Quote:
Originally Posted by jwhite1319
Hi Pcontour,
Do you know where the most recent update to the settings file can be found. It seems that you are the only one of the programmers left working on Phoenix!
I would like to play with it and see if i can get a new settings file
|
The way things worked is that the preferred settings in the program, were updated with the latest "best settings found". So if you move the values of the pref settings in the program to the settings file then you will have the latest best settings. People found other best settings. The truth is that the program did not have the latest settings that people found. In my opinion, you don't need anybody else's settings files to optimize. I think it is important that you don't use money management when working on you signal settings.
The settings document doesn't reflect my opinion on the best way to get good long term settings. Generally what people were doing was getting good settings over the most recent month and hoping they would work for the next month. My opinion is that the best way to get good long term settings is to use as long a period as you can and find something that works over a very long period. No settings are perfect.
How do you mitigate against a long string of losses. This is where the decrease factor comes in. It chops your bet down dramatically after two out of the last three trades are losers. That strategy was very effective in increasing profits after I changed it. It did have one market that killed it. A long series of every second trade being a loser. There are always some losses. FX is a percentages game. A simple repair to that strategy can result in a different pattern that would cause a problem. One key to the decrease strategy working well is the ability to ramp back up aggressively and trade again as fast as it ramps down.
How do you cut your losses and let your winners run. There is a limit to the losses in terms of pips. You never let it go lower than designed 80 pips. To let the winners run. I made a trailing stoploss that tightens up as you go higher into the profit zone. People don't realize it is a percentages game. They said to me, I was 40 points in the profit and the old code would have taken the profit but instead, I ended up with 5 pips only, can you make it only trail by 10 pips. On the flip side. I saw one trade go up 260 pips in the profit. In that case you miss out on 220 pips. When you tighten it to 10 pips you lose long term. The profits were drastically lower. You need to have a big picture mind set. When you make a change to try to rectify one particular situation where you lost, you create a different situation where you create a different loss. or miss out on a win.
Real trading doesn't match back testing for a number of reasons that Ian innumerated. You need strong backtests to hope to get good profits in real trading. Small tuning differences in backtests can have no difference in real testing. I feel that some of the original settings from Hendrick along with recent improvements to the money management can produce the best results. The key part of the settings file that you really want to concentrate on tuning is the signals themselves. You can go back to some original version and find Hendrick's settings that still may work well today.
10-25-2007, 04:46 AM
Similar Threads
Linear Mode
