Australian (ASX) Stock Market Forum

Developing a mechanical system from scratch

Have you tried testing an ORB on the Kospi?

I swing trade the Kospi, usually 10-15 trades/day but often let one or 2 contracts run from the open if it moves strongly. Perhaps I should just test it myself :) Just curious to know if you've had a look at the Kospi yet?
 
Have you tried testing an ORB on the Kospi?

I swing trade the Kospi, usually 10-15 trades/day but often let one or 2 contracts run from the open if it moves strongly. Perhaps I should just test it myself :) Just curious to know if you've had a look at the Kospi yet?

I'm getting around to testing it as well, but I've only got a little data for it as its IB data. I thought of getting some tick data for it but it looked too counter trend to me, I'll test it and see what its like. Being a TF system you can certainly find out which markets are trendy and which are CT.

CanOz
 
Being a TF system you can certainly find out which markets are trendy and which are CT.

CanOz refresh my memory on your time frames for these systems.
I basically only design mean reversion systems for time frames under 1 hour and consider TF Systems in the Longer Time Frames:2twocents
 
Its got potential, but i need more data.

This is based on the first 10 minute range.

Cheers,

CanOz
 

Attachments

  • K200 ORD.PNG
    K200 ORD.PNG
    73.7 KB · Views: 5
Its got potential, but i need more data.

What are you using? Tick or 1 min? I got both thats timestamped to Oz hours back to 2007-8ish. Its in NT format,

20120305 051500;264.55;264.6;264.45;264.55;473
20120305 051600;264.55;264.6;264.5;264.55;333
20120305 051700;264.55;264.55;264.5;264.55;181
20120305 051800;264.55;264.6;264.5;264.6;435
20120305 051900;264.55;264.6;264.5;264.55;451
20120305 052000;264.55;264.65;264.55;264.55;839
20120305 052100;264.55;264.65;264.5;264.65;357
20120305 052200;264.6;264.6;264.5;264.5;436
20120305 052300;264.55;264.6;264.5;264.6;620
20120305 052400;264.6;264.65;264.55;264.6;362
20120305 052500;264.6;264.6;264.55;264.55;105
20120305 052600;264.55;264.6;264.55;264.6;284
20120305 052700;264.6;264.6;264.55;264.6;173
20120305 052800;264.55;264.65;264.55;264.6;304
20120305 052900;264.6;264.6;264.5;264.55;583
20120305 053000;264.55;264.6;264.55;264.6;260
20120305 053100;264.6;264.65;264.55;264.6;485

Can email it to you if you want.
 
CanOz refresh my memory on your time frames for these systems.
I basically only design mean reversion systems for time frames under 1 hour and consider TF Systems in the Longer Time Frames:2twocents

I'm using a 1 minute chart to test because it enters next bar at market, but you could use anything up to 15 minute i suppose. One would wonder why you would use anything more than that unless you planned to hold overnight (that's something we can try too).

From what I've learned with it, most markets make their move in the first 15-30 minutes, if they move much.

I think it would perform better with a more clever trailing stop....perhaps you guys may have a more experience view on that than I.

The Kospi seems to trend and then reverse part way through the session, maybe when another market opens, the HSI maybe? Perhaps it would benefit from a time stop?

CanOz
 
What are you using? Tick or 1 min? I got both thats timestamped to Oz hours back to 2007-8ish. Its in NT format,

20120305 051500;264.55;264.6;264.45;264.55;473
20120305 051600;264.55;264.6;264.5;264.55;333
20120305 051700;264.55;264.55;264.5;264.55;181
20120305 051800;264.55;264.6;264.5;264.6;435
20120305 051900;264.55;264.6;264.5;264.55;451
20120305 052000;264.55;264.65;264.55;264.55;839
20120305 052100;264.55;264.65;264.5;264.65;357
20120305 052200;264.6;264.6;264.5;264.5;436
20120305 052300;264.55;264.6;264.5;264.6;620
20120305 052400;264.6;264.65;264.55;264.6;362
20120305 052500;264.6;264.6;264.55;264.55;105
20120305 052600;264.55;264.6;264.55;264.6;284
20120305 052700;264.6;264.6;264.55;264.6;173
20120305 052800;264.55;264.65;264.55;264.6;304
20120305 052900;264.6;264.6;264.5;264.55;583
20120305 053000;264.55;264.6;264.55;264.6;260
20120305 053100;264.6;264.65;264.55;264.6;485

Can email it to you if you want.

Thanks TH...

1 minute would do.

Perhaps 6 months for a start, then if its promising whatever you got:D.

Can i map that with an ASCII mapper? Waza, you ever convert your NT data for your MultiCharts using the ASCII mapper?

