Old Dog with New Tricks | Page 45 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Old Dog's Taming of the Beast? 34 replies

Old Dog Fib Trading - EA Development 18 replies

Old Dog Fib Trading - EURUSD 399 replies

5 Tricks of Trend Trading, by Rob Booker 25 replies

How To Work Around A Market Maker's Tricks 0 replies

Attachments: Old Dog with New Tricks
Exit Attachments

Old Dog with New Tricks

  • #881
  • Aug 28, 2010 8:39am Aug 28, 2010 8:39am
Hi

Ron, can you modify the EA to not open trades at Friday evening(e.g. after London session), because the brokers increase spread 2-3 times

Regards
  • #882
  • Aug 31, 2010 9:08am Aug 31, 2010 9:08am
Hi all (and especially Ronald, for whom these questions are),

I have discovered this thread a few days ago, and I've begun trying the EA on a demo account.

I would like to ask Ronald a question: once optimization is done, if we look at the Optimized Settings file, we see some lines with -1.

If I'm correct, this means that no optimal settings have been found for the corresponding hour.

My question is: does this mean that there will NEVER be any trade opened during that hour for that pair, unless a future re-optimization generates new settings?

And another question: Ronald, you say that it's better to use as many bars as possible. You are using 1.5M bars, which corresponds to 3 years I think. But I suppose this means that the weekly re-optimization has only little chance to provide different results, because only one week is added (which is less than 1% of total bars), so the impact of one single week is very limited.

In other words, with respect to my 1st question, this means if settings do not allow trades for a given hour, there is little chance this will change in the near future? Or maybe I've missed something...

I'm asking those questions, because with settings such as WinRate = 60% and RR = 1.3, some pairs contain almost only "-1" rows in the Optimized Settings... which means (correct me if I'm wrong) there will almost never be any trades for that pair?

Thanks for your replies!

JL
  • #883
  • Aug 31, 2010 9:34am Aug 31, 2010 9:34am
Quoting jl83
Hi all (and especially Ronald, for whom these questions are),

I have discovered this thread a few days ago, and I've begun trying the EA on a demo account.

I would like to ask Ronald a question: once optimization is done, if we look at the Optimized Settings file, we see some lines with -1.

If I'm correct, this means that no optimal settings have been found for the corresponding hour.

My question is: does this mean that there will NEVER be any trade opened during that hour for that pair, unless a future re-optimization generates new settings?...
Settings of -1 do mean the EA won't trade. It's basically illogical information which the EA cannot trade with so it ultimately disables the EA.

The impact of a single week is of course limited. When using such a large sample size, I'm able to locate the long-term patterns within the hilo trading system that are consistent and therefore predictable. A lot of people using bar counts under 100,000 bars will suffer from lag when the EA is adjusting for the winter slow session and the summer's occasional volatility. In theory, 1 year minimum (~300,000 bars) is all that's needed to make sure the EA optimizes itself against all market scenarios. I use 5 years so at least I've covered the pre-subprime market, the subprime-crash, and the post-crash.

The way I'm trading, generally what changes are the stoploss/takeprofits for a given pair every time I reoptimize. But overall whether or not there is a trade for a given hour, one hour changes every 2 weeks or so.

If you see only -1 in the optimized settings, that's perfectly acceptable. I never really understood why some users need this EA to trade at least once every day.
  • #884
  • Aug 31, 2010 11:07am Aug 31, 2010 11:07am
Thanks Ronald for your reply.

No problem with -1, it was just to make sure I fully understand how to interpret this.

Just 2 more questions...

You say everytime you re-optimize, the main changes are S/L and T/P. But I thought you based these settings on 10 x spread, and I guess the average spread does not vary? Or did I misinterpret what you mean?

Other question if I may... in one of your webinars you mention that this EA is the first of a series of 8 EA's. I suppose you mean 8 EA's based on the same auto-optimization principle... May I ask you which core strategy you plan to trade with those 8 EA's?

Thanks
  • #885
  • Aug 31, 2010 12:59pm Aug 31, 2010 12:59pm
Quoting jl83
Thanks Ronald for your reply.

No problem with -1, it was just to make sure I fully understand how to interpret this.

Just 2 more questions...

You say everytime you re-optimize, the main changes are S/L and T/P. But I thought you based these settings on 10 x spread, and I guess the average spread does not vary? Or did I misinterpret what you mean?

Other question if I may... in one of your webinars you mention that this EA is the first of a series of 8 EA's. I suppose you mean 8 EA's based on the same auto-optimization principle... May I ask...

  1. When optimizing using this EA, you choose whether or not you want a fixed stoploss or a fixed takeprofit. I have EAs where I am optimizing both--independently. That's where I notice the most changes.
  2. The strategies traded using those EAs are the ones that I either created alongside the people who have contracted me to create these systems for them, or manual systems that have been automated by myself.

  • #886
  • Aug 31, 2010 1:37pm Aug 31, 2010 1:37pm
Thanks.

I had started initially using M5 charts, and Rfking's version of the EA, but after watching the webinars, I switched to M1.

