Testing-account讨论

Programming changed in the weekend - now the EA only trades in the upper and lower quadrant of the historical range.
The risk has been raised to open larger trader trades and/or trade a tighter grid.

I will follow your EA now and than.

For some reason, Instaforex has changed the minimum lotsize to 0.1 instead of 0.01 - effective of 22-06-2010 (sometime before noon)

Deposited an extra 12.33 Euro (16 $) and got the extra 30% deposit-bonus 😉

I would love to see your comments on the system, all trades are disclosed, and you can analyze the system per magicnumber - 2 numbers per currencypair. Remember, that in the weekend 19-20 of June, the programming was changed to only trade the upper and lower quadrant of historical range.

Attached is 3 examples of settings for the EA - if anyone's interested to get anything explained in detail, just let me know...😎


<a href='https://www.myfxbook.com/files/GridsForPips/audnzd.set' target='_blank'>audnzd.set</a>
<a href='https://www.myfxbook.com/files/GridsForPips/cadchf.set' target='_blank'>cadchf.set</a>
<a href='https://www.myfxbook.com/files/GridsForPips/eurchf.set' target='_blank'>eurchf.set</a>

Your system is not viable mate....those open losers are going to come back at ya no matter what you do...

This is a gridtrading system, drawdown is expected. The old orders from before the 21st of June - 33 orders - are with small lots, thus creating low drawdown. Furthermore, its only 2 of the currencypairs (AUDCAD and NZDUSD) that has trades within 50% of the centerline which would be expected to prolong for a long timeperiod.

I expect zero net profit for the first month or two, because the system builds up a pool of orders on the currencypairs that are moving away from the centerline, but once they eventually return towards center they will close in profit (some not because of negative swap, some larger because of positive swap).
When they do return I also expect other currencypairs to begin wandering from center towards historical upper or lower recordvalues (for the record, all currencypairs has minimum 10 years of history - either in the chart or as fixed upper/lower/center values from other sources), thus always creating DD, but I'm confident the AccountEquity will begin to rise once the system has traded for a longer timeperiod (I've traded for less than 2 weeks with changing conditions, programchange and minimumlot raised from 0.01 to 0.1).

If you review my current open trades, you'll see that I'm actually in positive swap at the moment (0.20 Eurocents), and with the diversity in traded currencypairs, I have no worries regarding drawdown versus profitability.
I'm currently trading with an internal high Risk-setting 8 (lowest is 1), which means that the program divides the full historical range with 8 before it calculates gridseparation and lotsizes which involves Maximum number of open and pending orders minus total open orders, free margin (not just Accountbalance or equity), spread or chosen minimum gridseparation.
If it calculated the grid and lotsizes based on the accountbalance the grid would have the same spacing, and eventually blow the account, if priceaction went beyond the historical range. This is not the case with my program because of the above.
I will now do some backtests and upload to the strategies-section and furthermore attach them to this post. For each strategytest, there will be 2 following screenshots, taken in the beginning and end of each test to show which high/low/center-values that are being used in the beginning and end of the test, thus clarifying how much historical widening the program may be able to handle. It will take some days to finish, because a 1 year test usually takes hours, and I'd like to do some 5 year tests for clarification.

Sincerely yours

Soeren.

I suggest you read this authors work: https://www.fooledbyrandomness.com/

It's about the nature of change and it's roll in trading and it does not bode well for a grid based on lack of change.

Yep, there is absolutely no guarantee that those prices will ever be seen again. Anyone remember sterling at £2.40 per dollar?

I can and I can't see it returning there any time soon.

If I understand you both correctly, I'm unlikely to get trades in the upper and lower quadrant of the historical range, because those ranges may never be seen again? Hmm, then I guess its a good thing, that I have the 'GridArea' setting to control how many % from center the trading begins, along with 'FixedTop', 'FixedBottom' and 'FixedCenter'. Within the program, I have the possibility to control from when (1980, 1990, 2000?) the gridrange is being calculated, but it certainly is something i could change to an external input if that, in your opinion, is better (more desirable)?

This is an experiment, and I'm looking forward to see how it goes in the following months with the current settings and currencypairs. 15% a month, from my point of view, is a huge gain, if the corresponding DD is acceptable. My original goal was 6% a month equal to 100% profit a year...


The first test is ready - AUDUSD - as seen in the screendumps, the starting grid was 4246 pips, versus the end grid 5300 pips, a 25% widening solely in the upper direction. The grid, at some point was actually bigger, but I need to do some extra programming to have that shown in a screendump. 😕
The all-time lowest Equity only just manages to stay above zero: 1.65%!!! But the point is, it stayed above with high Risk...

And I'm aware that the test has poor quality, but for some reason, the downloaded data doesn't go as far back as the ordinary monthly data...

If you wan't me to change the 'age' of calculated data let me know, and I'll probably have it ready later today (tomorrow at the latest).
An example of age could be 01.01.2000 if we suspect the prices won't ever go back to the values of the good old days...😀
That would give 5 years of history if the new tests begin 01.01.2005 too, and the downloaded data, I think, goes back to Oct. 1999, this would give more accurate testresults.

Soeren,

You're obviously new at this. I can't fault you're enthusiasm, but you're just barking up the wrong tree. It's not going to happen no matter how clever you get.

There just is no way to deal with the dangler problem long term. You cannot out bank a big move.

Elkart, it seems you have nothing but negative things to write to other people about their ideas don't you?

I didn't register here until 2 weeks ago, so I don't know if you've just deleted some of all the profitable systems you have shared with the other people in this forum, but I must say, that I'm not impressed with your VG intraday-system, it doesn't seem to be do much better than my gridsystem - furthermore, you have only tested on Demo, I thought the only viable testresults would be from Real-money accounts that can run into all the usual problems with servers, pricefeeds and dusky brokers?

Attached is a screendump from a strategytest done with downloaded data.
EURCHF, datetimelimt for historical data: 01.01.2000, Minimum_months_of_history: 1 !!!, Risk: 1 (lowest possible),
GridArea: -0.1 (which means it both buys and sells 10% around the centerline), CleanUp=TRUE and AccountEquityPercent=101 which makes the EA close all orders once the AccountEquity has reached another 1% in profits...

I suspect the full report to be available sometime at the end of the weekend, then you can have a look...

I forgot to tell the initial gridrange was 237 pips, and you can see the current Equity in th chart...

I have plenty positive to say sometimes where appropriate, but on grids and open dangler systems I'm very negative yes. Especially if the idea is to manage client funds with the systems as the outcome is inevitable.

Last time I did what you're doing now it cost me $15 000. Over and above the money factor the lesson you're about to learn is going to take months. And that's a greater cost than the money.

Grids can't work, they're fundamentally flawed. You limit the profits while letting your losses run.

The VG system is a failure. Even though it made money it would eventually also crash out. I said so. Can't work. Just like the grids can't work. Any system that doesn't deal with the losers will eventually collapse.

That's where the CleanUp and AccountEquityPercent settings come into play... Losers? Yes, if I wan't to deal with them, no problem...

Pyramiding works, the exact opposite of what you're doing.

Position building. Stopping out the losses as they occur and letting the winners run. But then you have to have some insight into the market and make a directional call.

I have those to. Partial TP's and partial SL's. Doesn't work long term. In ranging market sure, but take a good look at early 2009 on USD.

If ever we are set up for a short squeeze on EUR it's now. Never in the history of the world has so many people had such a one sided position on a currency. Except on the Zim dollar maybe and we know how that ended.

I don't know what the trigger will be, but at some point there's going to be short covering on euro and then it's the dollars turn. No grid will survive that.