CFD pricing and the XJO/SPI

[Punta]

The XJO does not have anything like the large gap in price between the close of one day's trading and the start of the next day that individual stocks or the SPI can have.

Yeah I think I get most of this - the price jumps in XJO that are associated with up/down gaps in the individual stocks that make up the index, will be spread over the staggered open

.....

It may therefore perform well in backtesting when a tradeable security fails.

Yeah for sure, just exploring the possibilities here...

 

[elbee]

I still need to understand how a CFD can trade when it's underlying is undefined, if the price of the CFD itself is supposed to be locked to the underlying.

As skc pointed out before the CFD price won't necessarily reflect the underlying. Certainly not in the case of an index which is not itself tradeable.

 

[Punta]

As skc pointed out before the CFD price won't necessarily reflect the underlying. Certainly not in the case of an index which is not itself tradeable.

Ah right, I'm a slow learner!

So the CFD is a contract for the difference in the futures price, which is always well defined, and not for the difference in the index itself.

To test how well the futures price matches the index, it would make sense to put together a time series of whichever future contract was nearest to expiry, and plot that against the XJO?

I have just pulled a few months of data for the SPI future that expires in Dec 2011, and compared the closing price for each day, with the close of the XJO. They look pretty similar, considering the close is 30 mins apart, especially as the time approaches the expiry.



Unfortunately I can't really test the open-to-close change, because the opening times of the SPI are so different from the XJO...

Anyway, thanks for the heads up/tips - I will forward test with futures and CFD, and see how it goes...

 

[Punta]

Hmm I think my intraday index dreams might be over. Looking for a product that tracks the XJO, I just extracted data for APZ1 from IB, and compare it with the XJO.

During the last month, the XJO close matches APZ1 at 4pm quite well:



However, the XJO open does not match the APZ1 at 9.50 am at all



The APZ1 has been trading all night, so it already contains information that is going to move the XJO during the day.

Bottom line: you can't trade the index, so there's no point looking for open-to-close strategies for an index? Lower frequency, or higher frequency index strategies might be worth pursuing I would guess because there are products that more closely track the index at these periods...

 

[nomore4s]

Hmm I think my intraday index dreams might be over. Looking for a product that tracks the XJO, I just extracted data for APZ1 from IB, and compare it with the XJO.

During the last month, the XJO close matches APZ1 at 4pm quite well:

However, the XJO open does not match the APZ1 at 9.50 am at all

The APZ1 has been trading all night, so it already contains information that is going to move the XJO during the day.

Bottom line: you can't trade the index, so there's no point looking for open-to-close strategies for an index? Lower frequency, or higher frequency index strategies might be worth pursuing I would guess because there are products that more closely track the index at these periods...

There are no free lunches in trading, and if there are they are normally exploited very quickly by the pros.

The big problem you will get in doing testing on the XJO is the staggered opens, it makes any open data you use unreliable. Even if the futures contract didn't trade all night the opening gaps pretty much take care of any edge you might think you have, look at the HSI futures contract and see how the overnight gaps account for huge chunks of the overall moves even though the HSI still has very good daily ranges.

You are correct, you can't trade an index so no point testing trading strategies on it but why not do testing on actual products you can trade like a futures contract? There are edges to exploit in those markets.

 

[Punta]

You are correct, you can't trade an index so no point testing trading strategies on it but why not do testing on actual products you can trade like a futures contract? There are edges to exploit in those markets.

Yeah I have a few equity strategies paper trading at the moment, which seem to be ticking along alright. They're not particularly scalable though.

In looking for a scalable strategy, I just assumed that I would be able to find a product for an index-based strategy. Turns out it's not that simple!!

 

[Tannin]

(Oh dear! First post here and I'm dragging up an old, old thread .... This is not a good look. No matter, it's a point no-one else seems to have raised so far and it really ought to be mentioned, so I'll post it anyway. I don't usually drag up ancient history. Honest!)

If you are looking for a security that tracks the S&P ASX200 closely, simply buy a basket of the underlying securities.

Although the ASX200 seems from its name to track 200 different companies, in reality the largest 8 or 10 of them account for by far the greater part of it. BHP alone is one eighth of the XJO. In fact, you can just about throw away the smallest 190-odd constituents and just consider the big four banks, Telstra, the two supermarket giants, and the big two miners. That's more than three-quarters of the index right there. And you don't need to buy all nine. Surely you could pick just one of the banks as a proxy for the other three, don't worry about Rio (the BHP price is much more important, and they both depend on iron ore demand and prices more than any other factor so they tend to go up and down together), and a single supermarket. At this point we have just four shares to buy - BHP, Telstra, Westpac, and Wesfarmers. (Or BHP, Telstra, Woollies and NAB, whatever.)

Depending on how exactly you want to track the index, you could throw in one or two other stocks, perhaps by deciding which two other sectors need a small representation to mimic the index more closely. But even with just four stocks, you are holding about half the value underlying the index, so you are already going to be in the ballpark. And with ordinary shares to trade instead of CFDs or options or ETFs or any other fancy stuff, you have no problems with extra brokerage or greedy spreads or lack of liquidity and life is simple. Simple is good.

 

[Joules MM1]

some stuff from the weekend