Steve Hopwood's Fib Target-level retracement auto-trader: compliments Simone | Page 26 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Monster Muncher Manual Trading Discussion: Compliments SH 228 replies

Fib retracement line colors 0 replies

Fib Retracement 13 replies

Fib Retracement 2 replies

Fib Retracement Tool 1 reply

Attachments: Steve Hopwood's Fib Target-level retracement auto-trader: compliments Simone
Exit Attachments

Steve Hopwood's Fib Target-level retracement auto-trader: compliments Simone

  • #501
  • Aug 18, 2010 9:42am Aug 18, 2010 9:42am
Quoting hiredwhip
I'm running FTLR on 23 D1 charts....No trend filters with anywhere from 15% to 25% drawdown. I don't have a negative profit number yet this month. Very slow but about as smooth and steady as you could ask for.
After upgrading to the latest version it will go back on a live account in two weeks if everything holds true.
I take naps while I'm watching it sitting there thinking. MPTM is very important for closing out trades.....

whip
Whip,

Why do you think using MPTM is important? What have you observed when running FTLR without it?

I'm running with the RSI filter and so far all profitable trades with drawdown only as much as 4%. I presume the longer I run it I will see higher drawdown, but does using MPTM really make a difference in the profitability or the drawdown over a longer period of time?
  • #502
  • Aug 18, 2010 2:21pm Aug 18, 2010 2:21pm
Quoting octomanoak
Whip,

Why do you think using MPTM is important? What have you observed when running FTLR without it?

I'm running with the RSI filter and so far all profitable trades with drawdown only as much as 4%. I presume the longer I run it I will see higher drawdown, but does using MPTM really make a difference in the profitability or the drawdown over a longer period of time?
Hi Octo,
It's been awhile....hope everything is well. I'm back in the house with rain storms again.....sorry not to answer this sooner, but I'm not checking the threads when I'm working like I was.
IMHO MPTM gives me much more control of the pairs in the profit zone because of it's flexability. I will use it just about any time I can. On EA's that don't use a hard TP I feel like I'm flying blind ( that's my old scalper's opinion ).
This bot is at 10% DD at the moment but using a 1000 pip SL will keep the DD alittle higher than that normally.
I only get a half dozen trades a week but at least one of those will be a good one.
Got to go back.....hope this helps

whip
  • #503
  • Aug 31, 2010 2:34pm Aug 31, 2010 2:34pm
Hi Steve,

One of my hay guys turned a baler over in the middle of my back hay field and I'm waiting for a repair crew so I'm in looking at the charts of course.....

OK Steve...Unless I get yelled at with big fonts....FTLR is going back on one of my penny accounts next week. I'm sending along all the info and it is running with the last update and as good as gold.
Hope everybody tries it out....to me it is in your top 5 all-time..

Have a great week...

whip
Attached File(s)
File Type: pdf Statement_ 1144067 - Hiredwhip.pdf   58 KB | 495 downloads
File Type: mq4 FTLR.mq4   35 KB | 417 downloads
File Type: txt mptm.set.txt   4 KB | 390 downloads
  • #504
  • Aug 31, 2010 5:13pm Aug 31, 2010 5:13pm
Quoting hiredwhip
Hi Steve,

One of my hay guys turned a baler over in the middle of my back hay field and I'm waiting for a repair crew so I'm in looking at the charts of course.....

OK Steve...Unless I get yelled at with big fonts....FTLR is going back on one of my penny accounts next week. I'm sending along all the info and it is running with the last update and as good as gold.
Hope everybody tries it out....to me it is in your top 5 all-time..

Have a great week...

whip
Cheers whip. As it happens, I was going to post about my own progress here some time this week. I am still testing with mixed success - H1, no trend filter, rising balance, rising dd. Closing the position at just the right time would be excellent, but not exactly useful when contemplating use of an auto-trader.

I am a loooooonnnnnggggg way from ditching this robot, so your post is a welcome one.

You still demoing the D1?

