Market Statistics & Probabilities (requests welcome) | Page 6 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Odds and Edge - Probabilities in Day Trading 1 reply

Best book on randomness in markets and market probabilities? 4 replies

Trading and Probabilities 3 replies

enoc's psychology&probabili ties thread 23 replies

Trader Statistics and Broker Statistics 0 replies

Attachments: Market Statistics & Probabilities (requests welcome)
Exit Attachments

Market Statistics & Probabilities (requests welcome)

  • #101
  • Jan 11, 2011 8:52pm Jan 11, 2011 8:52pm
Can you also program a system that is almost well defined?

Can you test how much retracement after every major turning point from candles counting out fibo numbers; 3,5,8,13,21 etc.

Can you test when macd 10 min in direction of trade and 2 10 min consecutive candles in that direction approximate size of move ......

you know there is an avg. swing for every currency....
  • #102
  • Edited 9:34pm Jan 11, 2011 9:18pm | Edited 9:34pm
Quoting vx13 (quoting hanover)
[color=Blue][i]“An obligatory condition should be the binding stops to a technically significant level, that is, to support, resistance, or a trendline. Stops should be placed only where the probability of continuation of a movement in the direction against your position grows sharply.......

As far as I know, lots of traders usually place stops considering money management reasons only. ........ Technical levels such as trend lines, supports, or resistances (frequently being close at a short distance behind the stops) are not taken...
In wrestling, there becomes a point of desperation, where to prevent the onslaught from continuing, the losing wrestler must tap out. They have admitted that they no longer have the intention to stay in the ring.

Think about how this 'obligatoy condition' of placing your stops behind a technically significant level, divulges your point of desperation at which you can no longer bear the drawdown. Placing it behind a technically significant level, along with individuals who also sumbit to this modus operendi, are showing their hands to the worst possible people - professional traders. You are putting yourself in a compromising position every time, and as price nears your 'well-placed stop', it's only a matter of time before your position is chomped up to provide liquidity for theirs...1, 2, 3.

Sort of reminds me of a clustered elevator with too many people..when the cable snaps, boy oh boy does it drop.

PS. While I do find the number-crunching and statistical analysis interesting, I just hope that the intention of your research project is not to develop a system based on trying X amounts of pips for a stop-loss, and using the one which gives the most profitable results over backtested data. Beware of curve-fitting..it may come back to bite you.

Regards,
xXTrizzleXx
  • #103
  • Jan 12, 2011 3:51pm Jan 12, 2011 3:51pm
Quoting xXTrizzleXx
In wrestling, there becomes a point of desperation, where to prevent the onslaught from continuing, the losing wrestler must tap out. They have admitted that they no longer have the intention to stay in the ring.

Think about how this 'obligatoy condition' of placing your stops behind a technically significant level, divulges your point of desperation at which you can no longer bear the drawdown. Placing it behind a technically significant level, along with individuals who also sumbit to this modus operendi, are showing...
Agreed.
But what I was taking from the quote was that a stop should not be arbitrary in relation to significant price levels. It shouldn't be placed just because it's "20 pips from my entry".
Placing it by the criteria of money management, giving your trade room to breathe, and significant price levels is the best way to get the most of your stop.

I've read what you've stated in a few books and I wholeheartedly agree.
I keep the same in mind for my own personal stops but I was just stating that kind of stoploss of the statistic. I'm so it could be better designed of course. Mine at least.

@newyear
I would like to see your design for the psych level test, newyear
And I dropped you a PM
  • #104
  • Jan 13, 2011 12:05pm Jan 13, 2011 12:05pm
This is a touchy subject.. at least I perceive it to be

some people might be stuck to there belief and insult me for going against there belief..

like when someone challenges something from the bible the bible lovers are going to flip shit

so what I'm saying is as the results are adding up.. if some find them less then impressive those religious believers in the gap will never accept it..


OK RESULTS ARE IN!

Will the weekend gap fade?

Yes it will fade = 53.97% of the time

say what you want.. but this is done with 1:1 risk to reward.. so if the weekend gap was 50 pips.. will it fade the 50 pips.. or go the opposite direction 50 pips..

Well, there it is
  • #105
  • Jan 13, 2011 5:56pm Jan 13, 2011 5:56pm
AND the bible is wrong.

Discuss!
  • #106
  • Jan 13, 2011 6:09pm Jan 13, 2011 6:09pm
Quoting Marv
AND the bible is wrong.

Discuss!
I believe....... that it's a sin to discuss religion on this forum.

May the FF gods have mercy on your soul.
  • #107
  • Jan 13, 2011 6:17pm Jan 13, 2011 6:17pm
My point being these are just my results.. please don't crucify me

lol
  • #108
  • Jan 13, 2011 7:40pm Jan 13, 2011 7:40pm
I just ran a backtest of 10 years data on Standard Deviation

The results show a small statistical advantage.
Considering the upper and lower SD provided similar results it is convincing enough (for me at least) that there is an edge to be gained in the long run...

Here are the stats.. with 1:1 risk to reward,
if price goes to +1 SD or -1 SD.. you immediately place a trade in direction of the average with TP at the average and SL at +2 or -2 SD

chance you will get your TP = 54.23%

its a small advantage but its actually a pretty solid test result in all honesty.. and it surly could be filtered for even better results I would put my money on

anyways these are the first test results I have that show me some proof in the pudding of an entry exit point with better than 50% probability.. and good risk to reward

Well it might mean nothing to you but these are interesting to me
  • #109
  • Jan 14, 2011 1:53am Jan 14, 2011 1:53am
Quoting newyear498
I just ran a backtest of 10 years data on Standard Deviation

The results show a small statistical advantage.
Considering the upper and lower SD provided similar results it is convincing enough (for me at least) that there is an edge to be gained in the long run...

