misguided logic? | Forex Factory

Options

Permalinks
Print Thread

Similar Threads

help me confirm this logic! 5 replies

Applying Logic to trading 5 replies

Making money in spite of fundemental logic (as we may understand it) 5 replies

Having Trouble with Cross Logic for Indicator. 1 reply

Logic is missing is most systems 2 replies

misguided logic?

  • #1
  • First Post: May 2, 2010 12:40am May 2, 2010 12:40am
what am i missing here...
if i am testing a strategy and i get, lets say, 60% negative results. and I am using a simple MA cross as my buy/sell signals, and I have 20 tp and 20 sl....

if i were to turn my current buy signal into a sell signal and vice versa, and run the strategy tester for the same time period leaving everything else the same; shouldn't i now get a 60% positive result?

****I dont think this is a good way to build a strategy, but i'd just like to get a sense of how much i can rely on the strategy tester...****

i didn't work out the way i thought it should have..
  • #2
  • May 2, 2010 12:44am May 2, 2010 12:44am
Quoting cue232s
what am i missing here...
Quite a lot it would seem if you have been a member since 2006 and managed not to pick up on the millions of threads on this very subject.
The breaking of a wave cannot explain the whole sea.
  • #3
  • May 2, 2010 12:49am May 2, 2010 12:49am
yea.. i got interested in the forex in 2006.. opened and account.. started a business, got very busy, and now i'm back.. I have missed a lot..
Sorry, but I'm happy they kept my seat warm here at ff!!
  • #4
  • May 2, 2010 12:50am May 2, 2010 12:50am
maybe you could point me to a few threads?
  • #5
  • May 2, 2010 12:54am May 2, 2010 12:54am
Quoting cue232s
if i were to turn my current buy signal into a sell signal and vice versa, and run the strategy tester for the same time period leaving everything else the same; shouldn't i now get a 60% positive result?
No, you're not supposed to get 60% positive results. If your strategy produces 100 trades, each trade losing 1 pip on average, and running on pair which have 2 pip spread, after reversing, the 100 trades will still see 1 pip loss on average, due to spread..
  • #6
  • May 2, 2010 12:56am May 2, 2010 12:56am
Consider this...

Spread costs?
  • #7
  • May 2, 2010 12:58am May 2, 2010 12:58am
you have to deal with spread cost regardless of which way you play the market (i.e. short or long)
  • #8
  • May 2, 2010 9:58am May 2, 2010 9:58am
Quoting _Nopik_
No, you're not supposed to get 60% positive results. If your strategy produces 100 trades, each trade losing 1 pip on average, and running on pair which have 2 pip spread, after reversing, the 100 trades will still see 1 pip loss on average, due to spread..
two things:

1. for some reason when i use strategy tester, it doesn't include spread in the results

2. each losing trade loses exactly 20 pips, and each winning trade wins exactly 20 pips.

That's why i'm so confused.
  • #9
  • Edited 10:08am May 2, 2010 10:08am | Edited 10:08am
with spread cost,with fixed 20 pip tp and sl and spread is 2 pip, regardless your position (short/long):
- price has to move 22 pip in your direction to hit TP
- price only need 18 pip movement against your direction to hit SL
  • #10
  • May 2, 2010 10:40am May 2, 2010 10:40am
Quoting cue232s
two things:

1. for some reason when i use strategy tester, it doesn't include spread in the results

2. each losing trade loses exactly 20 pips, and each winning trade wins exactly 20 pips.

That's why i'm so confused.
Wrong.

Spread is counted. If you don't believe me, write an EA which buys and sells instantaneously and in rapid succession. You'll love money. Why? Because of the spread!
  • #11
  • May 2, 2010 10:47am May 2, 2010 10:47am
yea, you guys are right.. spread is in there.
but this is crazy, so that means that at least 20% of my trades are going at least 17 of the 20 pips in either direction, then retracing and going at least 17 of the 20 pips in the other direction... thats why, no matter if its a buy or sell signal they are always negative?

is that likely?
  • #12
  • May 2, 2010 11:09am May 2, 2010 11:09am
No.

Run your backtest in visual mode.
  • #13
  • May 2, 2010 12:16pm May 2, 2010 12:16pm
I'm back testing now in visual mode.. thanks for the advice.
Question though, does strategy tester account for each and every tick exactly ans the it happened live(or theoretically at least)? I ask because the history data i downloaded was for the 1min time frame. does this contain each and every tick?
  • #14
  • Last Post: May 2, 2010 1:05pm May 2, 2010 1:05pm
No
0 traders viewing now