  • #505
  • Aug 31, 2010 5:18pm Aug 31, 2010 5:18pm
Quoting SteveHopwood
Cheers whip. As it happens, I was going to post about my own progress here some time this week. I am still testing with mixed success - H1, no trend filter, rising balance, rising dd. Closing the position at just the right time would be excellent, but not exactly useful when contemplating use of an auto-trader.

I am a loooooonnnnnggggg way from ditching this robot, so your post is a welcome one.

You still demoing the D1?

Hi Steve,
I'm still running 23 pairs on D1.....My kind of trading....Plenty of time for naps...This thing is a real gem....No loosing trades in over a month and that is with 2 different versions.

Thanks for the fun.....whip
  • #506
  • Aug 31, 2010 5:30pm Aug 31, 2010 5:30pm
Quoting hiredwhip
Hi Steve,
I'm still running 23 pairs on D1.....My kind of trading....Plenty of time for naps...This thing is a real gem....No loosing trades in over a month and that is with 2 different versions.

Thanks for the fun.....whip
As I recall, way back in the distant depths of time, I originally envisaged this as running on the D1.

One of the probs I noticed when demoing this was that the original version fibbed the Sunday candle on Monday, so I added code to make it fib the Friday candle. Then it appeared to stick on the Friday candle. Has this been a problem for you?

  • #507
  • Aug 31, 2010 5:36pm Aug 31, 2010 5:36pm
Quoting SteveHopwood
As I recall, way back in the distant depths of time, I originally envisaged this as running on the D1.

One of the probs I noticed when demoing this was that the original version fibbed the Sunday candle on Monday, so I added code to make it fib the Friday candle. Then it appeared to stick on the Friday candle. Has this been a problem for you?

If it has Steve I've never noticed it......But then you and I both know what kind of a robot expert I am......I saw macman reading here about an hour ago....I'd sure like to see him tackle this thing on a long term demo and try to break it

whip
  • #508
  • Aug 31, 2010 5:40pm Aug 31, 2010 5:40pm
You up for the challenge, macman?

  • #509
  • Edited 6:07pm Aug 31, 2010 5:56pm | Edited 6:07pm
Quoting SteveHopwood
You up for the challenge, macman?

Waaaay ahead of you guys - have been testing for a while ...

As you probably know I always start testing on M15, irrespective of what the end TF may be. This gives me a good idea of what to expect on higher TF's. Generally speaking a poor M15 performance won't get much better on a longer TF.

Here is the data from FTLR so far, running on my usual 6 test pairs at M15.

Pic 1 is the floating draw-down against profit trace for the whole test period so far.

Pic 2 is account performance over the last 4 days.

Very nice so far ....

I will come back with some suggestions later -
Attached Image(s) (click to enlarge)
Click to Enlarge Name: FTLR_profit_dd_trace.png Size: 32 KB
Click to Enlarge Name: FTLR_performance.png Size: 32 KB
  • #510
  • Aug 31, 2010 6:03pm Aug 31, 2010 6:03pm
Quote
As the problem of multiple-trade sends appears to be because the crap platform combined with criminal behaviour means that the platform/criminal's server/Satan does not recognise that a trade has been sent, I do not see how keeping a record of the ticket number of a trade that the platform/criminal's server/Satan does not recognise will help.
Because it does.

Ticket number is reliable immediately after order send but the order history data base is not. This is experimental physics, not logic.

Likely because the real time processing and clearing at the server's end works on unique transaction identifiers (ticket numbers), so it has to be "reliable". So it does recognize the order if you have a ticket number, but the data base for the orders is not fully populated.

And because MT4 isn't remotely a sophisticated platform with good RT capabilities (which would make it harder to program still).

For instance one way to do it properly is to ask for a lock from the server to guarantee you get a snapshot of the state---and then have to deal with exception handlers in case the lock failed, or timed out. That's how real programming works and it's a PITA.