Here are the stats.. with 1:1 risk to reward,
if price goes to +1 SD or -1 SD.. you immediately place a trade in direction of the average with TP at the average and SL at +2 or -2 SD

chance you will get your TP = 54.23%

its a small advantage but its actually a pretty solid test...
Interesting result. I'm a bit confused by "standard deviation" though, can you maybe explain a little more practical? Can "standard deviation" be a set number of pips?
  • #110
  • Jan 14, 2011 3:57pm Jan 14, 2011 3:57pm
Quoting pasopvirpot
Interesting result. I'm a bit confused by "standard deviation" though, can you maybe explain a little more practical? Can "standard deviation" be a set number of pips?
Bollinger bands are a measure of standard deviation, set your BB to something like 14 period measure with Standard Deviation of 1.. could do a 30M chart

that would be a practical example..



in 2010 this method was true as well but with 52% stats

I am concerned it might be luck that there is a small positive outcome.. but the law of large numbers as well as every year seeming to favor a very slight positive outcome have my attention
  • #111
  • Jan 14, 2011 7:48pm Jan 14, 2011 7:48pm
What is the % for a long & short order to be opened @ 00:00 GMT and go 75 pips for take profit & 50 pips stop loss for last 260 days?
  • #112
  • Jan 19, 2011 12:16pm Jan 19, 2011 12:16pm
Quoting newyear498
I just ran a backtest of 10 years data on Standard Deviation

The results show a small statistical advantage.
Considering the upper and lower SD provided similar results it is convincing enough (for me at least) that there is an edge to be gained in the long run...

Here are the stats.. with 1:1 risk to reward,
if price goes to +1 SD or -1 SD.. you immediately place a trade in direction of the average with TP at the average and SL at +2 or -2 SD

chance you will get your TP = 54.23%

its a small advantage but its actually a pretty solid test...
I understand that one standard deviation away from the mean covers 68% of the data of a normal distribution curve so one deviation up is 34% and one down is 34% but where is the mean in this test?

Is it at the mean of every bar or PoC or a moving average?
Very interesting stuff
  • #113
  • Jan 19, 2011 12:54pm Jan 19, 2011 12:54pm
Quoting vx13
I understand that one standard deviation away from the mean covers 68% of the data of a normal distribution curve so one deviation up is 34% and one down is 34% but where is the mean in this test?

Is it at the mean of every bar or PoC or a moving average?
Very interesting stuff
on this test it was a moving average

I am personally still turn off by the positive results because its a hell of a long haul with this statistics

in other word draw down can large since the consecutive loses eventually add up big.. but in the long run yes things do turn positive..

In other words I am still favoring 50% probabilities and exploiting that.. this results are certainly in mind for some other ideas

thanks for your response
  • #114
  • Jan 19, 2011 1:39pm Jan 19, 2011 1:39pm
If price breaks the 38.6% fib how many times it touches 50% fib,
If price breaks the 50% fib how many times it touches 55% fib,
If price breaks the 50% fib how many times it touches 61.8% fib,

by break i mean,... Price "closing" above the fib level. I'd like to see the tests on 1H,4H and Daily TF.

Thanks
.
  • #115
  • Jan 19, 2011 5:42pm Jan 19, 2011 5:42pm
Quoting nasir.khan
If price breaks the 38.6% fib how many times it touches 50% fib,
If price breaks the 50% fib how many times it touches 55% fib,
If price breaks the 50% fib how many times it touches 61.8% fib,

by break i mean,... Price "closing" above the fib level. I'd like to see the tests on 1H,4H and Daily TF.

Thanks
.
fibs are too subjective for me to test.. I could draw a fib a million ways.. just like someone can draw a trend line a million ways..
  • #116
  • Jan 22, 2011 12:10am Jan 22, 2011 12:10am
Its Friday and just after London close

the week was up.. does the EUR/USD retrace in direction before forex closes for the weekend? Will everyone take their profits before the weekend?

NO! more often than not the last few hours of Friday will continue in the direction of the weeks trend..

Friday after London closes: (10 years data)

Continues in weeks direction = 56.10%

Retraces direction = 43.90%
  • #117
  • Dec 17, 2011 10:37pm Dec 17, 2011 10:37pm
are you still running probabilities and statistics?
AVT INVENIAM VIAM AVT FACIAM
  • #118
  • Last Post: May 10, 2015 6:51pm May 10, 2015 6:51pm
Hi all,

I know this is an old thread but I want to give it a shot anyway. I have a statistics related question. I am really no expert in statistics but I do have some basic knowledge.

Now, what am I trying to do? I want to simulate an R-multiple distribution in excel. It seems like this can be best represented by an F-distribution. Is this correct?

Assuming the answer to this question is yes then the next question I have is how to decide on the properties of my R-multiple distribution. I would like the mean of the F- distribution to be the expectancy of my system PLUS ONE. I think (expectancy + 1 = d2 / (d2 -2)) is the formule. Can I use this ?

In addition I would like to have a certain % of values to be below 1.

So I want to figure out a way to randomly generate an R-multiple distribution Generating an F-distribution is not hard per se, but I want the F distribution to be representative for my system. So i need to decide on the two values for the degrees of freedom.

anyway, it's probably not the clearest explanation but I hope somebody understands what I am trying to do and wants to give me a hand !

Thx in Advance
0 traders viewing now