Survivorship bias

Discussion in 'Strategy Development' started by indicator777, Nov 4, 2019.

  1. Is anyone using Norgate data with historical index constituent series? What is your experience? The data appear useful in avoiding survivorship bias as in this interesting article.
     
  2. Yes I've used it. Data is pretty good....I was able to reconstruct the Nasdaq 100 fairly accurately going back to 1995. Keep in mind I did the reconstruction manually....I'm not sure how accurate Norgate is for their inclusion and exclusion dates. I recommend you do a sanity check on these to avoid any bias.
     
  3. Metamega

    Metamega

    https://alvarezquanttrading.com/blo...-survivorship-free-data-changes-test-results/

    Both the article you posted and this link use the data in Amibroker which makes it super easy. They’ve built a plugin and some functions that make running a system through historical constituents a couple lines of code.

    Not sure how the data is setup outside of Amibroker plugin but as far as EOD data goes, I haven’t seen anything similar for handling index constituents or delisted data. Some might find it a bit pricey considering theirs lots of free data out there but I never hear complaints about their data so doubt theirs much cleaning to do and the thought of creating a historical index constituent list sounds like hours and hours of research.
     
  4. themickey

    themickey

    Norgate Australian data and world indexes has been flawless for me past 20 years therabouts within Amibroker. Never had a glitch and when I've needed assistance like loading data onto new computers they have bent over backward to do everything to get me running asap.
    Their downloader, never once had a problem.
     
    .sigma likes this.
  5. ronblack

    ronblack

    Forex, ETF and futures and you don't need all that expensive equity data.

    Thanks for the links anyway. The trader education section of PAL blog is gold mine.
     
    themickey likes this.
  6. themickey

    themickey

    Indeed a gold mine :thumbsup:
     
  7. I think the major cognitive defect of ET posters is an apparent inability to discriminate between expertise and utterly poseuristic bullshit. They'll have, for example, both Destriero and raVar on their follow lists without noticing the contradiction.

    RonBlack = Michael Harris

    Harris has been at this for 20 years, where's his yacht?
     
    ironchef likes this.
  8. jharmon

    jharmon

    Norgate is supporting more platforms now too. I'm now trying some of their Python libraries and the Zipline engine that is used by Quantopean. It's pretty powerful but these back-testing engines have a steep learning curve. It would be ideal if they did intra-day data too with these capabilities, but they don't.
     
  9. themickey

    themickey

    I have Dest blocked.
    As fot the yacht, gawd only knows, in here somewhere.....?
    port-hercule.jpg
     
  10. ronblack

    ronblack

    LOL I haven't posted for 6 months for exactly this reason. ET was a great forum until the Kevin Schmits and their likes took over.Kevin is one of those with the goal of convincing every newcomer to ET to quit trading. Not good for advertisers and their money.

    Here is one of his attempts to undermine one of the best books on machine learning ever written by Dr. De Prado, who was 2018 quant of the year.

    https://www.elitetrader.com/et/thre...s-coming-for-wall-street.336737/#post-4941650

    Clearly Kevin thinks he is the only one. Kevin, can you clean Prado's shoes?
     
    #10     Nov 6, 2019