Inside Bar Trading Revisited | Page 10 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

3 Bar Fractal with mid-bar and bar on each side 11 replies

Inside bar or outside bar identifier 7 replies

Journal related to: Advanced Inside Bar (IB) Trading 19 replies

Pin bar, pin bar, pin bar 0 replies

Attachments: Inside Bar Trading Revisited
Exit Attachments

Inside Bar Trading Revisited

  • #181
  • Jan 21, 2009 3:08pm Jan 21, 2009 3:08pm
Noticed a subtle bug that may stop the EA taking a trade when there's another bar that almost crosses the IB threshold. Posted the first chart after this fix. I'll post an updated EA in the first post.

Changed the test code to close an IB early if a second IB forms and it's crossed in the opposite trend direction (as discussed in an earlier post). This is shown in the second chart. The idea was that this would help reduce any major losses when trends go against us. Sort of works, but not enough for me to want to put it into the live code.
Attached Image(s) (click to enlarge)
Click to Enlarge Name: Picture 3.png Size: 111 KB
Click to Enlarge Name: Picture 4.png Size: 108 KB
  • #182
  • Jan 21, 2009 3:37pm Jan 21, 2009 3:37pm
Today's results. Slightly better
Attached Image (click to enlarge)
Click to Enlarge Name: Picture 5.png Size: 73 KB
  • #183
  • Jan 21, 2009 4:10pm Jan 21, 2009 4:10pm
Quoting 1ka
The ATR is used to determine the TP. With larger ATR's we average out the size of the bars, so the EA is not affected by any spikes in the bar size, so it will generally TP much sooner. I ran a few tests on EURUSD 30M charts and found that 40 was a good figure for those. (Look back a few posts)

Hmm, I'm not sure that the ATR in MetaTrader works correctly. I'd expect to see the ATR average out at higher periods. But in MetaTrader it still seems as susceptible to large bars at higher periods. The yellow line is 140 period.

There's a definition of ATR here -> http://stockcharts.com/school/doku.p...true_range_atr

Can someone else verify this with their setup (non Metatrader)?

Chart is USDJPY M30 for 2009/01/21.
Attached Image (click to enlarge)
Click to Enlarge Name: Picture 6.png Size: 26 KB
  • #184
  • Jan 21, 2009 4:32pm Jan 21, 2009 4:32pm
Ika,

Thanks for the great thread. I trade with GFT and the ATR is different to your platform. My platform for USD/JPY 30min ATR(20) = 0.3115 and ATR(140) = 0.2140

Regards
  • #185
  • Jan 21, 2009 7:12pm Jan 21, 2009 7:12pm
Quote
Thanks for the great thread. I trade with GFT and the ATR is different to your platform. My platform for USD/JPY 30min ATR(20) = 0.3115 and ATR(140) = 0.2140

These numbers make more sense for a true ATR as it is defined . Metatrader's ATR must be rear end weighted in order to climb as fast as it did
on the U/J chart
  • #186
  • Jan 22, 2009 12:49am Jan 22, 2009 12:49am
Garson, Abnormalone,

Thanks for the confirmation. ATR in MetaTrader does appear to be something different to the other platforms, and to the definition of ATR. As Abnormalone says, it must be a Weighted ATR or WATR

I think I'll code you my own version and see what the differences are. I want to use ATR so it averages out these large trending bars.
  • #187
  • Jan 22, 2009 1:24am Jan 22, 2009 1:24am
1ka,FWIW here are my M30 E/U ATR values at close of NY sesion on 1/21/09. I do not use MT4. ATR 20,.0044, ATR 40,.0040, ATR 80,.0038, ATR 160,.0036, ATR 320,.0035.
Vincit qui se vincit.
  • #188
  • Jan 22, 2009 9:23am Jan 22, 2009 9:23am
Let me give you my 1 cent.
I've been trading this method successfully for 5 months.
I trade Inside Bar (IB) and Containing Bar (CB) only if the CB is touching Pivot line or Support/Resistance line, and only 1 of them.
I.e: If the CB touches Pivot line and S1 (or R1) at same time, it's illegal.

I know it means that I don't trade many optional trades but the RIsk:Reward is much higher

* I also trade IB's which are before the CB and not only after it.

See attachment.
Attached Image (click to enlarge)
Click to Enlarge Name: ib example.gif Size: 11 KB
  • #189
  • Jan 22, 2009 9:41am Jan 22, 2009 9:41am
Quoting smilee89
Let me give you my 1 cent.
I've been trading this method successfully for 5 months.
I trade Inside Bar (IB) and Containing Bar (CB) only if the CB is touching Pivot line or Support/Resistance line, and only 1 of them.
I.e: If the CB touches Pivot line and S1 (or R1) at same time, it's illegal.

I know it means that I don't trade many optional trades but the RIsk:Reward is much higher

* I also trade IB's which are before the CB and not only after it.

