Still Basketting | Page 101 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Anyone still trading Mouteki? 45 replies

The EUR/USD Rally-Still Dead? 23 replies

The Euro Rally Is Still Dead-coreCPI/TIC Update 3 replies

USD in Toilet but still buys Chinese Goods!!! 11 replies

Immediate market reaction on news are still the same ? 18 replies

Attachments: Still Basketting
Exit Attachments

Still Basketting

  • #2,001
  • Nov 28, 2012 11:47am Nov 28, 2012 11:47am
Quoting omadivi
my "buy" basket....
260 pips closed.....thanks kos....the offline charts totally rock..
It is possible to make large sums trading, small losses, large wins
  • #2,002
  • Nov 28, 2012 12:15pm Nov 28, 2012 12:15pm
Here is an EA design concept I will share. I think it makes rule implementation much easier especially in the context of US rules.

The EA is BasketNetTrader. It simply reads a list of TextBox Names and sums the numbers in them. It then adjusts the basket similar to my BasketAdjust script. You can test just this in demo by making a textbox and naming it "NET" and entering numbers in the value. If you leave the Pair list blank then it will just run on the attached pair so it can be used even in a non basket setting.

I have also attached a version of Heikan ashi smoothed, for the offline chart, that figures out the current net position and updates the text box. Run this on the chart with the EA and you have a fully automated system. This indicator also has an alert on any change of position. You need to run this with RealTime = true if you want the BE to work properly but otherwhise the signals themselves are on close.

You can of course get more complex and even layer multiple conflicting ideasindicators together. The EA will just use the summed net position to open the minimum amount of trades. This is a great way to take any indicator you have the code too and efficiently automate it's signals.

DISCLAIMER: These files are capable of automatically entering trades without user input. If you attach them both to the same live chart you will lose real money.
Attached File(s)
File Type: mq4 Heiken_Ashi_Smoothed_Visual.mq4   17 KB | 872 downloads
File Type: mq4 BasketNetTrader.mq4   16 KB | 901 downloads
  • #2,003
  • Nov 28, 2012 1:09pm Nov 28, 2012 1:09pm
Quoting Paradox
Here is an EA design concept I will share. I think it makes rule implementation much easier especially in the context of US rules.

The EA is BasketNetTrader. It simply reads a list of TextBox Names and sums the numbers in them. It then adjusts the basket similar to my BasketAdjust script. You can test just this in demo by making a textbox and naming it "NET" and entering numbers in the value. If you leave the Pair list blank then it will just run on the attached pair so it can be used even in a non basket setting.

I have also attached a version...
Wow Paradox, what a fantastic idea.

Cant wait to play with it.....
Damn another late night...LOL

Thanks for sharing.
  • #2,004
  • Edited 2:57pm Nov 28, 2012 1:47pm | Edited 2:57pm
Quoting Kosomolate
The system: Basket trading
Pairs: 14 Universal pairs
Timeframe: 15min
SL = 560 pips almost 6% of the acct. VERY RISKY
TP = HAS Direction change
Hi everybody, Hi Kosomolate

The first question I ask myself when I look at a trading strategy is: what's exactly the Money Management (MM)?

So I do it for this strategy. Money Management:
SL: 560 pips (proposed here)
Risk: 6% (proposed here)
Lot size per pair: 0.01 (correspond to a micro lot, i.e minimum lot size)

So we get:
Global lot size (because we have 14 pairs): 0.14
This gives 1.4 USD/pip, so we risk 784 USD/trade (we have a 560 pips SL), and this corresponds to 6% of our capital.
So, we have to get a minimum capital of 13,000 USD if we want to respect our MM. This is not surprising, because we are trading 14 pairs at the same time!!! A consistent capital is necessary to respect our risk. Our goal is to win over the long-term and not to lose our capital after 2 or 3 losing trades.
Note: with this MM, if our capital goes under 13,000 USD, we couldn't respect our MM, because, we cannot choose a lot size under 0.01 So it is necessary to start with more than 13,000 USD. For example: 20,000 USD at least.

