Try My EA! Need Group Feedback... | Page 38 | Crypto Craft

Options

Go to
Permalinks
Print Thread

Related Instruments

Gold/USD

Similar Threads

USD/TRY and EUR/TRY 12 replies

Fibo Group Feedback 23 replies

Need feedback on my trading breakouts system 8 replies

One World Capital... Need feedback 2 replies

Need feedback on FXCM Educational workshops 3 replies

Attachments: Try My EA! Need Group Feedback...
Exit Attachments

Try My EA! Need Group Feedback...

  • #741
  • Jan 3, 2015 3:14pm Jan 3, 2015 3:14pm
Quote
Not to mention the dynamic grid - when it arrives, it will be pure fun to watch!

I'm thinking along the lines of an auto-grid. When you start it the grid will build to one step beyond your target, and fill in when the target moves due to hedging. That way there won't be nearly so many trades to close and it should be much more nimble No need to declare how many steps, it will be on auto.
A good man knows his limitations ~ Clint Eastwood
  • #742
  • Jan 3, 2015 3:20pm Jan 3, 2015 3:20pm
Quoting dambuster
Another thought, weekend gaps where you might have the following scenario: Friday night close You entered at $1995 with $0.50 gridsteps earlier in the day so you have a grid going from $1980 to $1210 Price had moved from $1995 to $1998 and you picked up 6 buys Price had also moved from $1995 to $1993 so you have 4 sells Market closes at $1997 Market opens with a gap down to $1985 and 16 sell stops all get triggered at that price Market then reverses back up Ouch...
Yeah, and another thing about that: Watch out around holidays! Ikon opened at 1AM GMT, while FXCM didn't open until 6AM GMT on Friday, with the market already moving for hours! Talk about getting hosed on the gap! FXCM showed a 50 pip gap on EURUSD while the other brokers had no gap at all.
A good man knows his limitations ~ Clint Eastwood
  • #743
  • Jan 3, 2015 3:35pm Jan 3, 2015 3:35pm
Hmmm, dynamic grid should reduce that risk dramatically
  • #744
  • Edited 4:11pm Jan 3, 2015 3:45pm | Edited 4:11pm
Yeah, a gap without the buy or sell stops present would just adjust the target and fill in the possible grid from the current point. It might get messy though.

Like, you have the hedge out to the target 6 steps away on the sell side ~$3 away and the market closes. When it opens it's $6 away and none of the trades that made the target possible at that level are executed, instead they all go off at the new price. Now you have a big hole in the grid $6 wide, and what goes there? What happens if the buys and sells overlap? Like you had 3 executed sells and the price gaps down leaving a hole. Do you fill in the hole with buystops to the new buy target, which will mean buystops below the open sell orders? Hmmmmm
A good man knows his limitations ~ Clint Eastwood
  • #745
  • Jan 3, 2015 5:49pm Jan 3, 2015 5:49pm
Quoting dgodfrey
Yeah, a gap without the buy or sell stops present would just adjust the target and fill in the possible grid from the current point. It might get messy though. Like, you have the hedge out to the target 6 steps away on the sell side ~$3 away and the market closes. When it opens it's $6 away and none of the trades that made the target possible at that level are executed, instead they all go off at the new price. Now you have a big hole in the grid $6 wide, and what goes there? What happens if the buys and sells overlap? Like you had 3 executed sells...
Wouldn't you have to add more pending sells as well in case it touches on those new buys and goes down further putting those buys into a loss position - I think without further pending sells - you wouldn't reach your profit target. I think we need to have the same amount of pending buys and sells to make this strategy work....not sure though! Just throwing it out there
  • #746
  • Edited 8:18pm Jan 3, 2015 6:40pm | Edited 8:18pm
Well, you only ever need as many as it takes to reach the target, but the target moves. Gaps, slippage and hedging will make it move, sometimes a lot. In that gap down, the slippage would push the sell target away, and the sell stops you mention would be put in place to reach the sell target (unless you got lucky and the new price put the gain beyond the new sell target value). The buy target would be pushed out as well, but perhaps to a different amount, and the buy stops would fill in the intervening gap. There will be a different number of them, but only enough to reach the long or short targets
A good man knows his limitations ~ Clint Eastwood
  • #747
  • Jan 3, 2015 7:54pm Jan 3, 2015 7:54pm
Quoting dgodfrey
{quote} I'm thinking along the lines of an auto-grid. When you start it the grid will build to one step beyond your target, and fill in when the target moves due to hedging. That way there won't be nearly so many trades to close and it should be much more nimble No need to declare how many steps, it will be on auto.
That would be a huge step forward. An intelligent grid!