See attachment.
Thanks Smilee89,

Do you take the break if it goes either way (short/long), or do you have a rule that determines which breaks to take?

I'm about to incorporate pivots/support/resistance into the code and this could form part of the 'rules'.
  • #190
  • Jan 22, 2009 10:59am Jan 22, 2009 10:59am
Quoting 1ka
Thanks Smilee89,

Do you take the break if it goes either way (short/long), or do you have a rule that determines which breaks to take? Yes, both long and short. I trade the break of the CB (and not the IB) so the break must be in both sides of the pivot line (or S&R). So if long - above the pivot, if short - below it.


I'm about to incorporate pivots/support/resistance into the code and this could form part of the 'rules'.
Good Idea (-:
Remember that Pivot Lines are determined by the end of the day of the broker time and it can be different between each broker.
  • #191
  • Jan 22, 2009 11:57am Jan 22, 2009 11:57am
Quoting 1ka
Garson, Abnormalone,

Thanks for the confirmation. ATR in MetaTrader does appear to be something different to the other platforms, and to the definition of ATR. As Abnormalone says, it must be a Weighted ATR or WATR

I think I'll code you my own version and see what the differences are. I want to use ATR so it averages out these large trending bars.
Having coded my own ATR, then I can say that there's nothing wrong with MT4's ATR. It really does get affected by a large bar to that degree.

Plan B.
  • #192
  • Jan 22, 2009 4:14pm Jan 22, 2009 4:14pm
Ouch, got hammered in the morning.
Attached Image (click to enlarge)
Click to Enlarge Name: Picture 1.png Size: 74 KB
  • #193
  • Jan 22, 2009 7:27pm Jan 22, 2009 7:27pm
I did a small study on how only going long when above Pivot point and short when below Pivot point point would come out. I used the trades the strategy tester produced and a visual from Pivot Point on a chart to separate the trades and the results were positive.
Strategy tester 117 trades 96 winners 82% -$251.86
Because strategy tester double counts winning trades this was really 71 trades 48 winners 67% Results with Pivot Points added were 32 trades 24 winners 75% -$1.20. There was one trade that should not have been included because it did not go offset plus spread above setup bar. This would add $78.00 to each total. These results were for the period 1/1/2009 to present. The EA used Max bars 9 Bar offset 0 Min ATR 10
  • #194
  • Jan 23, 2009 1:45am Jan 23, 2009 1:45am
Quoting Abnormalone
I did a small study on how only going long when above Pivot point and short when below Pivot point point would come out.
Thanks Abnormalone.

I'm still in two minds to try this, and s/r levels in general, but your analysis seems to indicate it might be worth the effort. Smilee89 would agree with you, I think.

If I get time over the weekend then I'll give it a go.
  • #195
  • Jan 23, 2009 2:33am Jan 23, 2009 2:33am
Important

Calculating daily pivot is accurate by doing it when midnight broker's time is at GMT+2.
All other broker's time are less accurate. This because GMT+2 is when NY session is closed.
  • #196
  • Jan 23, 2009 4:35am Jan 23, 2009 4:35am
Quoting smilee89
Important

Calculating daily pivot is accurate by doing it when midnight broker's time is at GMT+2.
All other broker's time are less accurate. This because GMT+2 is when NY session is closed.

Thanks Smilee89,

Do you know of any example indicators that calculate these? Saves me having to code something from scratch.
  • #197
  • Jan 23, 2009 5:54am Jan 23, 2009 5:54am
In fact, I have this indicator that calculates pivot by the time you enter (for example 23:00 instead of 24:00), but I prefer to calculate the pivot with a simple pivot indicator on a chart that its broker ends the day on GMT+2.

FXDD is a good sample. It ends on GMT+2. I dont have an account there but I only analyze with this broker.
  • #198
  • Edited 8:24am Jan 23, 2009 7:54am | Edited 8:24am
Quote
I'm still in two minds to try this, and s/r levels in general, but your analysis seems to indicate it might be worth the effort. Smilee89 would agree with you, I think.
My Back test was flawed to some degree because I only used the same trades the strategy tester made. In real time if you eliminate one trade then the next trade may have come on a different IB and really changed the results drastically.
  • #199
  • Jan 23, 2009 1:47pm Jan 23, 2009 1:47pm
Better run today. Stopped the EAs at 19:45 GMT. Some initial losers from last night that may have won at 1/2 TP.

Added the weekly view too.
Attached Image(s) (click to enlarge)
Click to Enlarge Name: Picture 2.png Size: 73 KB
Click to Enlarge Name: Picture 3.png Size: 76 KB
  • #200
  • Jan 23, 2009 2:48pm Jan 23, 2009 2:48pm
In profit, by the end of week; It is great thing

1ka have a extra cold Guinness beer....

Anto
2 traders viewing now