Here are my questions:
1/ Please, does anyone could tell me if I am wrong about the MM's calculation? Kosomolate, could you give us more details about MM you advice? Please, according to what I said previously, how could you respect your risk with a capital of only 400 USD (like you said here)?
2/ On this thread, I see a lot of screenshots presenting very huge profits. But, these profits are always get during "trend maket areas". So, obviously, they are positives. But, the most interesting, is to see: how goes this strategy during "ranging areas"? What is our risk? With how many losing trades, could we do bankrout?

As we are traders, the first question we have to ask ourself is not: how many money could we win? But instead, how many money could we loss? Because gains are unlimited but our capital is not unlimited, you see?

I hope someone could help me to learn more about this strategy.

Thanks a lot everybody
  • #2,005
  • Nov 28, 2012 2:51pm Nov 28, 2012 2:51pm
Paradox
Looking at the visual indi I don't see where the SL is at I see loses of -304 and -375
could you add a input for SL as mine is 40 pips per pair unless it is in there and I just don;'t see it
I think you are showing just TotalLosingPips with out consideration for SL
  • #2,006
  • Nov 28, 2012 2:56pm Nov 28, 2012 2:56pm
Here are my questions:
1/ Please, does anyone could tell me if I am wrong about the MM's calculation? Kosomolate, could you give us more details about Money Management you advice? Please, according to what I said previously, how could you respect your risk with a capital of only 400 USD (like you said here)?

14 pairs *40 pip per pair SL = 560 pips at 0.01 is $56 SL
$56/6% = $933
5 core5 pairs 40 per pair SL = 200 pips at 0.01 is $20 SL
$20/6% $333

You could trade this system for under $1000 or even $400 for that matter
  • #2,007
  • Nov 28, 2012 3:05pm Nov 28, 2012 3:05pm
Quoting vinze87
Hi everybody, Hi Kosomolate

The first question I ask myself when I look at a trading strategy is: what's exactly the Money Management (MM)?

So I do it for this strategy. Money Management:
SL: 560 pips (proposed here)
Risk: 6% (proposed here)
Lot size per pair: 0.01 (correspond to a micro lot, i.e minimum lot size)

So we get:
Global lot size (because...

You are wrong about your MM calculation. The pip total is 560 @ $0.10 (roughly) per pip. Not 560 @ $1.40 per pip.

The original method was figured at 40 pips per pair, or 560 total. Some may be larger, some may be smaller, but 560 total.

This amounts to a $56 loss, which is 6% of a $930 account.

FWIW, I have done some calculations with hourly data in my (very similar) method to kos, and it shows approximately a 50% win probability. The good news about that, is that the pip return distribution is heavy tailed towards the positive. The ratio is somewhere on the order of 6 or 7 to 1. Which translates to a positive expectancy... it also so happens that there is a positive expectancy over a very wide range of parameters, which is also a big positive.

For what it's worth, I modeled this strategy with both a trailing stop and with the exit being given by a signaled reversal. The signaled reversal won big. So that is what I use, plus the 560 pip stop.

I haven't optimized the stop loss yet, that may be something beneficial to do - because it seems to me iirc that the selection of the stoploss point at 560 pips was somewhat arbitrary.

Regards
Casey
  • #2,008
  • Nov 28, 2012 3:06pm Nov 28, 2012 3:06pm
I see that dryclean was quicker on the draw than me

Something to note, if you are in the US with a 2% margin requirement it costs approximately $340 in margin to open a basket.

Casey
  • #2,009
  • Nov 28, 2012 4:09pm Nov 28, 2012 4:09pm
Quoting caseyjw
You are wrong about your MM calculation. The pip total is 560 @ $0.10 (roughly) per pip. Not 560 @ $1.40 per pip.

The original method was figured at 40 pips per pair, or 560 total. Some may be larger, some may be smaller, but 560 total.

This amounts to a $56 loss, which is 6% of a $930 account.

