GridsForPips - New 討論

Starting out with EURCHF, since this is closest to its lowest value ever.

Cheers

Soeren

Added USDJPY and USDCHF, both within 20% of lowest, hopefully they'll begin climbing again.

Cheers

Soeren

Adding a lot more pairs this weeekend, waiting for the trades on the old account to close before transferring funds to this one.

If I'm making anymore changes in the programming, it will be small optimizations, to see if it will run faster in backtests, cos right now a 10 year backtest takes days...

Cheers and have a nice weekend.

Soeren

i'm also in grid trading
my planned grid on audusd

the grid lines are support and resistance lines that are display by my indicator

I've just come across another improvement to my EA, the OrderCloseBy()-function, in another thread. I will try to implement that one during this weekend.

Chers

Soeren

Hmm, maybe I should only have started out with one pair/500 cents like Gil suggests, DD is becoming a little high...

Nice to see that Equity is climbing fast 😀

From 60 to 70% during today.

Cheers

Soeren

Soeren,

Your system takes a lot of equity and very little return for the risk. You may want to try other systems until you 'hone' this one?

It seems to me you may be in love with the concept that this is YOURS!

Just an idea....

Gil

True Gil, I might be just a little in love...

As I stated earlier, maybe I should have chosen to trade less pairs, but for now, I'll just wait and see where it ends up.

Currently, my free margin is too small for the trendtrading part to start up, that won't begin until I've got at least 20'000 cents.

I've lowered the Risk and us the full history now as well, but trades opened before yesterday was using higher Risk and only 60 months of history, I'm assuming that's the main reason for the heavy drawdown...

We'll see in the future, if I just can keep my itching fingers off the settings and programming...😳

Cheers

Soeren

It seems like it's found a steady level of 30-40% drawdown, so the first months won't bring in any profits. I expected this, but if it can get the 6% a month in the long run, I'm happy :-)

Cheers

Soeren

Gil, if you're still taking a look sometimes, I just summed my actual deposits to 4492 cents, the rest is welcome bonus, rebate and interest. Nice to be able to utilize bonuses to trade and get a profit that way around. Currently it means I'm up by more than 6% in Equity 😉.

What do you think about the drawdown, I know it's a lot when looking at a glance, but when you realize it's trading 28 symbols, does that mean the average DD is 1-2% per symbol, or is that a wrong way to see things?

As an example, the historical range for AUDUSD is 4514 pips (history from 1989.06.01), I'm trading the grid from the centerline and out. Let's say the price is exactly at the centerline (0.7749) and my EA has opened a long order, suddenly, Reuter's states that a nuclear bomb has droppend on Australia and the Aussie falls like a bowlingball from the 50th floor (Sometimes I like to exaggerate, can't help it😄), causing a daily fall to the historical low. Because of my filter no new long orders are opened along the way, so I only have this minimum-lot order open with a loss of 2757 pips loss. With my current centsaccount, that is a 275.70 cents loss equal to ~6% loss of Equity. The more orders I have open the worse the case is, but it's just an example, besides, on the other symbols involving the Aussie, som might benefit from the freefalling and I'll make profits on those (diversify).
To take it a little further, the AUDUSD falls to 0!!!, My trade has lost 7749 pips, causing a staggering 774.90 cents loss, which means I've lost 16.15% of my Equity in one trade.
The point is, Drawdown is OK, as long as it's a controlled drawdown, which I believe, mine is.

Hope to get some comments on this post.

Cheers

Soeren

As a swing trader, that makes sense to me.

However, I would never go over 20% drawdown if I could avoid it and would focus on around 10% as my sweet spot. That is just me.

Because, when the 'it' hits the fan, you are going to want that cushion right?

Grid traders are fine, but I think it is like construction. First, get the bid, then double it, then double it again and that is what it will cost. haha... :)

GridsForPips posted:
    It seems like it's found a steady level of 30-40% drawdown, so the first months won't bring in any profits. I expected this, but if it can get the 6% a month in the long run, I'm happy :-)

Cheers

Soeren

6% is great. Unless your drawdowns dip to like 50% or more of your account. Like I said above, it seems too risky to me. And, if you ever plan on leveraging this with investors, I would never allow something that deep just to get 6%....

To get 6%, DD acceptable would be 3 times that = 18% OR LESS consistently. The tough part is I believe you have to program your mind to get through the May 5th/6th's 2010 with a grid trader. Even though you will only lose such and such % on one currency, when they ALL trend very hard in the 'wrong' direction, you will be sucking wind, regardless of your instaforex bonus'!

The necessary requirement is that when you are hitting that type of day that you come out looking good at less than 50% drawdown. That would be my goal.....

Hi Gil, thanks for your comments.

As a result of them, Ill try to make backtests of all 28 symbols and merge them with Report Manager, just to give an idea of how far the rabbithole goes (especially around May, if I understand it correctly). I believe it'll take a few days, but I'll try to discipline myself enough to see it through.

In the meantime, Ill look more into JForex API, and see if I can figure out to make the exact same EA to work with Dukascopy, and use their historical tester to get a more precise result...

Cheers

Soeren

JFOREX has a converter, but it never worked 100% with my Ea's when I tested it.

There have only been a few times when the runs get really bad for multiple pairs. On one of my EA's, EURJPY never retraced enough until 2700 pips! That was by far the longest run.

I believe that a good test is run from Nov 2009 through June 2010 and that is pretty good, but 07' and 08' had some nice runs as well on multiple pairs....

Gil

I tried the converter, but despite the adaptions I had to make just to compile the d... thing, it never took a trade anyway.

Hence, I have to work to figure out all the different obstacles, maybe by seeing how the converter interprets MQL4 to Java, and then have a second clean Java, where I put in the right commands one by one switching between the 2 programs and the API-documentation in my 1500+ lines program (man it's gonna take time...🙄)

Regarding backtests, I've just made on AUDUSD today, starting 01.01.2010 and ending 05.09.2010 which lasted from 08:18 to 12:51, roughly 4½ hours. 28 tests is gonna take a long time.
Would you consider making a backtest with only M1 open-prices alright - I'll miss some trades, but it goes a lot faster.

For now, I'll fire up 12 month backtests of the symbols ending 05.09.2010, that should only take roughly 28 * 12 * ½hour = 168 hours 😲!!!

I'm seriously looking forward to JForex testing, haha

Cheers

Soeren

By the way, would screenshots just before and after a 'CleanUp' be of any interest? I'll just zip them and attach them to strategies...

Decided to trade less of the grids to try to lower DD.

'GridArea' is now 0.5, which means beginning trading approx. 50% away from the centerline and out.

Open orders will be kept open until profit is reached (hopefully)

Cheers

Soeren

Sadly, my decision actually cuts off my most profitable symbol until now: CADCHF but for the benefit of DD, I'll stick with this decision...

Cheers