hi trendchaser.. i did not receive the EA about that strategy of continuation trend.
could you send for my e-mail?
could you send for my e-mail?
Supernova GBP/JPY Mini Trend Catcher
- #4,883
- Feb 8, 2008 3:50pm Feb 8, 2008 3:50pm
- Joined Jun 2007 | Status: Unconventionable Conventionist | 2,768 Posts
Hi All,
I have been pretty quiet while I've had my head buried in code. I have been working on Money/Risk management and reactive indicators (plus a couple of other requests like same bar reverse trade on close, etc). In addition, I am hoping to release an updated SuperNovaStats that will give us more data to help analyze our results. I hope everyone has been finding the indicator useful.
I am wondering if the testers might consider postponing the MM/RM testing for one week so that we can test the reactive code. I have added several optional filters, all of which can be used in reactive and standard modes. These filters are to help us avoid getting bad fills due to news spikes and other unforeseen events. If we try to test too many new items at once, we won't know what had the greatest affect. Also, that would give me another full week to test the RM/MM and I would feel a lot better with more testing as I have been coding a great deal and want to make sure all of the other features are stabile before unleashing the MM/RM.
Let me know how you feel about this plan.
Rick
Great features you have put into our little Mr. Roboto and I guess we will be more than happy to test the beta7 before we go into the MM side of the test!
Have a good weekend - I will be around here on Saturday and Sunday,too!
Ciao
Matt
- #4,884
- Feb 8, 2008 3:53pm Feb 8, 2008 3:53pm
- Joined Jun 2007 | Status: Unconventionable Conventionist | 2,768 Posts
Rick
Any way you would like us to test sounds fine to me.
I have been without an internet connection (connectus interruptus) for about 16 hours but everything seems fine now.
First week of live testing shows a profit of 131 pips but, when I lost my connection last night, I called my broker and had them close all four open trades (all were in profit).
In my utter paranoia I have DSL, a vodafone card for my laptops and a group of unprotected WLAN networks of my neighbors I can use here! Now I am thinking about cable as a backup! 
:
Hope you found a good medicine against that c.i. you caught there!
- #4,891
- Feb 9, 2008 3:29am Feb 9, 2008 3:29am
- Joined Jun 2007 | Status: Unconventionable Conventionist | 2,768 Posts
Here's a screenshot of my results as reported by the indie and also a statement.
IBFXm, sntest8
Great work there!
Would you care to share the excel sheet you designed? I guess the testers could just fill in their personal results there. 
Happy weekend!
Matt
- #4,895
- Feb 9, 2008 9:06am Feb 9, 2008 9:06am
Hallo Dee!
Great work there!
Happy weekend!
Matt
I will, but I'd prefer not to spend the time logging the stats. Personal observations to date suggest that CCI has not been a good indicator - often lagging for entries and too quick or even late getting out.
I also think the less indicators, the better. Time to put the CCI away perhaps?
Would you care to share the excel sheet you designed? I guess the testers could just fill in their personal results there.
or at the upper right hand paperclip (attachments) it's posted as Supernova_testing.xls on February 5th.
- #4,897
- Feb 9, 2008 11:57am Feb 9, 2008 11:57am
- Joined Jun 2007 | Status: Unconventionable Conventionist | 2,768 Posts
It can be found here http://www.forexfactory.com/showpost...postcount=4824
or at the upper right hand paperclip (attachments) it's posted as Supernova_testing.xls on February 5th.
Beg your pardon!
: You see - I download the things and if I do not have time to really look at that they go into a "for further use" folder so they do not congest my desktop. I am really sorry - I have it here! Thanks!!Next time please start a completel advertising campaign!
Thanks!
Matt
I agree with that. However, the CCI testing I've been doing has been consistently poor. Only the EURCHF on a 30 min chart, and 8 trades for $28, is not up to much.
I will, but I'd prefer not to spend the time logging the stats. Personal observations to date suggest that CCI has not been a good indicator - often lagging for entries and too quick or even late getting out.
I also think the less indicators, the better. Time to put the CCI away perhaps?
Hi Dave,
One of the goals of all this testing is to see what we should leave in and what can be removed. Some of the indicators are good for determining entries and exits, while others are better at confirming whether the enty/exit signals are valid. CCI falls into the confirming category by all observations and our testing confirms that as an exit signaller it well and truly sucks
. When we are finished with all the testing, we will have a post-mortem in which anything deemed unnecessary will be removed from the final version. After we are all happy with the feature set, if anyone wants a version for another platform (NinjaTrader, OpenQuant, TradeStation), I will port SuperNova. I have been really happy with the testing and as I have said before, I think MM/RM will make a world of difference. So I think the effort to port this to other platforms will be worth it.
Rick
- #4,899
- Edited 5:09pm Feb 9, 2008 3:00pm | Edited 5:09pm
Hi All,
Here is Beta 7. It doesn't have all the features I had hoped and I apologize for that. Beya 8 will have the Money and Risk Management. Promise
This version contains the reactive code and some mechanisms for filtering trades that get bad fills caused by price spikes (news) and fast moving markets that cause price to stray wildly from the moving average causing a valid, but usually poor, entry.
