Deadline for the GPM Precious Metals Competition

Answered
99
6
created on 2021-01-04 16:33:25

Comments (6)

Petitfort · about 1 week ago

Is there any function gives the information ratio metric directly(not in the backtest function)?

Yongcheng · about 1 week ago

Hello,


Theres a function called informationRatio in R but its not yet in python. It wouldnt be too difficult to derive though.

Yongcheng · about 1 week ago

Dear All,


To add on to Manas:


For your final version of your notebooks, please also make sure that your models are 'explainable'. If you are choosing to use a smaCross of unconventional parameters like short = 23, long = 164, there should be some sort of logic behind it.


i.e. you have to convince us that there is not a huge risk of overfitting and that you have taken the necessary steps to prevent it.


Bottom line is that we need to be confident that your strategy will produce results in the months/years to come for us to license it to the client.


Feel free to reply/start a new topic to clarify.


All the best!

Petitfort · about 1 week ago

Thank you Yongcheng.

Petitfort · about 1 week ago

Hello Yongcheng, i see in my last submission notebook the sharpe before and after transaction cost is 0.75 and 0.58, but in the dashboard is only 0.4. We used a different peroid of historical data for the public dashboard?

Yongcheng · about 1 week ago

Hello,


Yes, the time period we use for the public leaderboard is 2003::2016. You might have been backtesting from 2003::2014 in your notebook