Forex Backtests
How accurate is forex backtesting on metatrader?
The highest modeling quality that can be achieved when backtesting is 90%. That difference alone is the difference between most any system being profitable or not and does not begin to give a useful representation of the system.
I myself do not believe in backtesting because as a system developer it would be a huge disadvantage as the natural progression for anyone backtesting a trading system is to then curve fit the strategy to past data which produces system ruin when traded in real time. I believe this is why so many talented computer science engineers who are attracted to the markets and believe their coding abilities (close friend develops artificial intelligence systems for NASA and has tried in forex to no avail) will lead them to crack the code themselves and are unable to. They all always fall into the trap into over optimization, which is only natural. Now someone with those gifts and abilities working with a trader can produce very different results in my opinion. That is the ideal situation I believe.
Forex Backtesting
I keep the trade by trade breakdown as the system is traded in real time that you can refer back to. I myself only forward test and have much success with this approach. As you can see from my trading styling though everything about how I approach the markets is different. Which also makes sense as the norm seems to lead to the same results for the herd.
Guess off of the top of my head I would say the win/loss ration is close to about 35% (discounting all trades closed at +10 pips). The minimum risk to reward ratio which takes places on the first trade order is 1:3 which is why with a relatively low win/loss ration they system is able to be profitable. The second trade order has a risk/reward ratio of 1:6 and 1:12 on the third trade order.
