- This topic has 10 replies, 6 voices, and was last updated 10 months, 2 weeks ago by
.
- Posts
FYI, the Nasdaq 100 stock list needs to be updated as follows:
The Nasdaq 100 Index will have new additions and deletions effective on December 19, 2022.
The stocks jumping on the index as part of the annual reconstitution are CoStar Group (CSGP), Rivian Automotive (RIVN), Warner Bros. Discovery (WBD), GlobalFoundries Inc. (GFS), Baker Hughes Company (BKR), and Diamondback Energy (FANG).
As a result of the reconstitution, VeriSign (VRSN), Skyworks Solutions (SWKS), Splunk (SPLK), Baidu (BIDU), Match Group (MTCH), DocuSign (DOCU), and NetEase (NTES) will drop off the index.
The Nasdaq low-volatility sub-strategy has a lookback period of 70 days, so we must wait 70 days with the addition of new stocks to make sure that they have been in the Nasdaq during the whole lookback period.
Rivian (RIVN) was dropped from the NASDAQ 100 effective June 20. Yet, it was one of the stocks selected under the NASDAQ 100 strategy in August and in September. Does this have to do with the 70-day lookback period you mentioned in the earlier post? I’d like to better understand why this happens. Thanks.
You are right! I forgot to delete it. The 70 day look back period only matters for new companies. I will fix it by tomorrow. Thank you.
Suppose we run a 10 year back test for the NASDAQ 100 or DOW (2014 to 2023). When evaluating 2014, does the program only use stocks that were in the index in 2014, or does it use the stocks that are in the index in 2024?
You will have only the latest stocks in it so a 10 year back test is too long. However the stocks falling out of the Nasdaq100 are nearly never used in the strategy as they represent rather the worst performing stocks and the strategy selects good performing stocks.
Is there a plan to use the constitution at the historical rebalance day to construct the ranks? I’m just thinking NASDAQ 100 itself is like a stock selection portfolio, so what is in right now is after the selection. At the time of every rebalance, if we use the current constitution, we are backtesting on a current portfolio – if that makes sense ? Thank you.
No, it would be very complicated to do this in QuantTrader. We would need to set up a database with all the changes of the NASDAQ index for the last 10 years.
Thanks for your reply. In this case, here is an easy way help investors better understand the profile of the strategy. Instead of QQQ, a better comparison is from a new benchmark ETF that has all current NADAQ 100 stocks equally weighted – and of course, you need to give a new equal weight if a stock didn’t exist before some period. If you can make that benchmark ETF available, that will be very helpful.
Chyylq, they don’t use a data source that includes delisted stocks. So, yes, their backtesting will suffer from some survivorship bias. This is mitigated somewhat by their walk-forward testing. What I do to evaluate their strategies is go to the very bottom of the page for the strategy or portfolio, and there is a link for the actual historical performance of the portfolio or strategy. As you might expect, those metrics fall short of the backtest results, perhaps because of the survivorship bias, perhaps because of overfitting. Still, the historical numbers look quite good for many of their strategies and portfolios. The only strategy I particularly dislike is their crypto strategy. Since crypto has such crazy runs, I suspect their clients were after them to come up with a crypto strategy, but I would rank it least desirable, at least at this point.
Thank you very much, @Peticolas! Great to hear you inputs on the strategies and how to use the resources.
- You must be logged in to reply to this topic.