or something...

I also get 80% over 365 bars.

Attached Image
1 trader viewing now