FWIW, I have done some calculations with hourly data in my (very similar) method to kos, and it shows approximately a 50% win probability. The good news about that, is that the pip return distribution is heavy tailed towards the positive. The...
have you backtested 15min data?
  • #2,010
  • Nov 28, 2012 4:16pm Nov 28, 2012 4:16pm
you guys are using HAS with settings 2,15,2,15 in any timeframe?
  • #2,011
  • Edited 6:00pm Nov 28, 2012 5:31pm | Edited 6:00pm
Hi Dryclean and Casey,
Thank you very much for answered my question,

I think you are right, I have complicated the MM calculation (I multipiled twice by 14), excuse me please.

For 1 pair, we have: SL=40 pips, risk=6%, Lot size=0.01 we get: $0.1 /pip this gives a minimum required capital: 67 usd
For 5 pairs, we have: SL=200 pips, risk=6%, Lot size=0.01 we get: $0.1 /pip this gives a minimum required capital: 333 usd
For 14 pairs, we have: SL=560 pips, risk=6%, Lot size=0.01 we get: $0.1 /pip this gives a minimum required capital: 933 usd

Quoting dryclean
14 pairs *40 pip per pair SL = 560 pips at 0.01 is $56 SL
$56/6% = $933
Ok, so please, could you explained me, how Kos could trade on the 14 pairs basket, with a $400 account (as he said here #5, #11, #1816)?

And if we start to trade with this strategy, if our account goes under $933, we cannot choose a lot size smaller than 0.01, so what capital is necessary? If we want to risk 6%, this means we want to be able to accept 16 consecutive losing trades. So to be serious, we would get an account which allows us to be able to accept 16 consecutive losing trades.

Quoting caseyjw
I have done some calculations with hourly data in my (very similar) method to kos, and it shows approximately a 50% win probability. The good news about that, is that the pip return distribution is heavy tailed towards the positive. The ratio is somewhere on the order of 6 or 7 to 1.
Casey, you did a backtest, it's very interesting. Please, did you do it on Excel? Did you take the breakeven into consideration?

Thank you again
  • #2,012
  • Nov 28, 2012 6:15pm Nov 28, 2012 6:15pm
Quoting caseyjw
...I haven't optimized the stop loss yet, that may be something beneficial to do - because it seems to me iirc that the selection of the stoploss point at 560 pips was somewhat arbitrary.

Regards
Casey
Interested in how this turns out if you do follow through on it, if it were not for MATLAB I'd have said you'd probably be at this for days..
  • #2,013
  • Nov 28, 2012 8:41pm Nov 28, 2012 8:41pm
Quoting as002
have you backtested 15min data?

Yes and it was not profitable
  • #2,014
  • Nov 28, 2012 8:43pm Nov 28, 2012 8:43pm
Quoting vinze87

Casey, you did a backtest, it's very interesting. Please, did you do it on Excel? Did you take the breakeven into consideration?

I started in excel, trying to get a feel for the numbers... then I moved into MATLAB. The scale of the computation requires something with a bit more horsepower.
  • #2,015
  • Nov 28, 2012 8:45pm Nov 28, 2012 8:45pm
Quoting Thoughts
Interested in how this turns out if you do follow through on it, if it were not for MATLAB I'd have said you'd probably be at this for days..
I may or may not have time in the short term, but i will definitely do it and share the results.

I have some serious projects in the hopper for this weekend and next week (wife wants her x-mas tree )
  • #2,016
  • Edited 11:24pm Nov 28, 2012 11:10pm | Edited 11:24pm
core 5 buy fired anyone in

do we wait for the has body to change color or do we dive in the moment the wick changes color ???
  • #2,017
  • Nov 28, 2012 11:35pm Nov 28, 2012 11:35pm
Quoting unimak
core 5 buy fired anyone in

do we wait for the has body to change color or do we dive in the moment the wick changes color ???
I'm in the buy basket. My chart's 1H TF HAS changed to green an hour ago.
  • #2,018
  • Nov 28, 2012 11:49pm Nov 28, 2012 11:49pm
Quoting vinze87
Hi Dryclean and Casey,
Thank you very much for answered my question,

I think you are right, I have complicated the MM calculation (I multipiled twice by 14), excuse me please.