I have launched optimization with the following settings on 9 majors:
- 200 SL (i.e. 20 pips)
- reverse = false
- 150000 bars
- WR 60%
- RR 1.1
- Optimize based on Success Score
- everything else at default values

I'm quite surprised with the results, on average I have only 2 rows (i.e. 2 hours) with values that are not -1. On some pairs 3 or 4, on others only 1. I don't think my settings are very tight... If I use WR 60% and RR 1.3, not a single usable hour!

Does this seem normal to you?
  • #887
  • Aug 31, 2010 1:58pm Aug 31, 2010 1:58pm
Quoting jl83
Thanks.

I had started initially using M5 charts, and Rfking's version of the EA, but after watching the webinars, I switched to M1.

I have launched optimization with the following settings on 9 majors:
- 200 SL (i.e. 20 pips)
- reverse = false
- 150000 bars
- WR 60%
- RR 1.1
- Optimize based on Success Score
- everything else at default values

I'm quite surprised with the results, on average I have only 2 rows (i.e. 2 hours) with values that are not -1. On some pairs 3 or 4, on others only 1. I don't think my settings are very tight... If...
Yes
  • #888
  • Aug 31, 2010 2:26pm Aug 31, 2010 2:26pm
So this means that settings such as WR 60% and RR 1.3 are simply un-usable?
  • #889
  • Aug 31, 2010 3:02pm Aug 31, 2010 3:02pm
Quoting jl83
So this means that settings such as WR 60% and RR 1.3 are simply un-usable?

I have been using this on EU and UJ since last monday with an sl of 10 (or 100). Whilst I know that RR has suggested using 10 times spread and that you don't need it to trade every day I was just wanting to test the theory. I had previously been using a higher sl. This combined with a win rate of 50% and a min success score of 0.8 has resulted in some very good risk reward scores (sometimes 2 or higher). I am testing over 500,000 but may increase this. I posted my results last week and so far this week it has given up some of last weeks gains. This EA has real potential it is just a matter of everyone finding what they are comfortable with.

RFKing posted earlier that he simply sets a high win rate. I tried this over the weekend but win rates of 80% or 90% gave me lots of trading opportunuities but very low risk reward scores. The tp's were so low it was nearly like scapling which I know RR has suggested it is not designed for. Having said that I am tempted to give it a try.

Paul
  • #890
  • Aug 31, 2010 3:15pm Aug 31, 2010 3:15pm
Thanks for your reply.

I'm currently testing various combinations as well, but I remember having read some posts about combining WR above 60 and RR above 1, so that's why I have tried to do.

But the result is that my Optimization file contains only -1 rows, which means it will never take any trade.

That's why I asked this question...
  • #891
  • Aug 31, 2010 3:20pm Aug 31, 2010 3:20pm
Quoting jl83
Thanks.

I had started initially using M5 charts, and Rfking's version of the EA, but after watching the webinars, I switched to M1.

I have launched optimization with the following settings on 9 majors:
- 200 SL (i.e. 20 pips)
- reverse = false
- 150000 bars
- WR 60%
- RR 1.1
- Optimize based on Success Score
- everything else at default values

I'm quite surprised with the results, on average I have only 2 rows (i.e. 2 hours) with values that are not -1. On some pairs 3 or 4, on others only 1. I don't think my settings are very tight......
Ronald I noticed the less bars to optimise we have the more deficiencies we find e.g. tradable hours. Like in above example with 30 pips stop loss if we use more bars to optimize (700000) we have actually no tradable hour. Based on this findings, does it not make sense to you actually to take the opposite trades of this EA? The deficiences seem to actually balance in time.

So if at 150000 bars the hour 8 is tradable we suppose to get against the EA when we chosse the direction to trade because with 700000 the hour 8 is not tradable. Am I making sense?



Also How come you find anything to trade with 1.4 mil bars.... Am I missinfg something? Do you find anything over 60% winrate? with a decent risk/reward?
  • #892
  • Sep 1, 2010 12:47am Sep 1, 2010 12:47am
Quoting iulli
So if at 150000 bars the hour 8 is tradable we suppose to get against the EA when we chosse the direction to trade because with 700000 the hour 8 is not tradable. Am I making sense?

Also How come you find anything to trade with 1.4 mil bars.... Am I missinfg something? Do you find anything over 60% winrate? with a decent risk/reward?
Not sure I would go against the EA in this situation, anyway there is no option to do this I guess... but let's wait for Ronald's answer.

However, your second question makes more sense to me... I wonder how you can find tradeable hours if you tighten a bit the requirements. I'm OK not trading all hours of the day of course, but what if the reasonable settings do not allow any single tradeable hour?

(by reasonable settings, I mean in terms of risk and representativity of the sample, i.e. a decent WR, a decent RR and a decent number of bars)

I guess we probably miss something here...
  • #893
  • Sep 1, 2010 6:41am Sep 1, 2010 6:41am
Nobody is missing anything.

When I trade with this type of EA, I try to set it up so I get a handful of trades per week.

When you test against the larger datasets, it's expected that trade opportunities are harder to find. When you think about this, it makes sense. The EA is looking for a trading opportunity in the past x bars which is relatively consistent based on the requirements you have for the EA. In my case, 5 y ears ago, everybody on wall street was sitting pretty, it was the period before the financial collapse. FX market changed when the mortgage market crashed, and we are currently in the post-crash world.

