Steve Hopwood's Boris Schlossberg-inspired trading robots | Page 13 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Know Yourself, Know Your Setup, by Boris Schlossberg 9 replies

Mechanical day trading inspired by First Strike Plus 2 replies

Which movie has inspired you the most in your trading? 12 replies

New Webinar with Kathy Lien, Boris Schlossberg and Ken Calhoun 2 replies

Lien-Schlossberg book, Hayden:The RSI - any opinion? 1 reply

Attachments: Steve Hopwood's Boris Schlossberg-inspired trading robots
Exit Attachments

Steve Hopwood's Boris Schlossberg-inspired trading robots

  • #241
  • Feb 24, 2010 8:01am Feb 24, 2010 8:01am
Quoting sundog
Yes, I think we need to wait for the bar to close. .and a rest feature.
That makes sense if Steve can do that
  • #242
  • Feb 24, 2010 8:07am Feb 24, 2010 8:07am
Quoting sundog
Yes, I think we need to wait for the bar to close. .and a rest feature.
HI Guys

Mr SCHLOSSBERG says (and I Quote) " Wait for price to cross above the 20-period EMA, make sure that MACD is either in the process of crossing from negative to positive or have crossed into positive territory no longer than 5 bars ago"

Maybe we take what he says to literally. Anyhow attached is the statement of the last two days, as you can see we will have plenty of data to mess around with by the end of the week.

Kind Regards
P.S. Have 17 pairs running
Attached File(s)
File Type: xls forex statement.xls   56 KB | 255 downloads
  • #243
  • Feb 24, 2010 2:36pm Feb 24, 2010 2:36pm
Coding a wait until a new candle, (then a further waiting period if needed) is easy. Decide what you want and let me know.

Cheers

  • #244
  • Feb 24, 2010 5:08pm Feb 24, 2010 5:08pm
Quoting SteveHopwood
Coding a wait until a new candle, (then a further waiting period if needed) is easy. Decide what you want and let me know.

Cheers

I Do not know if it is an improvement or not. It seems schlossberg is not too worried about candle formations. I just do not know, maybe Scoobs, Macman or someone more knowledgeable can give us an opinion. For myself sometime I think the entry should happen sooner rather than later. Sorry can not be of much use.

Kind regards
  • #245
  • Feb 25, 2010 2:04am Feb 25, 2010 2:04am
Quoting astral77
I Do not know if it is an improvement or not. It seems schlossberg is not too worried about candle formations. I just do not know, maybe Scoobs, Macman or someone more knowledgeable can give us an opinion. For myself sometime I think the entry should happen sooner rather than later. Sorry can not be of much use.

Kind regards
Astral, thanks for that statement, interesting reading.

"my take" on it is we can do it either one of two ways.

1: Wait for the candle to close

2: If the price is triggered then have a pending order set for a better entry.


From watching my pair run, I really think we should wait for the candle to close, as what happened to me could happen fairly frequently, where a candle in the process of forming spikes into the target, but closes as a doji which means you're caught on the wrong side of the trade.

If you go #2, and you set a pending order but the candle is still a doji you just get a better entry on the wrong side of the trade.

I still think we need the candle to close before taking a trade.

This is just my opinion, as subject of debate.
  • #246
  • Feb 25, 2010 2:24am Feb 25, 2010 2:24am
Quoting sundog
Astral, ...From watching my pair run, I really think we should wait for the candle to close, as what happened to me could happen fairly frequently, where a candle in the process of forming spikes into the target, but closes as a doji which means you're caught on the wrong side of the trade.

If you go #2, and you set a pending order but the candle is still a doji you just get a better entry on the wrong side of the trade.
Thanks SUNDOG. You may have a point, when testing the 2SMA MACD EA using 1H Charts I cam across a sort of 200pip doji, which describes the situation you wish to avoid. I just looked at the results of the last 18 hours of 5min MOMO and 2SMA MACD. A lot of xxxJPY trades and 5min MOMO did not do so well, whereas 2SMA MACD made a number of very good trades. I do not think we should take what is happening with xxxJPY pairs into account due to the unusual movement (Guppy down nearly 400 and others taking a dive as well). It seems BOJ thinks the YEN is undervalued is letting it rise for the time being.

