Home › Forums › Logical Invest Forum › Survivorship bias
- This topic has 10 replies, 3 voices, and was last updated 4 years, 9 months ago by
alimcor.
- AuthorPosts
- 06/07/2020 at 8:05 pm #78987
alimcor
ParticipantHi, for Nasdaq 100 and Dow 30 strategies, the backtests take into account index past and current constituents (avoiding Survivorship Bias)? Thanks
06/09/2020 at 7:22 pm #78996Midnight122
ParticipantYou taking the CAIA exam tomorrow?
06/09/2020 at 7:26 pm #78998alimcor
ParticipantWhat?
06/09/2020 at 7:30 pm #78999alimcor
Participant06/09/2020 at 8:56 pm #79002alimcor
ParticipantFor dow30 this would not be a big problem, but it would be for nasdaq100. I would like to know the opinion of the administrators of the site. Thank you
06/10/2020 at 6:15 am #79007Vangelis
KeymasterWe are well aware of survivorship bias but both the Nasdaq 100 and Dow30 backtests do not take it into account. That means that long term backtested are optimistic. On the other hand the Nasdaq 100 has been ‘live’ for a few years now and has gathered historical performance data that is indeed survivorship-bias free.
06/10/2020 at 9:30 am #79009alimcor
Participantcan’t you backtest with Norgate premium data?
06/10/2020 at 9:56 am #79010Midnight122
ParticipantAlimcor – you seem high maintenance for a basic membership man. If you’re so resourceful why don’t you backrest it. Just asking no offense.
06/10/2020 at 10:03 am #79011alimcor
Participantbecause I don’t know the rules of the systems.
06/10/2020 at 10:27 am #79012Vangelis
KeymasterYou could do that using Norgate data and Amibroker. Unfortunately it is not that simple to input the data to QuantTrader, our proprietary software since we would have to create a whole ‘subsystem’ from scratch just to handle that data. It’s probably easier to create the N100 startegy in Amibroker and run the backtest there. But this too is complicated and inefficient since the N100 strategy (and most LI strategies) are not a good match for the AB ‘vector type’ logic. We have run our own private tests and our conclusion is that it’s not worth the effort. We have survivorship-free out-of-sample data since July 2015 when N100 went live that seem to validate the logic.
https://logical-invest.com/app/strategy.php?symbol=NAS100! (you need to be registered to view)
06/10/2020 at 10:58 am #79015alimcor
Participantok thanks for your reply and for your excellent work on the site
- AuthorPosts
- You must be logged in to reply to this topic.