Here are the new inputs:
extern bool UseReactiveEntryQQE = false;
When true, QQE will use the current bar for entry decisions.
extern bool UseReactiveExitQQE = false;
When true, QQE will use the current bar for exit decisions.
extern bool UseReactiveEntry = false;
When true, all non-QQE indis will use the current bar for entry decisions.
extern bool UseReactiveExit = false;
When true, all non-QQE indiswill use the current bar for exit decisions.
extern bool UseReactivePipsFilter = false;
extern int ReactivePipsFilter = 5;
These 2 are used together. When the boolean is true, the entry and exit conditions will have to be true for ReactivePipsFilter pips before the trade will occur. This is meant to filter brief crosses of QQE and the other indis which could cause false entries and exits. This will need optimization.
extern bool UseSlopeFilter = false;
When this is true, the entry condition must agree with the slope of the internal EMA. So if a long is triggered, it will only be placed if the ema slope is upward. Works with reactive and standard settings.
extern bool UsePriceFilter = false;
When this is true, the current price must be below the internal ema when entry condition is met for a trade to occur. So if the ask is below the ema price for a long, the trade is valid. This helps protect against taking trades after price spikes that cause valid crosses and helps to eliminate bad fills. Works with reactive and standard settings.
extern bool SameBarReverseTrade = false;
This allows a trade in the opposite direction to occur on the same bar as a just closed trade. So id a long closes and the conditions are valid for a short, the short will be placed on the same bar. Works with reactive and standard settings.
If all of the inputs are false, Beta 7 works just like beta 6.
Please let me know if anything is not working.
For the testers, I will come up witha test plan to include these features shortly.
Rick
Here is Beta 7. It doesn't have all the features I had hoped and I apologize for that. Beya 8 will have the Money and Risk Management. Promise
This version contains the reactive code and some mechanisms for filtering trades that get bad fills caused by price spikes (news) and fast moving markets that cause price to stray wildly from the moving average causing a valid, but usually poor, entry.
Here are the new inputs:
extern bool UseReactiveEntryQQE = false;
When true, QQE will use the current bar for entry decisions.
extern bool UseReactiveExitQQE = false;
When true, QQE will use the current bar for exit decisions.
extern bool UseReactiveEntry = false;
When true, all non-QQE indis will use the current bar for entry decisions.
extern bool UseReactiveExit = false;
When true, all non-QQE indiswill use the current bar for exit decisions.
extern bool UseReactivePipsFilter = false;
extern int ReactivePipsFilter = 5;
These 2 are used together. When the boolean is true, the entry and exit conditions will have to be true for ReactivePipsFilter pips before the trade will occur. This is meant to filter brief crosses of QQE and the other indis which could cause false entries and exits. This will need optimization.
extern bool UseSlopeFilter = false;
When this is true, the entry condition must agree with the slope of the internal EMA. So if a long is triggered, it will only be placed if the ema slope is upward. Works with reactive and standard settings.
extern bool UsePriceFilter = false;
When this is true, the current price must be below the internal ema when entry condition is met for a trade to occur. So if the ask is below the ema price for a long, the trade is valid. This helps protect against taking trades after price spikes that cause valid crosses and helps to eliminate bad fills. Works with reactive and standard settings.
extern bool SameBarReverseTrade = false;
This allows a trade in the opposite direction to occur on the same bar as a just closed trade. So id a long closes and the conditions are valid for a short, the short will be placed on the same bar. Works with reactive and standard settings.
If all of the inputs are false, Beta 7 works just like beta 6.
Please let me know if anything is not working.
For the testers, I will come up witha test plan to include these features shortly.
Rick
- #4,900
- Feb 9, 2008 3:10pm Feb 9, 2008 3:10pm
- Joined Jun 2007 | Status: Unconventionable Conventionist | 2,768 Posts
I agree with that. However, the CCI testing I've been doing has been consistently poor. Only the EURCHF on a 30 min chart, and 8 trades for $28, is not up to much.
I will, but I'd prefer not to spend the time logging the stats. Personal observations to date suggest that CCI has not been a good indicator - often lagging for entries and too quick or even late getting out.
I also think the less indicators, the better. Time to put the CCI away perhaps?
the influx of new indicators into the EA seemed unstoppable at some point, so there is really a lot inside that may be subject to question. The CCI was entered as the manual traders used it for a timely exit. But even that was just a back up system and they can always fade out within a candle.
So I guess there will be more indicators on the "De morituri te salutant" - list in the end. But we still have no real grip on the way the CCI may come in handy. I would always say "Get rid of the LaGuerre!", but first of all some people had major success with that and also my tests were not the worst, although I was carrying the LaGuerre burden again!
But surely there is a clock ticking on about everything that was not mentioned in the first post all these months ago - maybe apart from the PSAR in its multitude of ways to use.
Have a good weekend and sorry for skipping the post ... I only realized after Rick answered.
Ciao
Matt