Menu
Sister Sites

Options

Bookmark Thread

First Page First Unread Last Page Last Post

Printable Version

Similar Threads

The DIBS Method... No Free Lunch continues 11319 replies

Free lunch mentality 6 replies

No Free Lunch: Director's Cut 33 replies

Attachments: No Free Lunch but all the Free Coffee you can drink
Exit Attachments

No Free Lunch but all the Free Coffee you can drink

Quoting Pipfetcher
Does anybody have any input with regards to the amount of capital you can put at risk with this strategy on any single trade.I have always believed that no more than 2.5% of your capital should be at risk on any 1 trade and I would just like to know if you guys find this to be true with this strategy also as i have not proparly backtested it so far.

Tx
Hey PF,

I think that is a very personal thing. I myself risk only 0.5% of my equity as I donīt think that Iīm good enough right now, so I try to learn and just risk that amount. If you feel comfortable with risking 2.5% - go ahead. The strategy works, but you will get stopped out a lot. Most of the time with 0, sometime your SL is hit, before the market moves to the first TP. If you face that a couple times in a row, you would have large drawdowns by risking 2.5%.

Once I feel more comfortable with my own trading I might increase to 1%.

Hope I could help
DaBuschi
Quote
I think that is a very personal thing. I myself risk only 0.5% of my equity as I donīt think that Iīm good enough right now, so I try to learn and just risk that amount. If you feel comfortable with risking 2.5% - go ahead. The strategy works, but you will get stopped out a lot. Most of the time with 0, sometime your SL is hit, before the market moves to the first TP. If you face that a couple times in a row, you would have large drawdowns by risking 2.5%.

Once I feel more comfortable with my own trading I might increase to 1%.

Thanx for the speedy reply.

I agree with u that it depends on the personality of the trader but it also depends on the win/loss ratio of your system and that can only be determined by backtesting,backtesting and then some more back testing.

This seems to be a very good,solid and logical system but it is very difficult to backtest because you cannot simply take the "cable" for instance and start backtesting because one of the golden rules imho of this system is to trade the currency pair that shows the most "momentum" in a certain direction and therefore it would be,for lack of a better word,"incorrect" to just trade every signal u get on any given pair without taking momentum into account.

Basically what i am getting at is that it is difficult to backtest and if anybody has any idea of what the win/loss ratio of this system is I would be very glad if they could share it.

Thanx again for your reply.

Cheers
  • Post #363
  • Quote
  • Edited at 3:29pm May 28, 2008 2:43pm | Edited at 3:29pm
Quoting Pipfetcher
Basically what i am getting at is that it is difficult to backtest and if anybody has any idea of what the win/loss ratio of this system is I would be very glad if they could share it.
Still wouldn't change that you won't know when loser or winners will occur or in what order.
Or whether you'll be around at the times the winners occur. Or that you'll manage it as perfectly as in backtesting.
Or. Or. Or.

Basically pre-calculating stuff like this is near impossible unless you have actual, longer term live data you could throw into something like a monte carlo simulation.

Personally I think with just 1% you can make and lose quite enough money quite rapidly. Higher risk automatically increases theoretical(statistical) likelyhood of bust-out even at a "good" win/loss ratio. If you've ever had statistical analysis, just figure out how high the chance is for "enough losers in a row" with risk = your planned size to occur, with likelyhood per loser being even any kind of low low percentage. Even at only 30% losers the probability(P) for 30-40+ losers in a row should be well above 0.

It's been too long for me to just do the math spontaneously

The tricky bit is that it's even less possible to predict intermittent win-loss flip-flops and the fact that only full losses have a predictable size and anything else like adjusted stops to a 1000 pip runner are not predictable at any point in time..

Personally I'd just not be greedy-swingy with risk until you've actually figured out your own profile on demo or small stakes.

I'd like to add there's a good probability I've just written a huge load of rubbish, too. It's not really my area of expertise. But check out any and all posts by Darkstar on the topic if there are any around..I think this whole % topic was discussed quite a few times on FF.

P.S. Found a nice quick and dirty example maker ==> http://www.stat.sc.edu/~west/applets/binomialdemo.html
For N(amount of trades) = 200 and probability p = 30% for them to be losers, the outcome for 40 or more X is 0.9991, i.e. almost certain
Not that this is very applicable given that there is no clear 1 or 0 outcome as win or lose does not have an equal size and effect. But simplified: it still means that within 200 trades even with 70% win rate the chance for 40+ losers is almost certain(unless I missed something here). Assuming you have them in a row(seperate probability for that, but it's > 0%) and 2% on each, well, you take a guess how much is left then
Binomial probably isn't applicable, but as I say..I've no real clue
Just trying to give a thumbs-rule ballpark idea of what's all possible in the wild wild west..
Trust price. Know yourself.
Quoting SeekingLight
Even at only 30% losers the probability(P) for 30-40+ losers in a row should be well above 0.
For me every trade is an independant event, that means that next trade's outcome will never depend on any of the trades that has already happend. This means that statistically the probability to have 10, 100 or 1000+ losses in a row (or wins for that matter) has nothing to do with one systems probabilty. Backtesting will not show future outcome imwo.
[quote=SeekingLight;2024415]

Personally I think with just 1% you can make and lose quite enough money quite rapidly. Higher risk automatically increases theoretical(statistical) likelyhood of bust-out even at a "good" win/loss ratio.

Have to agree. Especially here. This strategy is supposed to be a low risk strategy. The idea is to risk little, many times and try to catch a runner now and then. If you are trading it properly you'll catch a few long running trades that will give you the chance to bring your position(s) to break even and as it goes further in your favor more setups may appear, allowing you to add on and build your position while keeping risk in check.
Don't change the low risk aspect by taking on too large of an initial position. If the trade does work out there will be time to add on.
Smaller initial positions will allow you to probe for the runners without rapid drawdowns.
i learn here....
  •  zinc6
  • | Joined May 2007 | Status: don't just do something, sit there! | 54 Posts
Hi All,

Been following this (& other) thread since the beginning and back in late April did a quick mechanical backtest and since there was some interest thought I'd post my findings here.

PLEASE NOTE: This was based on Oct 07 - Apr 08 NF data - 7 months of data is most likely NOT statistically significant ... but as I said above it was a quick test to see if DIBS was viable as an entry method and 7 months is all I could easily export from MT4 at the time.

TEST PARAMETERS:

Pairs: EUR/USD, GBP/USD, USD/CHF & USD/JPY
Times: All IB's 0600 GMT - 1600 GMT ... also 0500 GMT if an IB
Daily Open: 0600 GMT
Entry: Bar High +1 if above daily open, Bar low -1 if below daily open - spread included (oanda)
Stop: At bar low/high as applicable.

As this was a quick mechanical test all I tested was what the maximum % profit reached prior to being stopped out.

RESULTS:

Pair__ ____#T __1:1__1:2_1:3_1:4_1:5__1:10___%1:1
EUR/USD - 137 - 66 - 41 - 29 - 25 - 23 - 17 ------48.1%
GBP/USD - 125 - 72 - 52 - 43 - 32 - 26 - 12 ------57.6%
USD/JPY - 139 - 61 - 37 - 27 - 23 - 19 - 7 --------43.8%
USD/CHF - 138 - 57 - 35 - 27 - 23 - 21 - 14 ------41.3%

In essesence 50% of trades reach at least 1:1 R prior to being stopped out.

Using the above numbers you can "optimise" a % based TP ... but this varies from currency to currency and given the short time period of the test to base any system on this would be "curve fitting" and almost certain to lose money.

Another thing worth mentioning is the trend (based on weekly) had little noticable effect on stats other than the average maximum %profit reached ... and this was due to a few extream trades that managed to stay in for the whole run netting well in excess of 100% ... could this be the point???

My conclusion was that as a mechanical entry system it shows alot of promise ... the "secret" to making it profitable comes down to trade selection, management and exits ... and of course practise practise practise

I believe backtesting provides dimishing returns and commenced forward testing this system in demo 2 weeks ago using PC's close half at 1:1 and trailing stops behind minor s/r levels. Asian timezone so I'm only taking bars till 1300 GMT. So far:

10 trades - 5 b/e - 2 wins - 1 loss - 2 open - closed: +2.3% open +2.5%

Again promising ...

zinc6
Quoting zinc6
Hi All,

Been following this (& other) thread since the beginning and back in late April did a quick mechanical backtest and since there was some interest thought I'd post my findings here.

PLEASE NOTE: This was based on Oct 07 - Apr 08 NF data - 7 months of data is most likely NOT statistically significant ... but as I said above it was a quick test to see if DIBS was viable as an entry method and 7 months is all I could easily export from MT4 at the time.
Thanks for sharing these results with us. Good job
Quoting zinc6
Hi All,
RESULTS:

Pair__ ____#T __1:1__1:2_1:3_1:4_1:5__1:10___%1:1
EUR/USD - 137 - 66 - 41 - 29 - 25 - 23 - 17 ------48.1%
GBP/USD - 125 - 72 - 52 - 43 - 32 - 26 - 12 ------57.6%
USD/JPY - 139 - 61 - 37 - 27 - 23 - 19 - 7 --------43.8%
USD/CHF - 138 - 57 - 35 - 27 - 23 - 21 - 14 ------41.3%
zinc6
Great job! A question though: Does the "1:2" column show how many trades of the "1:1" that reached 1:2, and "1:3" show how many trades of those that reached 1:2 eventually reched 1:3 and so on?
Quoting NowAndLater
Great job! A question though: Does the "1:2" column show how many trades of the "1:1" that reached 1:2, and "1:3" show how many trades of those that reached 1:2 eventually reched 1:3 and so on?
Hey!

I am sure it was meant like you wrote. It can't be otherwise , because any 1:2 must have gone through 1:1 , and so on ..
Quoting mades
Hey!

I am sure it was meant like you wrote. It can't be otherwise , because any 1:2 must have gone through 1:1 , and so on ..
Yeah, well that means that ~10% reached 10:1 which isnt bad at all...
Another aspect to consider:

In spite of small % risk you can still have decent position size, given small SL range of IB.

No reason to consider unnecessary large % risk.


Thanks,
TS

[quote=billflet;2025091]
Quoting SeekingLight

Personally I think with just 1% you can make and lose quite enough money quite rapidly. Higher risk automatically increases theoretical(statistical) likelyhood of bust-out even at a "good" win/loss ratio.

Have to agree. Especially here. This strategy is supposed to be a low risk strategy. The idea is to risk little, many times and try to catch a runner now and then. If you are trading it properly you'll catch a few long running trades that will give you the chance to bring your position(s) to break even and as it goes further in your favor more setups may appear, allowing you to add on and build your position while keeping risk in check.
Don't change the low risk aspect by taking on too large of an initial position. If the trade does work out there will be time to add on.
Smaller initial positions will allow you to probe for the runners without rapid drawdowns.
Haste not to Enter AND Haste not to Exit !-TradeStar
Thanks Zinc6.

Few extreme trades netting above 100% means 1:100?

TS

Quoting zinc6
Hi All,

Been following this (& other) thread since the beginning and back in late April did a quick mechanical backtest and since there was some interest thought I'd post my findings here.

PLEASE NOTE: This was based on Oct 07 - Apr 08 NF data - 7 months of data is most likely NOT statistically significant ... but as I said above it was a quick test to see if DIBS was viable as an entry method and 7 months is all I could easily export from MT4 at the time.

TEST PARAMETERS:

Pairs: EUR/USD, GBP/USD, USD/CHF & USD/JPY
Times: All IB's 0600 GMT - 1600 GMT ... also 0500 GMT if an IB
Daily Open: 0600 GMT
Entry: Bar High +1 if above daily open, Bar low -1 if below daily open - spread included (oanda)
Stop: At bar low/high as applicable.

As this was a quick mechanical test all I tested was what the maximum % profit reached prior to being stopped out.

RESULTS:

Pair__ ____#T __1:1__1:2_1:3_1:4_1:5__1:10___%1:1
EUR/USD - 137 - 66 - 41 - 29 - 25 - 23 - 17 ------48.1%
GBP/USD - 125 - 72 - 52 - 43 - 32 - 26 - 12 ------57.6%
USD/JPY - 139 - 61 - 37 - 27 - 23 - 19 - 7 --------43.8%
USD/CHF - 138 - 57 - 35 - 27 - 23 - 21 - 14 ------41.3%

In essesence 50% of trades reach at least 1:1 R prior to being stopped out.

Using the above numbers you can "optimise" a % based TP ... but this varies from currency to currency and given the short time period of the test to base any system on this would be "curve fitting" and almost certain to lose money.

Another thing worth mentioning is the trend (based on weekly) had little noticable effect on stats other than the average maximum %profit reached ... and this was due to a few extream trades that managed to stay in for the whole run netting well in excess of 100% ... could this be the point???

My conclusion was that as a mechanical entry system it shows alot of promise ... the "secret" to making it profitable comes down to trade selection, management and exits ... and of course practise practise practise

I believe backtesting provides dimishing returns and commenced forward testing this system in demo 2 weeks ago using PC's close half at 1:1 and trailing stops behind minor s/r levels. Asian timezone so I'm only taking bars till 1300 GMT. So far:

10 trades - 5 b/e - 2 wins - 1 loss - 2 open - closed: +2.3% open +2.5%

Again promising ...

zinc6
Haste not to Enter AND Haste not to Exit !-TradeStar
Quoting Pipfetcher
This seems to be a very good,solid and logical system but it is very difficult to backtest because you cannot simply take the "cable" for instance and start backtesting because one of the golden rules imho of this system is to trade the currency pair that shows the most "momentum" in a certain direction and therefore it would be,for lack of a better word,"incorrect" to just trade every signal u get on any given pair without taking momentum into account.

Basically what i am getting at is that it is difficult to backtest and if anybody has any idea of what the win/loss ratio of this system is I would be very glad if they could share it.
I will make one huge point at this juncture.

The fact that a promising simple system is difficult to backtest should be a flashing light to you.

A good thing.

Anything easy to implement and test is also easy to destroy. DIBS will be more difficult to destroy because no two of you will trade the concept the same way.

-PeterCrowns-

PS: I am impressed with the effort so many of you have expended. -zinc6- Within your breakdown of potential profits you have given away the greatest potential of the method. PC
Quoting PeterCrowns
I will make one huge point at this juncture.

The fact that a promising simple system is difficult to backtest should be a flashing light to you.

A good thing.

Anything easy to implement and test is also easy to destroy. DIBS will be more difficult to destroy because no two of you will trade the concept the same way.

-PeterCrowns-

PS: I am impressed with the effort so many of you have expended. -zinc6- Within your breakdown of potential profits you have given away the greatest potential of the method. PC
You got me with that! Agree 110%!

The beauty of this system lies on its amazing simplicity. In this case, less is definitely more.

I've been following your posts since the beginning, but this one I colud not resist to reply.

Thanks a lot Peter for showing these concepts to us. You really opened my eyes.

Cheers!
Many share my views with me. But I don't share them with them...:cool:
Quoting PeterCrowns
I will make one huge point at this juncture.

The fact that a promising simple system is difficult to backtest should be a flashing light to you.

A good thing.

Anything easy to implement and test is also easy to destroy. DIBS will be more difficult to destroy because no two of you will trade the concept the same way.

-PeterCrowns-

PS: I am impressed with the effort so many of you have expended. -zinc6- Within your breakdown of potential profits you have given away the greatest potential of the method. PC
Peter,

Just wanted to drop a word of thanks for all of the information that you have shared and for serving as a catalyst for better trading for me.

Thanks again!
Quoting PeterCrowns
I will make one huge point at this juncture.

The fact that a promising simple system is difficult to backtest should be a flashing light to you.
...

-PeterCrowns-

PS: I am impressed with the effort so many of you have expended. -zinc6- Within your breakdown of potential profits you have given away the greatest potential of the method. PC
Hi Peter, thanks for dropping in... (Bemac places a full pot of Columbian {COFFEE } on PC's Table.)

It is amazing how many people want to be Traders but don't want to have to Think about it. I Wonder, Do you have an EA to Measure "Trader Discretion" or "The Hot Hand"

Sorry Peter, I know how ridiculess that is but I could just see the question coming.

Guys, & I know most of you understand this but,
How can you Backtest a Discretionary system?

ABC is in a Horizontal Channel going back 4 months. During that 4 Months it has had 26 Daily IB's. Am I Trading it? He|| no. But it's gonna Trade on a Backtest.

