Forums  > Trading  > Would you trade a product-specific anamoly?  
     
Page 1 of 1
Display using:  

riskPremium


Total Posts: 24
Joined: Nov 2018
 
Posted: 2020-02-16 03:51
I recently discovered a signal which seems OK (t>4 with a smooth equity curve) on one of the commodity futures. It also works reasonably well on a correlated one. Beyond that, performance becomes mediocre on rest products.

Only after this discovery did I come up with a theory for it, which is not convincing enough for me.

Generally I believe a robust model should work on a broad class of products without too much parameter tuning.

What would you do with similar situation? Ditch the signal entirely? Perhaps allocate a very small proportion of the portfolio?

gaj


Total Posts: 102
Joined: Apr 2018
 
Posted: 2020-02-16 14:53
walk forward validation. see if the signal persists for another 3 months.
Previous Thread :: Next Thread 
Page 1 of 1