Now that dambuster scared everyone with gaps (lol), we've got some serious thinking to do... maybe something along the lines of laying a fresh cycle over the one in trouble (but starting at the current price), so that the new one keeps beefing up the balance until the bad one is leveled out. It's adding more risk, as it'd burden the margin, but some of the open trades could actually get 'double hedged' that way... (?) Just brainstorming here.
mind over matter
  • #748
  • Jan 3, 2015 7:58pm Jan 3, 2015 7:58pm
Quoting dgodfrey
Mulling over the numbers today I came up with some analysis of what we're doing here. Just for your information.
Thanks for the analysis, dgodfrey. Very succinct and a great reference point for any setting adjustments.
mind over matter
  • #749
  • Jan 3, 2015 8:38pm Jan 3, 2015 8:38pm
Quoting kosmo
{quote} That would be a huge step forward. An intelligent grid! Now that dambuster scared everyone with gaps (lol), we've got some serious thinking to do... maybe something along the lines of laying a fresh cycle over the one in trouble (but starting at the current price), so that the new one keeps beefing up the balance until the bad one is leveled out. It's adding more risk, as it'd burden the margin, but some of the open trades could actually get 'double hedged' that way... (?) Just brainstorming here.
It would require a bigger account size for sure and how many cycles could you have open at one time? I dont think very many if you're the average Joe with a small account. Two, maybe three. If its already 0.01 per $1000.....
  • #750
  • Jan 3, 2015 10:26pm Jan 3, 2015 10:26pm
While working on the next version, I came across another potential issue.

There may be a conflict if you are running two instances of this program using the same broker on the same demo server. If you do and get trades closed down out of the blue, it could be that. The fix is easy but in the meantime, try not to set up your demos that way. Dont use the same demo server from the same broker to run more than one copy. Otherwise no troubles. May not be a problem anyway, but better safe than sorry.

I'm assuming there's no cowboys out there running two live versions from one broker on the same live server
A good man knows his limitations ~ Clint Eastwood
  • #751
  • Edited 7:58am Jan 4, 2015 2:27am | Edited 7:58am
Burning the late night oil, I guess I'll regret this in the morning

Anyway, here's a thought. Kosmo and I were thinking of catching this movement at likely pivots using indicators, but when I put up a Choros indi (which just puts horizontal lines on the screen at regular intervals), look what pops out at the $10 grid level on the 1H chart.

Attached Image (click to enlarge)
Click to Enlarge Name: XAUUSDH1.png Size: 47 KB


The idea would be to trigger a new grid at the $10 bands, like at the white dots on the chart. You can see how the price often blows through those or bounces off them completely. If you had a profit at that level, regardless of whether it was what your target was or not, close out and restart. I'm not saying to take a loss, if it's not profitable just keep going to the next band.

Going back to my earlier post about targets and exponential gain, a $10 move without hedge loss would be worth over 10% trading 0.1% per trade at $0.50 grid spacing. Surely 3% wouldn't be too much to ask then, assuming you did have some hedge loss to overcome. Like I said, anyway, I'd take it whatever the gain was off the table and reinitiate the grid at the $10 band when the price got there. Of course, the EA could handle all that if it was set up to do it.

