95% ASX100 success rate

[skyQuake]

Hmm Interesting Wysiwyg...

I'm surprised there exists data that goes back to 1980 for backtesting! Where do you get your data?
esp. historical XJO or All Ords constituents for the testing universe

 

[shouldaindex]

Similar question, where do you find (for example) the top 100 companies by market cap for 1992.

 

[CanOz]

I'm surprised there exists data that goes back to 1980 for backtesting! Where do you get your data?
esp. historical XJO or All Ords constituents for the testing universe

Premium Data has a fair bit of great quality data....

 

[skyQuake]

Premium Data has a fair bit of great quality data....

Thanks, will give that a go. Ive been fussing over incorrect Index constituents over a whole array of providers

 

[RobertoHood]

Hey guys,

Here are this mornings BUYS, i was a bit slow off the mark

BHP 630 @ $25.510

TWE 2870 @ $5.596

WPL 485 @ $32.64

As Monty Burns woulds say.... Exxxcelllent

 

[prawn_86]

As Monty Burns woulds say.... Exxxcelllent

Why do you monitor them throughout the week if the only way of testing the 'system' is EOD on Fridays? Ptherwise if you are exiting before that it isn't really what you have described being a system with a 95% success rate

 

[RobertoHood]

Here is a summary from my back testing for the 3 month period to date.

I found some flaws in my system , and i got a more realistic strike rate far from the 95% W/L rate i was initially getting.

Still going to trade this regardless, and may even be able to tweak a bit more and narrow in on losses.

 

[CanOz]

Here is a summary from my back testing for the 3 month period to date.

I found some flaws in my system , and i got a more realistic strike rate far from the 95% W/L rate i was initially getting.

Still going to trade this regardless, and may even be able to tweak a bit more and narrow in on losses.

Stop loss is -2.65%, how did you settle on such an accurate number?

 

[RobertoHood]

Stop loss is -2.65%, how did you settle on such an accurate number?

Measuring volatility, it seems to be a good range to allow for the most of the stocks to move on a weekly basis.

Falling below 2.65% seems to see the stocks finish the week down on the open.

 

[skc]

Here is a summary from my back testing for the 3 month period to date.

I found some flaws in my system , and i got a more realistic strike rate far from the 95% W/L rate i was initially getting.

Still going to trade this regardless, and may even be able to tweak a bit more and narrow in on losses.

Your position sizing and rate of compounding used in the back test may be unrealistic.

1. you'd never put all your eggs in one basket. Your back test period is so short you haven't hit a major adverse event. Imagine a profit downgrade that cut the stock price by 30% while you have it long with 100% of your account.

2. The opening match is getting very thin these days. I'd say something like $20-30k position would be about the size that you can trade on open without moving the match price.

 

[RobertoHood]

Your position sizing and rate of compounding used in the back test may be unrealistic.

1. you'd never put all your eggs in one basket. Your back test period is so short you haven't hit a major adverse event. Imagine a profit downgrade that cut the stock price by 30% while you have it long with 100% of your account.

2. The opening match is getting very thin these days. I'd say something like $20-30k position would be about the size that you can trade on open without moving the match price.

The starting capital for the backtest doesnt mean anything , it could realistically be any number.

I personally usually trade with around $50k split up by how ever many buy signals i get.

As for the product downgrade yes that could happen, but thats the market we trade.

 

[skyQuake]

The starting capital for the backtest doesnt mean anything , it could realistically be any number.

I personally usually trade with around $50k split up by how ever many buy signals i get.

As for the product downgrade yes that could happen, but thats the market we trade.

Think he meant capital allocation. Not dollar value.
eg LYC was in the 100s at one point I believe, then came a few 30% moves. If LYC was your only signal for the week would you have put in all your $ into it?

 

[RobertoHood]

Think he meant capital allocation. Not dollar value.
eg LYC was in the 100s at one point I believe, then came a few 30% moves. If LYC was your only signal for the week would you have put in all your $ into it?

I would never risk 100% capital on one trade, especially with a system which is hitting a 60.5% of the time.

I have also found that when i get a BUY signal, and their is no "fundamental" reason which could be contributing to the signall , the stocks appear to perform better and it smooths out the false BUY signals. Eg BHP , WPL this week, among others in the back test which also provided false buys.

Also the reason why i only trade ASX100 stocks , which compared to overall stock world in Australia is relatively less volatile although they can and do happen.

 

[RobertoHood]

What a week!

Bit of a balls up going through with BUY signals on WPL and BHP

Hope everyone survived the week all good

 

[tech/a]

So you close these out tonight and new prospects next week?

Win rate 66% this week.
Looking forward to next week

 

[skyQuake]

Nice exit on the WPL. Does your system account for discr exits? You mentioned it previously but I cant seem to find it anymore

 

[Porper]

What a week!

Bit of a balls up going through with BUY signals on WPL and BHP

Hope everyone survived the week all good

Just to clarify, are you using opening price on Monday and closing price on Friday? If so you seem to be getting in at a lower price and selling at a higher price?

 

[RobertoHood]

Hey Guys,


Had a bit of free time, so decided to run another back test starting from Aug-11-2014 and running through till Aug-21-2105

So a years worth of data , again just running through the ASX100 stocks.

It simply displays what would happen if every single BUY signal over the last year was acted upon.
(which may/may not be possible, depending on available capital, multiple BUY signals in a week etc)

I also included an accuracy rate on the spreadsheet which is an indication of what % of the BUY signals were indeed accurate for each particular stock.

This backtest is also a whopping 1% up from my previous back test which used a 12 week time frame.

The order at where the "back test" capital starts to BUY from is not by DATE , but from Alphabetical Order of the stock names , but looking at the data bomb , i really don't think it would make much difference , and is probably not far off being completely random.

I have uploaded a snap shot of a backtest with a -2.65% stoploss, and also a snap shot of one without a stoploss at all , as well as all the corresponding data.

Also, i realize the compounding does get a little bit ridiculous ... but hey play around with it!

Enjoy, also lowered the starting capital to $10,000.

 

[RobertoHood]

Just to clarify, are you using opening price on Monday and closing price on Friday? If so you seem to be getting in at a lower price and selling at a higher price?

Hey Porper,

I ended up selling early to take profits while i could! Although i wasn't so lucky with BHP.

The system ended the week with a 33.3% accuracy rate. With TWE being the only stock out of the 3 to end the week up from the open +6.33%

As for the opening prices, attached are my Confirmed BUYS of BHP, TWE and WPL , i was a bit late getting in with TWE, but as for the other two i got them bang on the open.

 

[RobertoHood]

Nice exit on the WPL. Does your system account for discr exits? You mentioned it previously but I cant seem to find it anymore

Thanks skyQuake,

The system only takes into account stocks up at Fridays close from Mondays open.

Any exits during the week, do not count toward the systems win% , only if the stocks closes the week up from the open do they count.

And as i am now starting to realize i'm more toward the 65% win rate as opposed to 95%!