CanOz
 
No not as yet ,as I understand it much easier to get data into Multicharts than the other platforms
 
I use this technique. It used to be difficult to play along, now there is an ETF for almost everything, it is quite easy.

http://www.cxoadvisory.com/43/big-ideas/beat-the-market-with-hot-anomaly-switching/
http://www.cxoadvisory.com/54/big-ideas/diversifying-across-equity-anomalies/

All my systems have a switch, sometimes implicit, sometimes explicit which stops the system trading if the market regime is not appropriate for that sort of trading.

Sinner,

Is there an inverse ETF available for the S&P500? I have searched but am unable to find one. I would like to create simple rebalance strategy using an inverse ETF and an ASX stock. Index falls buy more of the ASX stock. Stock rises buy more of the index. Keep it 50/50 and just rebalance it every month. Is this possible? Or do i have to use CFD type instrument to short the S&P500?

5 minutes of your time would be most appreciated.

Cheers

Oddson
 
Sinner,

Is there an inverse ETF available for the S&P500? I have searched but am unable to find one.

Cheers

Oddson

SPXU - Proshares UltraPro Short S&P500
 
I think bonds (ZN) deserve a better look with more data.

CanOz
 

Attachments

  • ZN ORB.PNG
    ZN ORB.PNG
    64 KB · Views: 4
Sinner,

Is there an inverse ETF available for the S&P500? I have searched but am unable to find one. I would like to create simple rebalance strategy using an inverse ETF and an ASX stock. Index falls buy more of the ASX stock. Stock rises buy more of the index. Keep it 50/50 and just rebalance it every month. Is this possible? Or do i have to use CFD type instrument to short the S&P500?

5 minutes of your time would be most appreciated.

Cheers

Oddson

There are heaps, very surprised you didn't find any considering how many there are. They are all listed on the NYSE or other US exchanges. SH is the 'correct' unleveraged instrument, the one canoz listed is leveraged.

It really depends on your timeframe, if it's less than 2-3 months you should really just use a futures contract imho.

Don't forget currency differences.
 
Well, although my system does ok on the SPI, it was mainly using the gaps from overnight trading and i think the typical SPI system performs better.

It failed to produce a profit over 3 years of 10 year T-note data. While i am optimizing the trailing stop on the DAX at the moment i don't really have any hope that it will be a viable system.

So looks like its back to the drawing board again...the search for an intra-day index system continues.

I think it would be easier to just buy a system:banghead:.

CanOz
 
Well, although my system does ok on the SPI, it was mainly using the gaps from overnight trading and i think the typical SPI system performs better.

It failed to produce a profit over 3 years of 10 year T-note data. While i am optimizing the trailing stop on the DAX at the moment i don't really have any hope that it will be a viable system.

So looks like its back to the drawing board again...the search for an intra-day index system continues.

I think it would be easier to just buy a system:banghead:.

CanOz

Not all markets are the same, multiplied by a million on the intraday scale. e.g. what I try and do in forex performs pretty badly in the local market equity index. In fact I don't even treat pairs the same, AUDUSD or EURJPY trading in Tokyo is different to EURUSD on the London Open GBPUSD on the LO as well! Liquidity, interest, fixing times, etc...Why expect the same results from t-notes as SPI? :2twocents
 
Not all markets are the same, multiplied by a million on the intraday scale. e.g. what I try and do in forex performs pretty badly in the local market equity index. In fact I don't even treat pairs the same, AUDUSD or EURJPY trading in Tokyo is different to EURUSD on the London Open GBPUSD on the LO as well! Liquidity, interest, fixing times, etc...Why expect the same results from t-notes as SPI? :2twocents

More expecting a profitable result on the T-notes, not needed to be similar. To me if its robust it should be able to be traded on many different markets.

I see you point too about the times but this system enters at about the same time each session.

CanOz
 
More expecting a profitable result on the T-notes, not needed to be similar. To me if its robust it should be able to be traded on many different markets.

I see you point too about the times but this system enters at about the same time each session.

CanOz

Man if its a runner you're gonna get hit trying to fade it every day and if it likes to move up a tick before giving it back then you got some guts hanging around waiting for it to run. Make hay where the sun is shining.

Courtesy bloomberg here is 3 years of MOVE index, does your t-notes equity graph look like this?

Selection_002.png
 
Man if its a runner you're gonna get hit trying to fade it every day and if it likes to move up a tick before giving it back then you got some guts hanging around waiting for it to run. Make hay where the sun is shining.

Courtesy bloomberg here is 3 years of MOVE index, does your t-notes equity graph look like this?

Yes, very similar, except it came back recently, almost even again. Basically the curve was underwater the whole time.

CanOz
 
Yes, very similar, except it came back recently, almost even again. Basically the curve was underwater the whole time.

CanOz

Q.E.D

You are trying to long intraday volatility in a market where the Fed is selling puts all day long.

Tried it on GBPUSD or the HSI?
 
Top