Here is the performance of a ranking system that I have been developing for some time. Equal weighted, 10 year performance and 25 stock portfolio gives me a 60% annualised performance.
But when trying to work with dynamic weight (1/(Rankpos) ), it gives with 25 stocks:
with 10 stocks:
This seems good. (I have also tested it out of sample from 2001–2013.)
Is there something in the settings with dynamic weight that makes it act weird?
Interesting - you may want to check the leverage on the model. Dynamic weighting can result in using leverage, even if it’s not intended. You can check the “cash invested” chart in the Charts Tab. If >100% cash invested, then the model uses leverage.
You can adjust settings so that leverage is not used. I can share what I use if this is what’s happening.
Cheers,
Ryan
What can I do to avoid this?
The key here is to set leverage < 1 in Settings and some others. Here’s a screenshot:
You can then play with the settings and check with “Cash Invested” to ensure you’re not using margin (if that’s the goal).
1 Like
