Options

Search
Search
Search

Subscribe to Thread

Bookmark Thread

First Page First Unread Last Page Last Post

Printable Version

Similar Threads

Changing an indicators colour depending on its value . . . 0 replies

Colour ADX Indicator depending on value 0 replies

EA wanted for sound alert depending on fractals condition 2 replies

Probability depending on distance

  • Post# 1
  • Quote
  • First Post: Jan 2, 2013 9:58am | Edited at 10:48am
Hi all..
I'm going to make calculated proof to myself that smaller time frames have worse results (because of spread). --> maybe they are 1% worse, and not 40% worse...
So... Let's assume that spread does not exist (for simplifying the calculation)...
For example, the distance to stop loss is 10 pips, and the distance to take profit is 20 pips.
How would we calculate the probability of moving price to SL or to TP, assuming that price movement is completely chaotic and touch of any of them stops the cycle?
Thanks.
Quoting Gravitus
Hi all..
I'm going to make calculated proof to myself that smaller time frames have worse results (because of spread). --> maybe they are 1% worse, and not 40% worse...
So... Let's assume that spread does not exist (for simplifying the calculation)...
For example, the distance to stop loss is 10 pips, and the distance to take profit is 20 pips.
How would we calculate the probability of moving price to SL or to TP, assuming that price movement is completely chaotic and touch of any of them stops the cycle?
Thanks.
not sure what the difficulty is here, math101. Assuming a normal distribution you can calculate it the following way:
1-(10+20)/20 = 33% that take profit is hit, 1-(10+20)/10 = 66% that stop loss is hit.

however, the premise of a normal distribution / chaotic price movement is already flawed, that's why all these calculations won't help much.
thanks, but according to that i can make some calculations that will certainly show me some difference between 1 pip spread and 5 pips spread for this 10/20 SL/TP example ..
Quoting Gravitus
thanks, but according to that i can make some calculations that will certainly show me some difference between 1 pip spread and 5 pips spread for this 10/20 SL/TP example ..
The result is the same for all time frame. The reason is SL/TP depends on price fluctuation. If you are based on time frame to determine the probability or generate some statistical data, there would be meaningless if want to determine which time frame have the worst result (To fix SL/TP for hypothesis test).

If based on certain pattern to judge the probability from time frame. Of course higher time frame provide high winning rate because the range is bigger. Example: Support resistance area/range is bigger in higher time frame.

Of course if spread is bigger, the probability of winning is lower.
**Keep Simple**
FxGrowth All Time Return: -1.9%
Quoting ~bull.bear~
Of course if spread is bigger, the probability of winning is lower.
So far, the above mentioned calculations are the best way i know in order to turn your sentence in figures So, smaller time frames give more signals, but they should be filtered depending of strength of signals with adjusting to spread affect.
Quoting Gravitus
So far, the above mentioned calculations are the best way i know in order to turn your sentence in figures So, smaller time frames give more signals, but they should be filtered depending of strength of signals with adjusting to spread affect.

Not sure how this to be done. But even there is some kind of calculation, probability still need to depend on underlying market for the price to strike SL/TP.

Please note it is impossible to have perfect value or indicator to depict current market probability. Because probability is still based on the calculation from historical data. Unless base on certain pattern that market always repeats, example ascending triangle, head and shoulder, symmetrical triangle and etc. Below link is the statistical proven for the pattern's probability.
http://www.autochartiststocks.com/pe...ce-statistics/
**Keep Simple**
FxGrowth All Time Return: -1.9%
I'm not really sure what you are asking here.
If you trade with a 10 pip SL and 20 pip TP, it is simple to work out what % of your trades would need to hit TP for you to break even.
I don't see how you could work out mathmatically, the probability of a trade hitting TP though. The factors that move markets are not purely mathmatical. If they were, it would be impossible for anyone to make a profit, everybody would lose because of the spread.
When you also consider that most traders enter after confirmation that the market is moving in the desired direction, how do you factor in that part of a move has already happened?
Eg. If market has moved 10 pips up from point A and you enter a long position at B (A+10) with SL at A and TP at B+20, you actually require a 30 pip move in total from A to hit TP.
Too complicated for me to try and work out.
In reality, for forex this "distance in pips" where the TP and SL are is the ONLY way to calculate probability... perhaps not from a pure mathematical theory but in the practical.

If we do not consider the spread:
If we place a trade 1 pip tp and 1 pip sl the probability is 50/50
If we place a trade 10 pips tp and 10 pips sl the probability is 50/50
If we place a 1 pip TP and 10 sl the probability is 90/10
If we place a 10 pip tp and a 1 SL it is 10/90
...and like this with any combination of TP and SL


Now, by default, any combinations of TP and SL will tend to be a Break Even system
or 50/50 in our account balance. .. Meaning it does not matter if you have a system with
a 10% prob win and 90% prob loss or vice-versa ..,


BUT breakeven LESS spreads ... so all those combination I mention above are loser to our accounts
becuase they have shorts stops. That is why scalpers system are losers in the long run... and that is why any system
use in a higher time frame will always do better.... but not because the system is good or the indicator
used work better in higher time frames...it is because the stops are longer from the entry.

