BACK TESTING

[Dr Iraj]

You designed a strategy. You back tested it for the last month. You got 100 wins no losses. WOW. God I have got it.
Does this mean that if you apply the same strategy for next month you get 100 winners no losers. Ok lets be fair, Do you think you get 60 trades as wins?
60 winners can’t be bad!!!
I need you all to participate in this discussion. Please. If you don’t participate we never find out where we are going wrong?
I contribute late tomorrow night.

 

[ChartMan]

You need a system that gives you a real chance. If you get 100 winners, which stocks are you going to pick? No doubt you will only enter 2 or 3 for practical financial reasons.
What we really need is a system that will pick 5 stocks, 3 winners 2 losers. You can bet on all 5, watch the potential losers like a hawk, and exit any bad trades pronto.
This works from a PRACTICAL point of view.
It goes without saying that you need a decent profit out of the winners to cover the potential loss of the losers.

Martin

 

[Dr Iraj]

The figuer 100 was just a number. I correct myself. If a certain strategy choose 5 wins 1 lose last month can you trust this system to choose 5 wins and 1 lose for the next month?

 

[Uncle]

Here we go again........looking for the HOLY GRAIL!!!!

My strategy is to look for bottom reversals with sufficient risk/reward in them for profit, check confirmation from indicators, and go for it. Exit promptly if it turns against - hence taking care of money management.

1 - 10 shares any one time, in a swing/short term trading portfolio.

As for the HOLY GRAIL....... it will never arise.

Best wishes

 

[Skimbleshanks]

As we all know, the key to successful trading is keeping your losses to an absolute minimum - irrespective of whether you believe you have found the Holy Grail (sorry Uncle) or are happily trading your own system (I just love pullbacks).

If you can pick a winner which gives you, say, an increase of 25% and limit any losses to 2%, then you can be far more flexible in the stocks you buy. This strategy assumes you can keep a watching eye on your stocks a couple of times a day.

Skim

 

[Uncle]

Skimble.......thou dost misinterpret Uncle. He hasnt found, nor looking for the HOLY GRAIL. There is none, nor will there ever be one, but methinks there are those who believe such a thing may be possible. Whatever strategy, I find it works for a time then needs adjusting to suit changing conditions, therefore one has to be alert at all times.

 

[Dr Iraj]

oops

The question was:- Should we trust an EDS routine or any other strategy which returned a good result in the last 4 weeks to give us similar result for the next 4 weeks?

 

[Skimbleshanks]

Uncle:

I know you don't think there's a Holy Grail! Every time the words are mentioned you post saying so - I was merely acknowledging the fact that my uttering of the words Holy Grail were bound to rattle you!

With love and kisses,
Skim



[Edited by Skimbleshanks on 16-01-2001 at 01:32 PM]

 

[Uncle]

Beware the snorting raging bull my friend.....

Sent with a smile of course!!!!!

 

[martymcc]

I can't wait till June when my exams will be finished and I can begin my search for the holy grail as I do believe there is such a thing! I'll show you Uncle (or maybe all this studying with no sleep is playing tricks on my mind?)

Marty

 

[ChartMan]

If, by any chance some of this is directed at me,then let me put something straight for all of you.
This Grail routine is something I knocked up one evening when I was learning how to write and develop ESDS routines.
It has no substance or logical foundation whatsover and could have been written by a 5 year old.SO I make no claim for it or it's performance whatsoever.
All I am saying is that this routine has picked stocks that have shown very substantial rises over a few days, and those that show losses have been slow to do so.
3 weeks of backtesting data is meaningless and provides the routine with no substance!
Hey, the clever one among us know that. I don't profess to be clever, and I'm not trying to persuade anyone that this is a Holy Grail- that was "tongue in cheek" a long time ago.I'm sure most don't want to be bored with the results of 5 years worth of backtesting. That's for the acedemics amongst us - and I'm not one of them.
The reason for posting 1/2/3 weeks of backtest is to allow people who don't have AIQ to see an overall picture of what a stock selection has done in recent times.
I am selfish I suppose, in that my trading style is biased towards short term. I want to make 50% in 4 days. I don't want to be safe and make 25% in 6 months, watching the paint dry.My postings, therefore reflect my style.
If anyone wants a precis of the performance of the grail over 5 years, I'll gladly post it .I suggest it will be a waste of space, if you get my drift.

Martin.

 

[martymcc]

Hi Martin, I'm sorry if you thought my post was written in sarcasm but, after completing a module in Intelliget Systems, I believe that we can find a system which comes very close to the Holy Grail. Also, your system seems to represent a fair success rate even though you would need to backtest it over a more substantial period of time as you admit. So keep posting its selections and let that be the judge of your system. Then you can say told you so!

Marty

 

[ChartMan]

No Marty, it was not directed at you but thanks for your concern.

Martin

 

[Uncle]

"Holy Grail"

Martin....I dont think anything is directed at you, or any other member. As you say, those of us who have been trading long enough, KNOW there is no such a thing.. I persist in drawing attention to it in order the less experienced, believe it to be possible. Sometimes we have hit a patch of good fortune where it looks as though we have found the unobtainble, only to find that changing market conditions throw a spanner in the works, so my advice to anyone out there who finds a winning system go with it, but be prepared to drop it as soon as it shows signs of failure.