That's two, possibly three paradigm-shifting events that happened in the market. The fact that I'm fortunate enough to have a system that seems to be working through that period with the expectations that I have for it is really interesting. I always try to force the EA to optimize through two different markets. Why? It gives me some peace of mind knowing that the system that the EA is currently trading is not dependent on a particular market.
  • #894
  • Sep 1, 2010 8:28am Sep 1, 2010 8:28am
I have been trying to import additional data into a MT platform. I have done it before and didn't have any problems, but for some reason, today it won't import. Here is what I have done and perhaps one of you have some suggestions that I could try that I have overlooked. I will try to be as detailed as possible with my explanations.

What lead me to do this in the first place, is I have a lot of problems with 'Old Tick Data;errors showing up on different EAs I try to run. I loaded Steve's version on Monday morning using 300,000 bars and up until this morning, there were 3 trading opportunities, but with Old Tick Data errors, so the trades never got placed. I decided to try the EA on another platform but I needed to import additional data.


I am trying to import the data to IBFX, although I have also tried this morning to import to FXDD and the same problem exists. I also have allowed for the time shift in FXDD. None needed in IBFX since they are GMT.

RR posted a site to d/l data from and I believe its Forex Tester. I went there and d/l to a folder on my computer. After, I opened the zip files and placed the txt files in a folder.

I set up a new account on IBFX, opened a M1 E/U chart, went to Tools > History Center. I select the currency and M1 for the time frame.

Next I clicked the 'Import' button. At the top of the box a line appears that says File. I enter the file that I have stored the data on my computer, making sure the file corresponds with the currency I have selected.

There is a large amount of space at the bottom of the box, where the opened file USUALLY appears. This is where it is not happening. The space stays blank. It never shows the open txt file.

I read a tip earlier today that someone said you have to go to Tools> Options> Charts and enter 999999999 in the Max Files in History. I did that. Didn't help.

I have 20 IBFX demo files on my computer and I opened # 21. That didn't help. It still won't import.
Tried going back to the Forex tester site and I re-d/l a couple of files and that didn't help.

I tried every one of the 21 demo platforms and none of them will import the data.

I've been working on this problem this morning close to 4 hours now and I'm starting to get a wee bit frustrated. LOL

TIA.

Carol
  • #895
  • Sep 1, 2010 10:44am Sep 1, 2010 10:44am
Quoting Eureka
I have been trying to import additional data into a MT platform. I have done it before and didn't have any problems, but for some reason, today it won't import. Here is what I have done and perhaps one of you have some suggestions that I could try that I have overlooked. I will try to be as detailed as possible with my explanations.

What lead me to do this in the first place, is I have a lot of problems with 'Old Tick Data;errors showing up on different EAs I try to run. I loaded Steve's version on Monday morning using 300,000 bars and up until...

The first thing I would check is how much ram you are using when the chart is loaded up. If you are close to maxing out then you should lower the "max bars on chart" (just below the "max bars in history" in the options tab) I say this because your computer allocates ram to the bars on your charts. If you dont have enough ram to operate then it wont load your history correctly.

Hope this helps.
  • #896
  • Sep 1, 2010 11:45am Sep 1, 2010 11:45am
I just bought a new computer about 2 months ago and it's an Intel Core 2 Quad 2.66 Ghz with 4GB of RAM.

A few minutes ago, I reloaded the platform that I was getting the Old Tick Data on and after it optimized, I have right at 2 GB of RAM available.

The platform has 300,000 bars in the History and 300,000 bars on the Chart. I have the EA placed on 10 charts in that platform.

When I was trying to import data, I had closed everything down but the platform I was trying to import the data to and my AOL software.

Thanks for your help.

Carol
  • #897
  • Sep 2, 2010 5:09am Sep 2, 2010 5:09am
Hi

Is it possible to modify the EA to save the .csv files in other folder??
(e.g. MT4/experts/files/HiLoTrader/ *.csv)

Regards
  • #898
  • Sep 2, 2010 10:26am Sep 2, 2010 10:26am
Anyone getting this error in the Experts folder?

HiLo Self Optimizing [TD] stevegee58 EURGBPm,M1:EURGBP Error:>!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"> <html> <head>...

The above error is happening on each currency I have the EA attached to.

Thanks.

Carol
  • #899
  • Sep 2, 2010 2:25pm Sep 2, 2010 2:25pm
I had to take down the tick database. I'm trying to setup a backup database tonight. Not sure how long it will take.

Those errors will affect all of the [TD] EAs. Simplest way to disable them is to look for the line in the code where you see "!IsOptimization()" or "!IsTesting()" and comment it out. Some EAs have TickDatabase as an input and you can just set that to false.
  • #900
  • Edited 9:21pm Sep 2, 2010 8:47pm | Edited 9:21pm
Ronald,
if you don't mind, please publishing one of your result on mt4i.com, it would be a great help to view and analyze the profit factors of certain pairs

thank you in advance.
Josh
0 traders viewing now