Steve Hopwood's Epsilon Megastar fib auto-trading robot | Page 12 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Time dependent auto fib? 1 reply

Auto fib bar count 0 replies

My Daily Fib Trading Journal 2 replies

Daily hi/low fib trading 35 replies

Steve Trading Journal 3 replies

Attachments: Steve Hopwood's Epsilon Megastar fib auto-trading robot
Exit Attachments

Steve Hopwood's Epsilon Megastar fib auto-trading robot

  • #221
  • May 22, 2010 7:55am May 22, 2010 7:55am
The latest version is available from post 1. I have persuaded the htf fibs to display correctly.

They are intended to aid manual trading decisions. If people would care to discuss their use, we can think about how to incorporate them into the robot's trading decisions.

The robot looks for htf swings that encompass the ltf swings. Using ZigZag for the htf swings produced swings on the default medium tf of H1 that are too close to those of the trading fib. I solved this by using ZigZagMe; I am unsure how accurate this is.

  • #222
  • May 22, 2010 1:05pm May 22, 2010 1:05pm
Quoting SteveHopwood
I solved this by using ZigZagMe; I am unsure how accurate this is.

When this thread first started I was testing zigzagme and frequently got fib not yet valid when I could clearly see it. Does the bot drawing the fibs solve this problem? Or are we setting ourselves up to backtrack? I am going to start testing random zigzags to see if we can benefit from a different one.

With the bot drawing its own fibs are we able to put all pairs on one chart? I have a six core processor but 57 charts with EAs still slows me down a little.
  • #223
  • May 22, 2010 1:29pm May 22, 2010 1:29pm
Quoting rfking2
When this thread first started I was testing zigzagme and frequently got fib not yet valid when I could clearly see it. Does the bot drawing the fibs solve this problem?
ZigZagMe is too quick to mark a new swing, so a fib can never be validated by a retrace to the inner levels. This is why I use ZigZag for this purpose. The trading fib is chart tf dependant and defaults to colour yellow.

Quote
Or are we setting ourselves up to backtrack? I am going to start testing random zigzags to see if we can benefit from a different one.
No. I explained why I use Me for the higher tf fibs. They are intended as an aid for manual traders. The robot does not use the info they provide in its trading decisions.

Quote
With the bot drawing its own fibs are we able to put all pairs on one chart? I have a six core processor but 57 charts with EAs still slows me down a little.
No. This is not a multi-pair robot.

  • #224
  • May 22, 2010 8:17pm May 22, 2010 8:17pm
Hey people... just a heads up about spreads for all of you using Alpari. Just take a look at the frozen platform of this weekend. A demo account shows 6.4 pips on eurusd; 24.6 on gpbjpy!!!

My live account shows a spread of 15 pips on gbpusd, 16 pips on gbpjpy, and so on...

Did some earthquake hit Great Britain, or is just the crim?
  • #225
  • Edited May 23, 2010 1:13am May 22, 2010 11:18pm | Edited May 23, 2010 1:13am
Still finding the rare occurence of a trade not closing before atttempting to open an opposite order (on backtest). Perhaps no one has encountered this, but it raises the question in my mind whether there could be a rare confluence of conditions where the bot could do this. If not, perhpas it is a glitch in MT4/Strategy Tester. Below is a screen shot of a portion of the strategy tester report showing the odd occurence. (BTW, this does not occur running the original strategy on the bot).

Transactions-1.jpg
Attached Image (click to enlarge)
Click to Enlarge Name: Transactions-1.jpg Size: 47 KB
  • #226
  • May 23, 2010 3:26am May 23, 2010 3:26am
Quoting octomanoak
Still finding the rare occurence of a trade not closing before atttempting to open an opposite order (on backtest)....
Take a look through the journal tab on the strategy tester for any errors. This will help pinpoint where the problem is.
  • #227
  • May 23, 2010 5:23am May 23, 2010 5:23am
You guys might like to play with the attached swing-spotting indi. It picks up a swing very quickly; if the market quickly resumes its previous movement, the indi moves the arrow, so it might not be suitable for us.

Let's watch it for a few days and see what we think.

One of our contributors sent it to me on Friday to see if it is a candidate for a trading robot, so the attached bot does this. I include it mostly so that coders have an easy copy/paste of the required variables and iCustom call (void ReadSuperSignal()), but backtesters might like to play with it to see if they can find profitable settings. If it looks to have legs, I will put it in the swing thread.

