Intro
I am conscious of how WallStreetBets-y this publish sounds, and to be honest I did publish one thing comparable there earlier at this time. However bear with me – I attempted to be extra sober with this evaluation.
I made this Inverse Cramer Index, which is thrashing the S&P by *20% YTD – I additionally made this publish asking people what I ought to do subsequent, and the best upvoted remark was a portfolio that invested in shares primarily based on South Park mentions. So I constructed it.
The Query
Is a portfolio that invests in shares when South Park mentions them a viable funding technique?
The Information
I initially used the South Park Fandom web site to try to brute drive scrape each episode script and evaluate to each firm title within the NYSE/NASDAQ database. This picked up a variety of shit although (and it additionally did not choose up a variety of stuff that it ought to have, like a point out of Campbell’s Soup within the 2015 episode Pink Man’s Greed), so had to determine a distinct option to get all that knowledge. I obtained artistic – came upon that Wikipedia has a abstract for each episode, and hyperlinks to the corporate names, so I constructed out a script that pulled tickers by scraping for these firm names.
The Technique
I attempted a bunch of issues. As a result of the indicators we get from the present are solely purchase indicators, we’ve to create the promote indicators ourselves primarily based on some kind of unbiased metric. First, I attempted making just a few funding home windows, aka promoting all purchase picks after x days for the reason that final purchase point out – I attempted with 30 days, 60 days, 120 days, and 360 days.
Most funding home windows have been on par or barely outperformed the S&P. The iteration that labored finest was promoting utterly as soon as the following episode point out happens, so that you simply’re at all times absolutely invested in the newest episode’s picks. This finally ends up exposing you fairly strongly to South Park as a stockpicker aka sign producer, which I assume is strictly what you need. If I have been doing this in an precise portfolio just like the Inverse Cramer, I would do the funding home windows, albeit perhaps a bigger window, better than twelve months, simply due to tax causes.
The Rebalances
Date | Shares |
---|---|
2000-12-13 | ‘DENN’ |
2000-12-13 | ‘DENN’ |
2001-11-07 | ‘DIS’ |
2003-04-30 | ‘CPB’ |
2004-11-03 | ‘WMT’ |
2006-10-04 | ‘BBY’ |
2008-03-19 | ‘CHH’ |
2008-04-16 | ‘SBUX’ |
2008-10-29 | ‘BBY’ |
2009-03-11 | ‘DIS’ |
2009-03-25 | ‘AXP’ |
2009-10-07 | ‘CMG’, ‘MCD’, ