Dow 96 point entry

[JonnyT]

Your figure of 6.3% means nothing.

If you understood trading then you would know that there are other more important factors:

1) What are you risking per trade in terms of your total equity?
2) What drawdowns have you had or are likely to have?
3) What is your average win?
4) What is your average loss?

I could double my money this afternoon simply backing the favourite in the 14:30 at Southwell with my life savings. Would this be worth the risk? Of course not.

The 6.3% is nice, but what risk have you taken to achieve it?

Putting into context a system that I trade averages 30% pcm with a 1% equity risk per day. I am not trying to be clever but just indicating that a system has to have a large margin for error to be viable. I don't think this method is viable.

JonnyT

 

[The Wanderer]

Mr Chill



So what was the points total for January?


Enough to make you rich ?

Or was it a loser over the full month??


Don't want to raise false optimism until I have 3 months of clear trading figures

 

[sidinuk]

Am I missing something here? JonnyT has backtested this system using proper data and says it's not viable with the current rules. Have those people that are trading it backtested it themselves and discovered that JonnyT was wrong or are they just choosing to ignore the fact that it's not a good long term system.

I'm not saying that the idea should be scrapped altogether but that it needs more work before you could even consider using real money to trade it.

 

[The Wanderer]

ChowClown

Thanks for the info.

Obviously I'm trading on something akin to the original system and not up to speed on the updated version as yet.Be interesting to see how each system compares during February.I'll post results at end of Feb assuming of course that I don't get wiped out in the middle of the month!!

 

[The Wanderer]

No backtesting done by me.Just wanted to see if it produced results using minimum stakes-50p a point with Finspreads.

Looks O.K. so far but I'm well aware that 1 weeks figures prove nothing.Even 3 months will prove nothing except to give you a little more faith if still profitable.Even if it is who can say if it will still work in different market conditions?

At this stage no more than an interesting little exercise to prompt a debate.I make no comment,good or bad,about the system-just thought that results produced may be of some interest.

 

[ChowClown]

JonnyT,

I'm testing the system with 5% risk per trade. The max drawdown so far (from 3rd Dec) is 3%.

SELL points: 54 (average of 1.4/day)
BUY points: 248 (average of 6.5/day)

Average loss: 26 pts
Average win: 41 pts

 

[Mr Chill]

Been following this system since 29/12 not long i know ,using daily cash with fins ,so far +345 -44 early days
yet, still not bad results

But strongly advise against following this system cos JonnyT
said it dose'nt work :cheesy: :cheesy:

 

[TheBramble]

Am I missing something here? JonnyT has backtested this system using proper data and says it's not viable with the current rules. Have those people that are trading it backtested it themselves and discovered that JonnyT was wrong or are they just choosing to ignore the fact that it's not a good long term system.

I'm not saying that the idea should be scrapped altogether but that it needs more work before you could even consider using real money to trade it.

All depends how much weight you place on back-testing.

I'm not getting into the argument about whether 'this' system, works or not (haven't looked at it or tried it), but some traders put WAY too much faith in back-testing.

Historical performance is not necessarily a guide to future action.

There's a whole gaggle of crisis-hunters out there playing the odds on precisely that fact - that the future MAY NOT perfectly fit the past.

Sure it makes sense to check your systems and methodologies against the instrument(s) in question, but the market may well have substantially and fundamentally changed (yesterday, day before, whatever) and your historical data and analyses would be worthless.

I guess all I'm suggesting is that keeping an open mind doesn't cost anything - the reverse might well do!

 

[Mr Chill]

Well said Mr Bramble .

 

[rglenn]

JohnyT is King you have to follow what he says! :cheesy:

On a more serious note though - a modified version of Jpwone's system does indeed work for those who are not afraid to take up the challenge.

But be careful of the KING he says you can't trade the Dow CASH!

 

[Mr Chill]

Nice one rglenn,i agree you have to have a open mind and not right things off to quickly ,i'm not saying this system works or dose'nt work, only time will tell .

 

[sidinuk]

Personally I feel more comfortable if I know that a system has worked in the past over several years and hundreds or even thousands of trades, through different market conditons (up, down, sideways) and different market volatilities. By backtesting you gain a good feel for the system, you know the likely drawdowns and under what conditions it works the best. You use different system variables to ensure that what results you have are not just flukes.

Knowing how a system has performed in the past gives you a lot of confidence in trading it in the future. Your more likely to stick with a system through a losing run if you've seen it all before in your testing and can accept that it is par for the course.

But that is my final word on this thread, I shall leave you to your 'system' and wish you all well with it.

 

[JonnyT]

CC,

If you risk 5% of equity per trade then how come the maximum drawdown has been 3%?

One loser would likely drop the 3%, several in a row are to be expected even for the best systems.

I wouldn't trade a dow system that averages 6.5 points per trade even if it did work. I have a breakout system for the Dow that averages 11 points per trade, I don't trade that.

The King

 

[ChowClown]

JT,

I've only forward tested since early Dec '03, therefore early days and since then there have been many days when triggers haven't been hit, as well as only a handful of losing trades:

Losing SELL trades: 5th Dec; 7th/13th/23rd Jan
Losing BUY trades: 3rd Jan

I won't be trading this either, as other variants I'm forward testing are more profitable. I just enjoy testing ideas.

 

[oatman]

Chill, do you actually trade, or just throw comments at other posts?

 

[Mr Chill]

Oatman

Its Mr Chill to you .

 

[JonnyT]

I doubt that Mr Chill has any real trading experience, probably 'plays' with Spreadbets. Of course I could always be wrong.

JonnyT

 

[nkruger]

Just came across this old thread - is anybody still ttrading it ? I have used a quite similiar strat since xmas ('06) which has worked ok. Take the Previous Day's Close(PDC)and multiply it by 0.35% and add or subtract this small number from the PDC to give long and short entry points. That number has been around 47 for a while now, considerably less than the 96 originally mooted in 2003 on this thread.......

 

[mmverk]

Just came across this old thread - is anybody still ttrading it ? I have used a quite similiar strat since xmas ('06) which has worked ok. Take the Previous Day's Close(PDC)and multiply it by 0.35% and add or subtract this small number from the PDC to give long and short entry points. That number has been around 47 for a while now, considerably less than the 96 originally mooted in 2003 on this thread.......

Hi Nkruger,

I came accross this system because you revived the thread. I have written code to run it on the platform my broker uses and have made some adjustments. I treat gaps up in a different way than gaps down which makes the system more robust. The indexes tend to behave differently after downward gaps.

I discoverd that this sytem as far as I have tested it now works best on indexes. I have tested it on a large Dutch bank but it generated too much false signals, whatever the chosen gapsize. It works fine on the Dow and also on the AEX, the Amsterdam exchange index. I will test it on more in the future.

Rules are as follows :
Buy long/short if gap is larger than the predefined gapsize in %
Predifined gapsize depens on index and is different between long and short.
Stoploss is previous close.
If one percent profit is gained (in indexpoints) implement stoploss of 2%, but never lower as original previous close.

Next adjustments I will make is to work with different initial stoploss levels. The system would have won on the AEX 28K with 1 future but would have lost 18K in the period form 1st January 2006. If I can cut the losses without affecting the profits too much, the system could be a big winner. I tested it with one future, max drawdown was 2K, needed capital 6K and 50% in the market.

If anyone is interested, I will post the results on this forum. I am working with 15min charts by the way.

Michel

 

[nkruger]

Hello Michel, yes - always interested. I'm still using the basic 48 point strategy on the Dow every day and now looking at one for the DAX and the Spanish IBEX. Let me know how you get on.

Nick