It is funny sometimes we read about systems with "high probability" wining trades .. sure, perhaps it is true that has a 90% probability of wining because the TP is 10 pips and the SL is 100 ... but that would not help my account balance.

Criteria does not change the probability. Market behavior does not change the probability...we chose the probability
by placing the stops.. but at the end, It is what the trader do with the given 50/50 by default that does increase or decrease the account. So.. stop using short stops... over 40 pips is the minimal were the spreads start to be absorbed

If every trader increase the stops in whatever system there are using, they will see and immediate improvement.
(more profits .. or less loss)

J.
.
Quoting Custos
not sure what the difficulty is here, math101. Assuming a normal distribution you can calculate it the following way:
1-(10+20)/20 = 33% that take profit is hit, 1-(10+20)/10 = 66% that stop loss is hit.

however, the premise of a normal distribution / chaotic price movement is already flawed, that's why all these calculations won't help much.
Just a minor clarification of the formula:
The formula above should be for a 10 pip sl, 20 pip tp, probability that take profit of 20 is hit: 1-20/(10+20) = 33%, and for stop loss of 10 1-10/(10+20) = 67%.

Quote
If we place a 1 pip TP and 10 sl the probability is 90/10
Not trying to bust chops, but just for clarification for those who come after attempting to work the exact numbers for their situation, the probability of 10 pip SL being hit first is 1/11 = 0.0909... rather than the 10% implied above for a 1 pip take profit, 10 pip stop loss with no friction.

Quote
It is funny sometimes we read about systems with "high probability" wining trades .. sure, perhaps it is true that has a 90% probability of wining because the TP is 10 pips and the SL is 100 ... but that would not help my account balance.
10 pip take profit, 100 pip stop loss works out more precisely to: 1-10/110 = 0.909
Quoting FXEZ
Just a minor clarification of the formula:
The formula above should be for a 10 pip sl, 20 pip tp, probability that take profit of 20 is hit: 1-20/(10+20) = 33%, and for stop loss of 10 1-10/(10+20) = 67%.
ah yeah, sorry, mixed up the nominator and denominator when posting the formula. thanks
.
Hi FXEZ... no chops busted.
Yea.. you have the correct math... I was going after the concepts.

a) The Probability of each trade is determinant by the distance in pips of
the TP and SL. (using FXEZ math) Nothing to do with criteria/discretion nor market behaviors.


b) Because of the spreads, using short stops greatly affects the possibility of making money,
regardless of the probability rate used.

J.
if you are wonderig why to trade higher tff over lower one its just a difference of time. Spread plays a big part but say on eurusd you might get 0.2 pip so should be beneficial to play lower tf here.

Again think about it.. even the time i spend writing this is priceless. The time it takes you to turn a prodit and buy your wife a trip to africa can be the difference
between a safari vacation or a dine.. Sleep and rest type vacation.. Depends which type you like better Ill let you to tell netween the two.

So take your spread and add it to your sl then figure out your chances
to win. Now calculate your win rate. This should give you everythingto calculate expectancy. Any expectancy greater than 1 is above breakeven.
Quoting jeuro
.
Hi FXEZ... no chops busted.
Yea.. you have the correct math... I was going after the concepts.

a) The Probability of each trade is determinant by the distance in pips of
the TP and SL. (using FXEZ math) Nothing to do with criteria/discretion nor market behaviors.
not sure if that is true. The probability calculations change when a normal distribution is NOT assumed. And several papers have pointed out that the markets are not normally distributed, thus you are required to calculate probabilities differently. If the markets were chaotic and the probability of 10 pip stop and 10 pip tp getting hit were exactly 50%, then the spread would make any system into a negative expectancy (unprofitable) system.

Quoting jeuro
b) Because of the spreads, using short stops greatly affects the possibility of making money,
regardless of the probability rate used.
J.
yeah, the spread does have an enormous effect, when stop loss and tp are very small.
Quoting ~bull.bear~
Not sure how this to be done. But even there is some kind of calculation, probability still need to depend on underlying market for the price to strike SL/TP.
It's not a forecasting method, it's just an adjustment of how spread affects the probabilities of results of a trading system.
(3)---------20=====> (3)-(1)=20



(1)---------
(2)---------spread

(4)---------10=====> (1)-(4)=10

-----------------------------
so, we buy at price (1) while bid is (2)
in this case for TP we'll have this probability:

1-(20+spread)/(30+spread-spread)

if spread is 1 pip:
Prob(TP) = 1-21/30 = 30%, then Prob(sl) = 70%

if spread is 4 pip:
Prob(TP) = 1-24/30 = 20%, then Prob(sl) = 80%

So, it says how spread deteriorates our results for these values of sl, tp and spread. That's what i wanted to say.
We can come to conclusion that model of comission (3pip) + spread (1 pip) is better for a trader than just spread (4 pip), even if they are equal, because comission doesn't affect the probabilities of SL/TP
Quoting Gravitus
It's not a forecasting method, it's just an adjustment of how spread affects the probabilities of results of a trading system.
..........
So, it says how spread deteriorates our results for these values of sl, tp and spread. That's what i wanted to say.
We can come to conclusion that model of comission (3pip) + spread (1 pip) is better for a trader than just spread (4 pip), even if they are equal, because comission doesn't affect the probabilities of SL/TP
OK. I got your point, you are looking for some kind of opportunity value to enter position at current movement. Example: Enter position at 1 pips spread is better than 3 pips spread for the same strategy, obviously this could reduce the spread risk. So you could have better probability of winning and less losing.

So is there any coding for this? Not sure have such code to determine spread value and open new position. One thing, sometime price reacts faster due to high frequency of trading, so does spread, but i think it is hard for the rule/EA to run this even we could code this scenario.
**Keep Simple**
FxGrowth All Time Return: -1.9%
Quoting ~bull.bear~

So is there any coding for this? Not sure have such code to determine spread value and open new position. One thing, sometime price reacts faster due to high frequency of trading, so does spread, but i think it is hard for the rule/EA to run this even we could code this scenario.
mql4 allows to do that, plus it's possible to use "kinda" average spread for current time of day/currency pair
Probabilities for touching a defined SL or TP...

There is a hidden component here. This component is time... Indeed, at any given moment, there is a probability for going up by xx pips and another one for going down by xx pips. Most of the time, these probabilities won't be the same for bear or bull move... For example, in an uptrending market, the probability of going up by let's say 5 pips in the next move is higher than that of going down by 5 pips.

So in definitive, the probability of touching a given point (TP or SL) is not a single (nor simple) calculation. It is a sum of probabilities.

Here is another detail that will make things even worse : the probability of going up by x pips in the next move is not the same as the probability of going up by the same amount of pips at another time (even just a few seconds/minutes from the first calculation).

Because quantifiable moves on low tf most often are smaller than the spread, trading such kind of principle can prove disastrous.

No need to say that whoever comes with a nice formula for such a problem could get nice profits really fast
Quoting Gravitus
mql4 allows to do that, plus it's possible to use "kinda" average spread for current time of day/currency pair

Nice! Sound you have a plan. Looking forward to your code and test in LIVE.
**Keep Simple**
FxGrowth All Time Return: -1.9%
Quoting Custos
If the markets were chaotic and the probability of 10 pip stop and 10 pip tp getting hit were exactly 50%, then the spread would make any system into a negative expectancy (unprofitable) system. yeah, the spread does have an enormous effect, when stop loss and tp are very small.
- Markets are chaotic but this doesn't mean completely random

- If the probability of 10 pip stop and 10 pip tp getting hit were exactly 50%, that would be... great! The market would be a 1 dimension (price can only go up or down) isotropic random walk. The price would stay in the vicinity of the first price => mean reversion strategies would make wonder.

Unfortunately that's not the case.

Bayesian probability may be useful here. Kind of "What is the probability of the price going 10 pips down before 10 pips up when I'm sure at 80% that it is on a support in a market that is 75% chance being in a uptrend...?"

Regarding the original question from the OP, you may have a look here http://www.forexfactory.com/showthread.php?t=336715
Quoting PipMeUp
- Markets are chaotic but this doesn't mean completely random
okay, maybe wrong word, I meant random.

Quoting PipMeUp
- If the probability of 10 pip stop and 10 pip tp getting hit were exactly 50%, that would be... great! The market would be a 1 dimension (price can only go up or down) isotropic random walk. The price would stay in the vicinity of the first price => mean reversion strategies would make wonder.
okay, random markets anyways don't exist, but just for the sake of discussion: I think your premise is wrong (for truly random markets).
If we are talking about truly random markets, then each (e.g.) 10 pip move up or down should always have a 50% chance. For example: price moves up 10 pips, then it moves up another 10 pips, then it moves down 10 pips.
Each 10 pip move here has a 50% chance of occuring, and is not dependent on the move that happened beforehand (if we talk about truly random markets).

For mean reversion strategies to work, our premise has to be that after a 10 pip move up, a 10 pip move down is more likely to occur than another 10 pip move upwards. But this is then not a random market at all.
Thread Tools Search this Thread
Search this Thread:

Advanced Search

0 traders viewing now