The robot's trading strategy:

 

  1. Use Rsi D1 to ascertain the trend a la EM
  2. Only trade when current Rsi is also on the right side of the D1 i.e. lower in a down trend etc
  3. Wait for a strong signal in the direction of the trend
  4. tp/sl 100(0) pips


CandleShift tells the robot which candle to examine for a signal:

 

  1. 0 = current
  2. 1 = previous
  3. 2 = 2 candles ago

Any value >0 means the robot will only look for trading ops at the opening of each candle.

Don't ask me what the indi inputs mean; not a clue.

Attached File(s)
File Type: mq4 Super_Signal_v3f_-_fixed_dista.mq4   6 KB | 1,095 downloads
File Type: mq4 Steve Hopwood's Super_Signal auto trading robot.mq4   11 KB | 1,008 downloads
  • #228
  • May 23, 2010 2:27pm May 23, 2010 2:27pm
Quoting fxdrive
Take a look through the journal tab on the strategy tester for any errors. This will help pinpoint where the problem is.
There are no errors in the journal. Nothing to indicate a cause.
  • #229
  • May 23, 2010 5:33pm May 23, 2010 5:33pm
Quoting SteveHopwood
You guys might like to play with the attached swing-spotting indi. It picks up a swing very quickly; if the market quickly resumes its previous movement, the indi moves the arrow, so it might not be suitable for us...
Hi Steve

Knowing your feelings regarding exotic indicators I am surprised you have devised a Bot framework for this one already. Lets hope it proves worthy.

kind regards
  • #230
  • May 23, 2010 5:45pm May 23, 2010 5:45pm
Quoting astral77
Hi Steve

Knowing your feelings regarding exotic indicators I am surprised you have devised a Bot framework for this one already. Lets hope it proves worthy.

kind regards
I doubt it will, in its own right. I offer it only in the hope that it might prove useful. There are coders here who play with my code all the time, in the hope that a little tweak here or there might prove a winner. Some of them might like to play with the ss indi to see if it makes a better swing spotter than ZZ.

My antipathy towards custom indis turns out to have arisen from my inability to understand how to code calls to them and make sense of their buffers. I am starting to get on top of this now.

  • #231
  • May 24, 2010 5:19am May 24, 2010 5:19am
I have been running the new bot for 8 hours and in that time I have had 9 trades for 9 winners with one trade in progress showing a small gain at present. Total gain is $64 with .04 lots. It is too early to start getting excited but it sure looks interesting.

I am running 9 pairs on M15 charts. The bot is running on Steve's defaults and I have MPTM loaded with pretty much default settings. I will continue to run the bot and report again later.

Thanks Steve.

Karl
  • #232
  • May 24, 2010 1:49pm May 24, 2010 1:49pm
Quoting Middleastern
Hey people... just a heads up about spreads for all of you using Alpari. Just take a look at the frozen platform of this weekend. A demo account shows 6.4 pips on eurusd; 24.6 on gpbjpy!!!

My live account shows a spread of 15 pips on gbpusd, 16 pips on gbpjpy, and so on...

Did some earthquake hit Great Britain, or is just the crim?
Hi Middleastern,

For some unknown reason (at least to me), Alpari raise their variable spreads to crazy values immediately before the close on a Friday.

It's worth noting as well that if you backtest an EA over the weekend, the EA will use the latest spread (i.e. the very high one) for the backtests.

Cheers,
swnke2.
  • #233
  • May 24, 2010 4:04pm May 24, 2010 4:04pm
Many thanks to Steve for his bright ideas and sharing of these robots for us to test and configure.

One suggestion that I posted in the question thread but now that I have priviledges I will repost. This is concerning the reinforcement of when and in which direction to enter.

This Epsilon robot is using the RSI on a daily chart to determine if the entry is long or short. Why not apply the williams %R to it since the RSI is a trailing indicator of the williams which might be better reinforcement to determine when to enter. So if both the Williams %R and RSI is > 50 then it would be long and vice versa for short. But if the Williams %R is below the -50 between -50 and -80 but the RSI is still above the 50 then that would be a holdoff until it can determine the markets direction. I believe this will better (faster) determine if markets are retracing as to not hand the pips bagged already by continuously entering in a direction that has lost momentum.
  • #234
  • May 24, 2010 4:45pm May 24, 2010 4:45pm
Quoting obnoxious2
Many thanks to Steve for his bright ideas and sharing of these robots for us to test and configure.

One suggestion that I posted in the question thread but now that I have priviledges I will repost. This is concerning the reinforcement of when and in which direction to enter.

