Index | Recent Threads | Who's Online | Search

Posts: 100    Pages: 10    Prev 1 2 3 4 5 6 7 8 9 10 Next
Last Post
New Thread
This topic has been viewed 5080 times and has 99 replies
judgetrade
Re: Improvement of my Ranking System

with 0.3% Slippage (possible with a smaller port and good trading)

Attachment aaa.png (330408 bytes) (Download count: 207)


Dec 2, 2019 4:38:22 AM       
geov
Re: Improvement of my Ranking System

Andreas,
Thanks for posting the performance from 2004 for your small/micro-cap model.

It is interesting that I can get a similar simulated performance and max drawdown with 10 positions and less micro-caps (10%). See liquidity filter rule on picture. Performance curve is very similar to your model.

Your 20+ position model's holds about 40% micro- and 40% small-caps according to your QuantStrike post. Because of the many micro-caps your trading cost (variable slippage) are very high, 17.7% of ending market value as per your Nov-18 forum posting of the trade stats. For my 10-pos small-cap model, also with variable slippage, trading costs are only 8.0% of ending market value because MktCap of stocks is greater, and possibly also because of fewer trades, 549 vs 1273.

Attachment 10 Small Caps.png (207330 bytes) (Download count: 197)


Dec 2, 2019 10:37:47 AM       
Edit 1 times, last edit by geov at Dec 2, 2019 11:24:22 AM
Nicoletta
Re: Improvement of my Ranking System

Andreas and Georg,

can I Know how many nodes have your RS?

Dec 2, 2019 11:14:29 AM       
geov
Re: Improvement of my Ranking System

Nicoletta,
As the name Multi-Factor implies, there are lots of them - 21 nodes all together.

Dec 2, 2019 12:15:22 PM       
Nicoletta
Re: Improvement of my Ranking System

Thank you Georg, I though more....

Dec 2, 2019 12:21:19 PM       
geov
Re: Improvement of my Ranking System

The annualized return can be increased from 40% to 52% with little reduction in Sharpe by using rank based weighting. Min weight is specified as 5% for this 10 position model. Reason for the increase in return is that the lower ranked stocks tend to perform worse, and therefor it is advantageous if they have less weight in the portfolio.

Attachment 10 Small Caps rank based weight.png (200867 bytes) (Download count: 176)


Dec 2, 2019 12:55:47 PM       
judgetrade
Re: Improvement of my Ranking System

Georg,
wow, super great model!!!
please give me a hint how you get to that 5% minimum weight, thank you
Best Regards
Andreas

Dec 3, 2019 9:43:12 AM       
geov
Re: Improvement of my Ranking System

Andreas,
In the Rebalance module you need to use
Sizing Method= Formula Weight
and
Position Weight Formula= pow(Rank,xx)

Depending on your ranking system adjust the power to any number that works. In my case the higher the power, the higher the return. Should work for your model as well.

Also set Weight Constraints to a min value. For a 25 pos model the equal weight is 4%, so you don't want to go much lower than 2% for your model. For my 10 pos model I set min weight to 6%, and max weight to 100%. Also you should adjust Transaction Control to reduce the number of small rebalancing transactions.

Let us know if you can improve return without lowering Sharpe too much.

Dec 3, 2019 10:57:14 AM       
judgetrade
Re: Improvement of my Ranking System

Thank you Georg, will try it out and post results!

Attached is a paper on Quality, what is robust and what not...

Attachment robust!.png (480537 bytes) (Download count: 123)


Dec 3, 2019 11:56:18 AM       
judgetrade
Re: Improvement of my Ranking System

here we go with the pow(Rank,25) weighting...

Attachment aaa.png (402678 bytes) (Download count: 116)


Dec 3, 2019 12:33:36 PM       
Posts: 100    Pages: 10    Prev 1 2 3 4 5 6 7 8 9 10 Next
 Last Post