Kind Regards
  • #247
  • Feb 25, 2010 3:39am Feb 25, 2010 3:39am
Quoting sundog
Yes, I think we need to wait for the bar to close. .and a rest feature.
Hi Guys,
Interesting discussion...I'm still in favor of the bar closing and seeing that 2nd bar on the MACD before entry if everything else is correct. Sleep function, pending order all OK. The 2sma macd bot has a much better entry but it also is on a H1 chart, or in my case a D1. Later entry is a safer entry on a 5 min chart.

whip
  • #248
  • Feb 25, 2010 4:12am Feb 25, 2010 4:12am
Quoting hiredwhip
Hi Guys,
Interesting discussion...I'm still in favor of the bar closing and seeing that 2nd bar on the MACD before entry if everything else is correct. Sleep function, pending order all OK. The 2sma macd bot has a much better entry but it also is on a H1 chart, or in my case a D1. Later entry is a safer entry on a 5 min chart.
whip
HI

Thanks for the input. It seems us 3 are the only people doing the testing and as I stated I am neutral, therefore, I will abide with the decision of the majority. Let`s do it and see how it goes. At some point in future we should be thinking about the S/L rules. Having looked at figures at present the value of losing trades are 3 or 4 times higher than winning trades. If for this BOT to have a future some sort of solution needs to be found in order to rectify this problem.

Anyhow lets sort out the entry then worry about other aspects. STEVE can you be kind enough to change the code for the BAR CLOSING.

Kind regards
  • #249
  • Feb 25, 2010 4:38am Feb 25, 2010 4:38am
Quoting astral77
HI

Thanks for the input. It seems us 3 are the only people doing the testing and as I stated I am neutral, therefore, I will abide with the decision of the majority. Let`s do it and see how it goes. At some point in future we should be thinking about the S/L rules. Having looked at figures at present the value of losing trades are 3 or 4 times higher than winning trades. If for this BOT to have a future some sort of solution needs to be found in order to rectify this problem.

Anyhow lets sort out the entry then worry about other aspects....
OK, so we move forward a step.

Astral it would be very interesting to keep that excel sheet you provided and use it as a comparison to the next version of the Bot. It would shine a light on whether we are moving in the right direction with our decisions.

As to the SL/TP I agree with you, for now I think I will move my SL to aggressive and see if there is any noticable difference in the results.

Regards
  • #250
  • Feb 25, 2010 4:41am Feb 25, 2010 4:41am
Also if we're going to ask Steve to amend his Bot, could I suggest we get the pending and the sleep time done at the same time. So that we bother him less.

Thanks
  • #251
  • Feb 25, 2010 4:49am Feb 25, 2010 4:49am
Quoting sundog
OK, so we move forward a step.

Astral it would be very interesting to keep that excel sheet you provided and use it as a comparison to the next version of the Bot. It would shine a light on whether we are moving in the right direction with our decisions.

As to the SL/TP I agree with you, for now I think I will move my SL to aggressive and see if there is any noticable difference in the results.

Regards
Thanks very much. I will keep it (the copy here in FF is saved as well). Good idea to compare. A good point with this system is that due to the low TF of 5min we get plenty of examples to analyze.

I just hope STEVE does not produce any new BOTS for at least 3 to 4 weeks so we get a chance to test the ones on offer. At present I am testing The Holy Grail by Scoobs, 5min Momo, 2SMA MACD and Trend Retracement EAs, using my VPS and two other PCs. The problem is not the RAM but the CPU. Started getting non responsive mt4 platforms on my VPS so had to cut back abit.

Kind regards
P.S. Just saw your latest post regarding pending order. Good IDea.
  • #252
  • Feb 25, 2010 5:19am Feb 25, 2010 5:19am
Quoting astral77
Thanks very much. I will keep it (the copy here in FF is saved as well). Good idea to compare. A good point with this system is that due to the low TF of 5min we get plenty of examples to analyze.

I just hope STEVE does not produce any new BOTS for at least 3 to 4 weeks so we get a chance to test the ones on offer. At present I am testing The Holy Grail by Scoobs, 5min Momo, 2SMA MACD and Trend Retracement EAs, using my VPS and two other PCs. The problem is not the RAM but the CPU. Started getting non responsive mt4 platforms on my VPS...
yes I have the same problems with the CPU due to the number of bots im testing, as well as my live trading.
  • #253
  • Feb 25, 2010 8:35am Feb 25, 2010 8:35am
Hi Steve,
Got a question....Where do we stand on the combo bot? I know you're about ridden into the ground and put up wet, but did I miss the second position exit fix or have you just been to busy to get to it.

Astral mentioned it in one of his posts and jogged my memory and I went looking back through the posts and the one in the zip file has the same date as the one I'm using.

