Stormin_Norman
Currency Trader
- Joined
- 12 January 2008
- Posts
- 1,256
- Reactions
- 0
yep. simple.
Autotrader test results
- Thread starter Stormin_Norman
- Start date
yep. simple.
sounds like the futures is a more efficient market .... but there are more trading opportunities in an inefficient market
- Joined
- 18 June 2008
- Posts
- 1,071
- Reactions
- 2
sounds like the futures is a more efficient market .... but there are more trading opportunities in an inefficient market
But you have people using futures to hedge currency risk, thus creating inefficiencies
Stormin_Norman
Currency Trader
- Joined
- 12 January 2008
- Posts
- 1,256
- Reactions
- 0
that's true scalping. not the misused common meaning of the term.
the EURJPY looks the most promising, but what's the difference between the two graphs: is that different versions of the EA, different data providers, or different time periods?
BentRod
ZZZZzzzzZZZZzzzzzzzz
- Joined
- 16 December 2005
- Posts
- 694
- Reactions
- 0
But you have people using futures to hedge currency risk, thus creating inefficiencies
Could this be the holy grail???
J/K
Reckon your right there mate.
Thats what keeps the markets inline.
Stormin_Norman
Currency Trader
- Joined
- 12 January 2008
- Posts
- 1,256
- Reactions
- 0
one of the variables which tells the EA how far back to look when making its decision making.
got a lot of result data which we have to go through and look at - these are just the 'best' results.
but finding the right settings isnt about just selecting the 'best' setting, but selecting the setting that are most likely to work.
there is a difference.
Regarding what's most likely to work, I've found monte-carlo style methods are a good way to sort out the "lucky" from the "likely".
I think the reason why the "best" (biggest) result might not be the right setting to choose is due to drawdowns. I start by maximising the best result in the long term, but then enforce a strict drawdown policy for the medium term.
For example I'm looking at a system that trades between 50 and 100 trades a day (each trade is short term), so I enforce the drawdown limit on daily results (medium term), using initial parameters that were chosen to maximise the result over a week (long term). The aim is to have losing days but no losing weeks, with a straight line equity curve showing the edge.
I think the reason why the "best" (biggest) result might not be the right setting to choose is due to drawdowns. I start by maximising the best result in the long term, but then enforce a strict drawdown policy for the medium term.
For example I'm looking at a system that trades between 50 and 100 trades a day (each trade is short term), so I enforce the drawdown limit on daily results (medium term), using initial parameters that were chosen to maximise the result over a week (long term). The aim is to have losing days but no losing weeks, with a straight line equity curve showing the edge.
Stormin_Norman
Currency Trader
- Joined
- 12 January 2008
- Posts
- 1,256
- Reactions
- 0
- Joined
- 18 June 2008
- Posts
- 1,071
- Reactions
- 2
Why is it called "franklin - fixed exit v2" ..... it looks to me that your system moves its stop losses to break even..?
That drawdown at trade 148 must of hurt
Stormin_Norman
Currency Trader
- Joined
- 12 January 2008
- Posts
- 1,256
- Reactions
- 0
Why is it called "franklin - fixed exit v2" ..... it looks to me that your system moves its stop losses to break even..?
That drawdown at trade 148 must of hurt
cause its not a variable stop loss, it does things due to 'fixed' variables on exit.
v3 uses an exit strategy which will change depending on how the market moves resulting in different stop losses and 'stop profits' for each trade depending on the market.
the drawdown @ 148 happens. its looking for trends and can go sideways waiting for said trends.