This Epsilon robot is using the RSI on a daily chart to determine if the entry is long or short. Why not apply the williams %R to it since the RSI is a trailing indicator of the williams which might be better reinforcement to determine when to enter. So if both the Williams...
I just Googled williams %R. What I read suggests that williams %R is another ob/os indi like Stoch and Rsi. Using both of them merely doubles up the indis showing the same information. There is a name for this, but I cannot remember what it is.

  • #235
  • May 24, 2010 5:24pm May 24, 2010 5:24pm
Quoting SteveHopwood
I just Googled williams %R. What I read suggests that williams %R is another ob/os indi like Stoch and Rsi. Using both of them merely doubles up the indis showing the same information. There is a name for this, but I cannot remember what it is.

Your are absolutely correct (have a look at this post for a visual example http://www.forexfactory.com/showpost...postcount=3611), there is no difference between Williams%R and Stoch. There are so many indicators that you can spend a life time studying them. When you say lagging indicators, I suppose they are all lagging to some extent since they are based on past events. Some are leading (see this artcile in investopedia http://www.investopedia.com/articles.../02/092402.asp) but I am not convinced and think the person who invents an indicator which shows the future would be richer than bill gates.

Kind Regards
  • #236
  • May 24, 2010 5:50pm May 24, 2010 5:50pm
Quoting astral77
Your are absolutely correct (have a look at this post for a visual example http://www.forexfactory.com/showpost...postcount=3611), there is no difference between Williams%R and Stoch. There are so many indicators that you can spend a life time studying them. When you say lagging indicators, I suppose they are all lagging to some extent since they are based on past events. Some are leading (see this artcile in investopedia http://www.investopedia.com/articles.../02/092402.asp)...
Hehe. We could indeed.

There are two common complaints about indicators:

 

  1. It is a lagging indicator

    1. Of course it bloody-well is. Every indicator relies on historical data and can only show what has happened in the past.

  2. It repaints - the worst crime of all, it seems.

    1. It bloody well better do. What the f*** use is an indicator that: shows an arrow; bleeps; does a lap dance; commits ritual hari-kari; when a certain market condition occurs, if it does not show by 'repainting' that said market conditions have disappeared.

I just love it when coders add "Non-Lag" to their indi names.

From now on:

 

  1. EM is Epsilon Megastar Non-Lag
  2. Trend is confirmed by:

    1. Rsi Non-Lag
    2. a clever combination of 15 Non-Lag Ma's.

  3. The EM robot will only work in combination with my patented non-lag, executively-predictive, scientifically-proven and peer-researched, neurally-networked Revolutionary Fibonacci-Drawing tool.

This tool is only available to those of you who can answer the question.

  • #237
  • May 24, 2010 6:28pm May 24, 2010 6:28pm
Quoting astral77
Your are absolutely correct (have a look at this post for a visual example http://www.forexfactory.com/showpost...postcount=3611), there is no difference between Williams%R and Stoch. There are so many indicators that you can spend a life time studying them. When you say lagging indicators, I suppose they are all lagging to some extent since they are based on past events. Some are leading (see this artcile in investopedia http://www.investopedia.com/articles.../02/092402.asp) but I am not convinced and think...
Lagging or leading is just a reference to how fast a trend direction is set. To put it simply, the RSI is weighted heavier than the Williams %R so therefore the Williams %R will set a trend direction faster. The compromise of setting a direction faster is that it is a weaker indicator so the RSI is there as reinforcement.
  • #238
  • May 24, 2010 7:13pm May 24, 2010 7:13pm
Quoting obnoxious2
Lagging or leading is just a reference to how fast a trend direction is set. To put it simply, the RSI is weighted heavier than the Williams %R so therefore the Williams %R will set a trend direction faster. The compromise of setting a direction faster is that it is a weaker indicator so the RSI is there as reinforcement.
There was a time I was collecting as many indicators as I could find and have a collection of hundreds. However, nothing beats S&R, trend lines, MA and RSI. Take the super signal indicator sent to Steve by a contributor, it is identical to a S&R indicator provided in another thread in FF.

Very good explanation of leading and lagging indicators is here http://www.investopedia.com/universi...r/ind_osc1.asp

kind regards
  • #239
  • May 24, 2010 7:27pm May 24, 2010 7:27pm
Hey Steve,

Nice work with Super Signal bot. Has the robot been enabled for live accounts ?

Thanks buddy,

getdown71
  • #240
  • May 25, 2010 12:42am May 25, 2010 12:42am
For those just testing the robot without messing with the settings, just a headsup that the RSI period on the robot is set to 21 instead of the 14 mentioned in the pdf.
0 traders viewing now