For 1 pair, we have: SL=40 pips, risk=6%, Lot size=0.01 we get: $0.1 /pip this gives a minimum required capital: 67 usd
For 5 pairs, we have: SL=200 pips, risk=6%, Lot size=0.01 we get: $0.1 /pip this gives a minimum required capital: 333 usd
For 14 pairs, we have: SL=560 pips, risk=6%, Lot size=0.01 we get: $0.1 /pip this gives...
Hi vinze,

Your questions are beginning to become slightly annoying. It sounds like you've totally lost youself in a bunch of numbers (no offence meant). It also sounds like you have serious doubts about the authenticity of this setup. I am going to try and explain it.

First of,your account leverage should be 1:500
On a 400 usd, you are trading 0.01 for 14 pairs, that's a total of .14lots. The required margin for the 14 trades would be less than 60usd(you can do the margin maths)
The recommended SL is 560pips which is 56usd, that's 14%. Ideally, you should start the account with 1000usd. I hope this is helpful.
And fyi: we are in buy mode..
It is possible to make large sums trading, small losses, large wins
  • #2,019
  • Nov 29, 2012 1:24am Nov 29, 2012 1:24am
Quoting vinze87
Hi Dryclean and Casey,
Thank you very much for answered my question,

I think you are right, I have complicated the MM calculation (I multipiled twice by 14), excuse me please.

For 1 pair, we have: SL=40 pips, risk=6%, Lot size=0.01 we get: $0.1 /pip this gives a minimum required capital: 67 usd
For 5 pairs, we have: SL=200 pips, risk=6%, Lot size=0.01 we get: $0.1 /pip this gives a minimum required capital: 333 usd
For 14 pairs, we have: SL=560 pips, risk=6%, Lot size=0.01 we get: $0.1 /pip this gives...
If you follow up through the math you'd see that 40p SL on each 0.01 trade sums up to 4$ (considering that the pip value is 10$ per round lot). 40 * 14 trades sums up to 560 pips, which in turn sums up to 56$ on 0.01 lot trades (again, considering that the pip value is 10$/lot). Seeing as that isn't, in most cases you'd have a 60$/560p SL.

Now, get that into an 1000$ account. A 60$ SL (hard cash) would get you in 6% risk per basket trade, not per actual trade. Theoretically, per actual trade you'd get 0.4% risk. Kinda lightens your calculations up.

And seeing as you all pose in Matlab profficiency and such likes, did you gain your data before or after integration? 'Cos if you did it before, it nullifies it altogether. This isn't a normal gain-per-ratio trade, it's a cummulus of gain-per-ratios, and it doesn't behave (theoretically) as a normal trade would. Do a PID after the data, and normalize data in the scope. You'll see something more than you did.

P.S. : 15m isn't profitable, but 15m entries based on H1 is. I wonder why that would be?!
Please, just slap me around if I'm being silly ...
  • #2,020
  • Nov 29, 2012 2:21am Nov 29, 2012 2:21am
Quoting daemon_delux

And seeing as you all pose in Matlab profficiency and such likes, did you gain your data before or after integration? 'Cos if you did it before, it nullifies it altogether. This isn't a normal gain-per-ratio trade, it's a cummulus of gain-per-ratios, and it doesn't behave (theoretically) as a normal trade would. Do a PID after the data, and normalize data in the scope. You'll see something more than you did.

P.S. : 15m isn't profitable, but 15m entries based on H1 is. I wonder why that would be?!
Interesting. No, I do not gain or filter my data at all... I get my entries and exits from the individual pair data with a randomly generated slippage parameter and another parameter that estimates spread on entries and exits. Why would you run a closed loop response algorithm on a (virtually) stochastic time series?

Also I would be interested to see the results from your M15 entries based on H1.... or would that just be an H1 system on an M15 chart? How much data did you back test on? I don't have very much, only 2 years of hourly bars.

Also please note, when I say not profitable that means that there isn't a positive expectancy within the time period that I tested.
1 trader viewing now