If I have'nt been paying attention I'll just have to take the beating....

whip
  • #254
  • Feb 25, 2010 3:17pm Feb 25, 2010 3:17pm
Quoting hiredwhip
Hi Steve,
Got a question....Where do we stand on the combo bot? I know you're about ridden into the ground and put up wet, but did I miss the second position exit fix or have you just been to busy to get to it.
Forgotten about it and cannot remember anything about it. Refresh my memory and I will have a look.

As you guys are seriously testing the robots you wanted me to code inthe first place, I am only too happy to help with updates. Sing out with what you want. After all, if you arrive at something successful, I will benefit as well.

Cheers

  • #255
  • Feb 25, 2010 3:30pm Feb 25, 2010 3:30pm
Quoting SteveHopwood
Nice bit of boobo spotting here.

I looked at the code. In English, and using a buy trade as an example, what I meant to code was: if the Ask is < sma 50 - 10(0) and the trade is in profit, then close it, because there must have been a profitable move followed by a retrace.

What I actually coded was: if the Ask is < sma50 and the Ask is < sma 50 - 10(0), then close the trade. This is an obvious bit of mis-coding that led to the 2nd half exit routine is engaged and is becoming a very short SL on a retracing pair.

I have fixed this, and will...
Hi Steve,
Post #195 is your ideas on how to solve the problem of early stopping out of the bot's trades by the second position exit stop. It also has post numbers of some symptoms of the problem that I had that I posted about.

whip
  • #256
  • Feb 25, 2010 3:37pm Feb 25, 2010 3:37pm
Quoting hiredwhip
Hi Steve,
Post #195 is your ideas on how to solve the problem of early stopping out of the bot's trades by the second position exit stop. It also has post numbers of some symptoms of the problem that I had that I posted about.

whip
Gimme a break, whip. How many back-posts with links to other posts do you want to plough through?

What is the problem and what do you want me to do about it?

  • #257
  • Feb 25, 2010 5:54pm Feb 25, 2010 5:54pm
Quoting SteveHopwood
Gimme a break, whip. How many back-posts with links to other posts do you want to plough through?

What is the problem and what do you want me to do about it?

Steve this is the problem and the solution that you posted....Thankz whip


Hi Steve,
Today I ran into a problem with the bot that I've had twice before. Today it was a long trade EURCHF, perfect entry on the 2nd bar of Swingman's great colored macd true indi. Started up and then retraced on the 2nd candle back over the 50 sma and was stopped out long before the SL.
I think the 2nd half exit routine is engaged and is becoming a very short SL on a retracing pair.
If my logic is correct, and I sure hope you set me straight if it isn't (I'm sure you will) can that routine be held off until the 2nd half of the position is in place.


Nice bit of boobo spotting here.

I looked at the code. In English, and using a buy trade as an example, what I meant to code was: if the Ask is < sma 50 - 10(0) and the trade is in profit, then close it, because there must have been a profitable move followed by a retrace.

What I actually coded was: if the Ask is < sma50 and the Ask is < sma 50 - 10(0), then close the trade. This is an obvious bit of mis-coding that led to the 2nd half exit routine is engaged and is becoming a very short SL on a retracing pair.

I have fixed this, and will put it up in time for market opening next week.

The robot might need a third part of the conditional i.e. close the trade if (buy again):

1. it is in profit
2. the Ask is < sma 50 - 10(0)
3. the order lot size < Lot, indicating that half the trade has been closed

I am merely thinking aloud here; 3 clearly needs adding. Watch this space for the pre-market-opening updated version.

Thanks for the attention to detail. Frankly, I do not have time for it and rely on others to spot blunders like this one (or two). Cheers
  • #258
  • Feb 25, 2010 6:04pm Feb 25, 2010 6:04pm
Quoting hiredwhip
Steve this is the problem and the solution that you posted....Thankz...
Thanks whip. I am pretty sure I fixed this - I recall writing the code. Maybe I did not actually put up the fixed version. I cannot remember, and so will look into this.

Nag me if I do not respond over the weekend.

Cheers

  • #259
  • Feb 25, 2010 6:22pm Feb 25, 2010 6:22pm
Thanks for all your work Steve. Have a great day....

whip
  • #260
  • Feb 26, 2010 2:43am Feb 26, 2010 2:43am
I have had a look at the code; the half-close function is only called if the trade is in profit.

It looks as though I did not actually upload the fixed version into post 1.

It is there now, so sing out next week if it is still not working. It should be fine; it is the same code as the 5 min momo, which appears to be correct.

0 traders viewing now