MARCH 11th 2008
Peter Crown's First Post to this great Forum.
" That is the way I remembered it.
I've been away from the CME for 9 years. I much rather trade from my home office. Less noise.
The only pressure on me is what I place on myself. A truth that has always held true, even if I didn't recognize it.
Peter"

Not yet 90 days and we have multiple suggestions on ways to improve on this "Discretionary System."

HELLO! WAKE UP!
If you can't afford to Forward Test this Method with Real Money,
At Least, Forward Test it with Demo Money.
The Method Works but,
YOU, may be required to take some Training Modifications.
This is my first post since joining FF, although I have wandered a number of times as a guest.
Bemac's point is one that stands out the more you wander, some very good, & promising systems are destroyed by 'add this, I think this, It didn't work for me' kind of approach. Discretionary trading systems ARE the way to go...but you got to work at it. No system, of whatever kind, can replace the brain of an educated, experienced trader, & that only comes froms from trading. So please guys & girls...'IF IT AIN'T BROKE..............'
Run with it.
My thanks to Peter, to Bemac, for a truly great system. To everyone else for a great thread, & a great forum.

(It is amazing how many people want to be Traders but don't want to have to Think about it. I Wonder, Do you have an EA to Measure "Trader Discretion" or "The Hot Hand"
How can you Backtest a Discretionary system?
Not yet 90 days and we have multiple suggestions on ways to improve on this "Discretionary System.")
So make sure it's an exciting one, 'cos it's the only one you get!
Quoted from Bemac -

'HELLO! WAKE UP!
 If you can't afford to Forward Test this Method with Real Money,
At Least, Forward Test it with Demo Money.
The Method Works but,
YOU, may be required to take some Training Modifications.'

I don't believe there is any other way to realistically test any system/method etc. You can by all means put some indicators on a chart and look back to see if they're productive and especially if you look at the chart from right to left they will look great, even the worst system might look good from this angle!!

With any system, no matter how good, there is no accounting for the personal input of the user. If we could all trade like Peter Crowns, Jacko etc it would be paradise for all. Sadly we all introduce our own tweaks and personalities into our trading. That is why Bemac's suggestion is spot on, YOU have to trade the system yourself to see how it performs for you. Indeed I would go further than Bemac, forget demo trading it's a waste of time unless you plan on becoming an eternal student of forex trading. I have noticed others recommend places where you can trade for pennies, that would be far more rewarding and realistic a way to test something out.

Sorry to barge in to your house Bemac, just wanted to agree with your sentiments. No coffee?? I understand! btw, I detect a very Scottish background here, where do you hail from?

rgds, strada.
Quoting Bemac
Hi Peter, thanks for dropping in... (Bemac places a full pot of Columbian {COFFEE } on PC's Table.)

It is amazing how many people want to be Traders but don't want to have to Think about it. I Wonder, Do you have an EA to Measure "Trader Discretion" or "The Hot Hand"

Sorry Peter, I know how ridiculess that is but I could just see the question coming.

Guys, & I know most of you understand this but,
How can you Backtest a Discretionary system?

ABC is in a Horizontal Channel going back 4 months. During that 4 Months it has had 26 Daily IB's. Am I Trading it? He|| no. But it's gonna Trade on a Backtest.

MARCH 11th 2008
Peter Crown's First Post to this great Forum.
" That is the way I remembered it.
I've been away from the CME for 9 years. I much rather trade from my home office. Less noise.
The only pressure on me is what I place on myself. A truth that has always held true, even if I didn't recognize it.
Peter"

Not yet 90 days and we have multiple suggestions on ways to improve on this "Discretionary System."

HELLO! WAKE UP!
If you can't afford to Forward Test this Method with Real Money,
At Least, Forward Test it with Demo Money.
The Method Works but,
YOU, may be required to take some Training Modifications.


Hello, Bemac...

I have just quoted this particular post in the DIBS thread...

Somebody (whom I guess have not read Peter's and your posts...) asked if "we should not be a little discretionary with this system?"

I guess your post clears all the question...

Cheers!
Many share my views with me. But I don't share them with them...:cool:
0 traders viewing now