Probability Question | Forex Factory

Options

Permalinks
Print Thread

Similar Threads

Question on probability based trading idea 11 replies

Simple Question, maybe a dumb question 2 replies

Just a fast question - Need help for an exam trading question 0 replies

Direction probability question. 0 replies

Hedging Question - Quick Question Looking for A Quick Answer 1 reply

Probability Question

  • #1
  • First Post: Jan 19, 2009 9:51pm Jan 19, 2009 9:51pm
I am not very good at maths..

Can someone with better maths verify below:

If there only 2 trades per day. BUY SL10 TP20 && SELL SL10 TP20
when one hit TP, Close All.
5 days per week. If both hit SL 2 times in a week, no further trades for the week.

e.g:
bestweek
Day 1=+20
Day 2=+20
Day 3=+20
Day 4=+20
Day 5=+20

= total +100

Worseweek
Day 1=-20
Day 2=-20
Day 3 no trade
Day 4 no trade
Day 5 no trade
=total -40

I check the variances for all scenario.
Probability for breakeven(0) and positive gain week is 12/15 scenario.
Probability for losing week is 3/15 scenario.

Is it correct?
  • #2
  • Jan 19, 2009 10:27pm Jan 19, 2009 10:27pm
One thing you might not have thought about:

Your SL is 1/2 your TP, so you will stop out twice as much as TPing in a random market.

And you will pay spread on each trade.
  • #3
  • Jan 19, 2009 11:47pm Jan 19, 2009 11:47pm
The problem with this is defining the probabilities for the single day, I think it is safe to simplify it to one trade per day, going randomly long and short does not increase the odds of your t/p being hit. If your t/p is twice as big as your s/l, the best mapping we can make is that you are twice as likely to lose than win, so win probability = 33% and loss probability is = 66% for any given event. We assume independence of events.

Best week = .33 ^ 5 = 0.004 (or 0.4%)
Worst week = .66 ^ 2 = 0.43 (or 43%)
Break Event = .66 * .33 * .33 *.66 = 0.05 (or 5%)

As you can see, the odds are not in your favor, the assumptions though simple are probably not far wrong given the largely random nature of markets.
Even if you assume my simplification of the initial problem is wrong, it would require a significant skewing of the odds (i.e. an edge) to give you a positive expectation.
The breaking of a wave cannot explain the whole sea.
  • #4
  • Jan 20, 2009 2:20am Jan 20, 2009 2:20am
Quoting Craig
The problem with this is defining the probabilities for the single day, I think it is safe to simplify it to one trade per day, going randomly long and short does not increase the odds of your t/p being hit. If your t/p is twice as big as your s/l, the best mapping we can make is that you are twice as likely to lose than win, so win probability = 33% and loss probability is = 66% for any given event. We assume independence of events.

Best week = .33 ^ 5 = 0.004 (or 0.4%)
Worst week = .66 ^ 2 = 0.43 (or 43%)
Break Event = .66 * .33 * .33 *.66 =...
Great Craig, I think the explaination is quite clear.

How about if the trade based on some indication or statistics

Example of daily life:
If we see black cloud, the probability of today raining is high.
If we see the sunshine, the probaility of today raining is low.

Do in Forex we also have some indication for trending up, trending low, etc..

Just a thought..
  • #5
  • Jan 20, 2009 11:03am Jan 20, 2009 11:03am
Well that is the definition of an edge, basically that you are going to get enough positive outcomes due to none random behavior to give yourself a positive expectancy in the long run, even then nothing is guaranteed.

For example, I have 335 trades in a current strategy, win probability is 55%, and win loss ratio is 0.94, search for 'expectancy'.
The breaking of a wave cannot explain the whole sea.
  • #6
  • Jan 21, 2009 1:31am Jan 21, 2009 1:31am
Quoting Craig
Well that is the definition of an edge, basically that you are going to get enough positive outcomes due to none random behavior to give yourself a positive expectancy in the long run, even then nothing is guaranteed.

For example, I have 335 trades in a current strategy, win probability is 55%, and win loss ratio is 0.94, search for 'expectancy'.
I learn a lot by your suggestion. I google "expectancy"

I think anyone who like to create trading system, need to start read and understand this. Should this articles been read for all who post their trading system in FF.

http://www.articlesbase.com/finance-...ent-76483.html

All the while, i am more into coding system.
Now, a lot more to learn and read.

Thanks Craig.
  • #7
  • Jan 21, 2009 2:00am Jan 21, 2009 2:00am
Check out this http://www.elitetrader.com/vb/showth...threadid=33654, maybe slightly heavy going for the beginner but worth a look.
The breaking of a wave cannot explain the whole sea.
  • #8
  • Last Post: Jan 23, 2009 8:31am Jan 23, 2009 8:31am
Quoting Craig
Check out this http://www.elitetrader.com/vb/showth...threadid=33654, maybe slightly heavy going for the beginner but worth a look.

Thanks Craig,


nice link.
__Thanks__ MJ
0 traders viewing now