Analysis wanted...

[tech/a]

Oh you are missing the point. Which is surprising, as a system trader, are you not looking for the stocks that outperform everything else.

Actually no.
Id have thought that more discretionary.

Perhaps we should make up our own indicies that better reflect the type of stocks we trade

Did some work on this quite a while ago.
There are many variations that can be considered.
If you have a constantly changing stocks in your
portfolio you can make up an index of those stocks adding and subtracting from it.
You then have two plots.
(1) Equity curve
(2) Your index.

We tracked this idea for a while with T/T (off of the boards)
Then you can make an index from a universe of stocks you may wish to look for signals in.
This of course is endless.
Stocks $1-$5 trading over $500K a day and above their 150Ema for instance.
If thats your universe your looking for signals from you can plot a custom index.

You want your index to be out performing the univese index.

Must get a few months to look into that again.

 

[sinner]

Basically we can't use one index as an overall filter.

Depends what you're up to now, doesn't it? I would be careful in reducing this discussoin to what you've said.

TH is simply asking the question of "why" (correct me if I'm wrong), to get a discussion going, rather than saying "don't ever use the XAO for anything".

It's important to recognise what the XAO is versus the XJO, or NYCOMP versus SP500 for example. They represent the equity curve of an investment profile. If you are buying the composite, generally you're buying smaller sized companies and in holding the index you "reward" losers (by investing more) while "punishing" winners (by taking profits).

Buying the top weighted index is a natural way to reward winners and punish losers and generally buy bigger sized companies. Size also happens to be a good proxy for information, in the information theory sense of the word.

I would also point out that statistically if a index has an up day then something in excess of 70% of component stocks will also finish the day up. Think of it like reverse breadth, it's definitely useful.

Perhaps we should make up our own indicies that better reflect the type of stocks we trade

Check out the Livermore Active Issues http://cssanalytics.wordpress.com/2009/11/25/the-livermore-active-issues-index-using-the-nasdaq-100/

pretty easy to build an index like that.

 

[Trembling Hand]

Actually no.
Id have thought that more discretionary.

How come Tech?

 

[kid hustlr]

Had a re-read of this. Good thread.

@ Sinner:

I like how you put it that we are 'rewarding winners' by buying the index. It's like following a momentum/trend following strategy.

When comparing the XJO vs the XAO, I note that over the last year the 'gap' between the XJO and XAO has shrunk. The XJO is outperforming. This tells me the top 200 companies have outperformed when compared to the wider top 500 index that is the XAO. I also note this is on a cap weighted basis.

@ TH

I understand what you are getting at in regards to the XAO being completely different now than it was 4 years ago, but isn't that difference in performance you've shown between the XAO2007 and 'XAO NOW' to be expected? The 'XAO NOW' is effectively a cherry picked version of the 2009 index. Keeping the winners and dumping the losers?

For example if I was to compare the 2012 performance of the XAO constituents on the 1st Jan 2012 and the XAO constituents on the 1st Jan 2013, the 2103 XAO will always outperform as the winners are kept and the duds dropped from the index.

Still ridiculous to think on a price sum basis we are at the 2007 highs when the XAO has us 1500 points away.

 

[IFocus]

In the 90's I think it was News Corp that moved or influenced the main index against the general market trend due to its weighting. At the time most punters were not aware of the issue.
If you are discretionary trading forever flicking through charts of the ASX you see which areas are moving anyway.

From a fundamental point the big caps will have a impact on the market general if they are increasing profits (if this is reflected in share pricing) as the small companies that connect to the business inputs / outputs will rise with them. This certainly applies to mining / banking.

 

[tech/a]

How come Tech?

In direct response to this statement

Oh you are missing the point. Which is surprising, as a system trader, are you not looking for the stocks that outperform everything else.

I said

Actually no

To me Systems trading is creating a set of Conditions with a set of Parameters and applying them to an individual instrument or a portfolio of instruments on a repetitive basis with the intention of replicating a return within the returns found during testing.
Whether stocks outperform everything else is irrelevant---Nice but not the goal.

Discretionary trading---mine anyway--- is using my skill as a trading practitioner to return as much as I can for as little risk as I can. This requires me to best apply my skills in selection of trade---and management of trade.
The skill set I use maybe and often is very different from trade to trade---entry to entry--stop to stop---exit to exit--time frame to time frame.I want to outperform on every trade---whether I do or not is proven by the market over any given time.