I know you don't like my pimping my LibOrderReliable, but it does handle some of the issues (not the you-must-check-by-ticket-number business) and you can get two-stage order handling easily, and it will check the error conditions (e.g. trade context) and do proper randomized sleeping. I could add an additional (optional) check that after you send an order which is accepted the library could wait to see if it has populated in the order history library. (there is a tiny change the order could be accepted, take profit/stop loss, and closed before it shows up as open).

The goal of the library is that you should be able to write clean order code (e.g. something which works in backtester) and then have it pretty much work in real time.
  • #511
  • Aug 31, 2010 6:15pm Aug 31, 2010 6:15pm
Quoting mbkennel
I know you don't like my pimping my LibOrderReliable, but it does handle some of the issues (not the you-must-check-by-ticket-number business) and you can get two-stage order handling easily.
You are wrong. Not only do I not mind you pimping your lor, but I love it.

What I hate are the problems using libraries cause me as the thread runner; people are unable to cope with them. If you ever get to the point of running a few of your own threads, you will understand. Until then, you will just have to take my word for it; libraries are more trouble than they are worth, yours included. Sorry.

For this robot and indeed most others, sending it to sleep for a while after a successful trade send solves the only really important problem - that of repeated trade sends caused by the combination of crap platform and criminal intent.

I have found a different way to you of solving this problem. It may not suit you, but it sure suits me.

  • #512
  • Aug 31, 2010 9:30pm Aug 31, 2010 9:30pm
I have been running the new version on FXDD demo for 2 weeks:

D1, 14 pairs, RSI=true, no MPTM

30 closed profitable trades
No losing trades
Maximum negative UPL less than 10%.

Seems to be very stable.

Great work Steve!
  • #513
  • Sep 1, 2010 8:32am Sep 1, 2010 8:32am
Fantastic JR.

  • #514
  • Sep 1, 2010 5:23pm Sep 1, 2010 5:23pm
I just thought in case we get a runaway trade it might be a better option in having a recovery system as per the NanningbobV5 robot instead of a straight stop loss for the FTLR robot.
Obviously the idea is once the trade has gone beyond the L4 level we could easily send a second or third trade and just try to get break even before looking for a new set up.
Does anybody think that this is possible to code into the FTLR robot?

Cheers
  • #515
  • Sep 1, 2010 5:26pm Sep 1, 2010 5:26pm
Quoting Pipstrumpf
I just thought in case we get a runaway trade it might be a better option in having a recovery system as per the NanningbobV5 robot instead of a straight stop loss for the FTLR robot.
Obviously the idea is once the trade has gone beyond the L4 level we could easily send a second or third trade and just try to get break even before looking for a new set up.
Does anybody think that this is possible to code into the FTLR robot?

Cheers
This sounds like a great idea to me. Any further thoughts, guys?

  • #516
  • Sep 1, 2010 5:28pm Sep 1, 2010 5:28pm
Quoting SteveHopwood
This sounds like a great idea to me. Any further thoughts, guys?

Yes, a good idea IMHO.
I am doing the same, only manually trading a recovery to BE -
  • #517
  • Sep 1, 2010 5:30pm Sep 1, 2010 5:30pm
Quoting SteveHopwood
This sounds like a great idea to me. Any further thoughts, guys?

It could only help, anything is better than a flat SL....

whip
  • #518
  • Sep 1, 2010 5:33pm Sep 1, 2010 5:33pm
Quoting hiredwhip
It could only help, anything is better than a flat SL....

whip
Instead of losing a trade to a flat S/L, I would rather add a zero hedge and wait for the market to retrace, then work on getting out around BE.

Once your S/L is hit, your money has gone ...
  • #519
  • Sep 1, 2010 5:36pm Sep 1, 2010 5:36pm
Quoting macman
Instead of losing a trade to a flat S/L, I would rather add a zero hedge and wait for the market to retrace, then work on getting out around BE.

Once your S/L is hit, your money has gone ...
Agree.....NBob has proven that time and again.......

whip
  • #520
  • Sep 1, 2010 5:48pm Sep 1, 2010 5:48pm
Fabulous gentlemen. I need no further encouragement.

macman, any suggested levels and lot sizes?

1 trader viewing now