Spread or transaction cost?

Viewing 5 posts - 1 through 5 (of 5 total)

Author

Posts
04/16/2018 at 12:55 AM #68431 quote

bezieh
Participant

Average

Hey guys, i got a system for OMX30 SEK20 on a 1h timeframe, making about 1 trade per day in average on 4 years.
And when i backtest without any spread or transaction cost, it turns out to be nice, about 300% profit in 4 years, but when i add the spread on 1 point, it goes from +300%, to -100%, haha, is the spread this extremely expensive, or is it something wrong here?

One omx30 costs lets say 1600, and if you buy it cost 1601.5, so its 1.5 points u pay basically, which is around 0.1%, and when i remove spread cost in PRT backtest, and add 0.1% transaction cost, suddenty it makes sense, and goes from +300% to like +250% in 4 years, isn’t that the correct way? Or is it really spread 1 point on each trade, instead of transaction cost?

Because i really thought my system was nice, until i added the spread cost of “1 point”.

What about the spread cost of other index, for example portugal got 50 points in spread, how is this possible to profit from this on a low timeframe? Isn’t it transaction cost that you should apply?

Thanks for answers guys!

04/16/2018 at 10:10 AM #68452 quote

Eric
Participant

Master

You should look at average gain

personally i would not trade (for example) a system on DAX with less than 10 points avg gain

04/16/2018 at 10:58 AM #68460 quote

GraHal
Participant

Master

Just to add a bit more on Eric’s well spotted observation … what is the Average gain on the System you describe above?

See attached for where Average gain is shown.

Bez-1.jpg

05/26/2019 at 1:46 PM #99454 quote
Paul
Participant

Master
same thing with this code on gbpusd 15min with no spread. Changed the buy/sellshort of original code.

Looks nice, but average gain is too low.
```
// REVERSI-4
//www.doctrading.fr

Defparam cumulateorders = false

// REVERSI HAUSSIER
ca1 = close > open and close[1] > open[1] and close[2] < open[2] and close[3] < open[3]

IF ca1 THEN
sellshort 5 contracts at market nextbaropen
amplitude = close - low
ENDIF

// REVERSI BAISSIER
cv1 = close < open and close[1] < open[1] and close[2] > open[2] and close[3] > open[3]

IF cv1 THEN
buy 5 contracts at market nextbaropen
amplitude = high - close
ENDIF

// STOP LOSS & OBJECTIF
set stop loss amplitude
set target profit amplitude
```
Screenshot-2019-05-26-at-14.26.41.jpg
05/26/2019 at 1:57 PM #99456 quote

Paul
Participant

Master

to compare, with an optimised factor set for amplitude.

Left without spread, right with spread set to 1.

Nicolas thanked this post

Screenshot-2019-05-26-at-14.54.29.jpg Screenshot-2019-05-26-at-14.54.35.jpg
Author

Posts