By the way, here are the Choros files if you want them. The points are set up for 5 digit, so I think if you use 200 gridspace with 100 band it gives you the $10 spacing.
Attached File(s)
File Type: ex4 Choros 3.ex4   10 KB | 359 downloads
Attached File(s)
File Type: mq4 Choros 3.mq4   5 KB | 398 downloads
A good man knows his limitations ~ Clint Eastwood
  • #752
  • Jan 4, 2015 6:26am Jan 4, 2015 6:26am
Another thought, would closing all pending orders 5 mins before market close and then reopen them when the market opens avoid a possible black swan event?

The EA would need inputs such as DayOfClose, HourOfClose, MinuteOfClose and then similar imputs for market reopen

Just a 2 cents thought...
  • #753
  • Jan 4, 2015 7:46am Jan 4, 2015 7:46am
Thats already in there except for the day of close. The "UseTimeToTrade" function could be modified for that, But what would you do at the market open if it did the big gap? Deleting trades out of the gap manually would mess things up, but I guess it would be better than them all going off at once. I guess you would also need a variable for day of open as well.
A good man knows his limitations ~ Clint Eastwood
  • #754
  • Jan 4, 2015 7:57am Jan 4, 2015 7:57am
I don't think that the UseTimeToTrade would close pending orders?

So, the vraiables would need to be:

DayOfClose
HourOfClose
MinuteOfClose

To close all pending orders and then:

DayOfOpen
HourOfOpen
MinuteOfOpen

for it to reopen all previously closed pendings as long as the market hasn't gapped?
  • #755
  • Jan 4, 2015 8:17am Jan 4, 2015 8:17am
I see what you're saying now. I was thinking of it wrong. Yeah, that would work. Also, you see that big gap in the chart around Christmas? That's FXCM's feed through ATC. But they didn't open back up until 08:00 GMT on the 26th. Ikon's chart shows they opened at 01:00 GMT. And Ikon has a $3.50 gap while FXCM has an $11.50 gap. Nothing like making a bad situation worse.

More to the point, if you close out all the pendings and not the open orders, you are exposed to the unhedged trades. For each unhedged trade you had, assuming 0.1% on your trades as exposure, you would either make or lose 1.15% of your account on that FXCM gap. That could add up pretty quick. Maybe the prudent thing to do would be to hedge the unhedged trades and lift the hedges you put on when you open, assuming you had any. It would cost you the spread on as many trades as you hedged, but that's just like insurance.
A good man knows his limitations ~ Clint Eastwood
  • #756
  • Jan 4, 2015 8:31am Jan 4, 2015 8:31am
That sounds like a plan
  • #757
  • Jan 4, 2015 5:19pm Jan 4, 2015 5:19pm
Quoting dgodfrey
Here's the new release. Because of the issue of the arbitrary shutdowns and not being able to do normal things without messing it up, I would advise you all to try this version. It's my hope that all those stability problems are fixed. Please read the explanation carefully. There are some important changes. {image} The variables: TradeAmount: The percentage of your account balance to trade in dollars. The program uses this to calculate the correct lot size for your boroker. This should work regardless of what your broker calls 1 lot. Example: $1000...
Hello - I got a message on v2.11 which says the "Grid is corrupted". This was at the point of closing a profitable grid. It left a huge numbrt of sells open which I closed manually.

Only difference from the default settings is I set slippage to 0 and true.
  • #758
  • Jan 4, 2015 5:31pm Jan 4, 2015 5:31pm
Quoting fxtrader08
{quote} Hello - I got a message on v2.11 which says the "Grid is corrupted". This was at the point of closing a profitable grid. It left a huge numbrt of sells open which I closed manually. Only difference from the default settings is I set slippage to 0 and true.
That's why it was corrupted. Your broker was unable to close the orders without slippage so it left orphan trades
  • #759
  • Edited 8:06pm Jan 4, 2015 5:37pm | Edited 8:06pm
Just like spyderman says The program won't try to close trades that the broker rejects due to slippage if you have that protection turned on.
A good man knows his limitations ~ Clint Eastwood
  • #760
  • Jan 4, 2015 5:49pm Jan 4, 2015 5:49pm
Got two quick closes, nice way to start
A good man knows his limitations ~ Clint Eastwood
1 trader viewing now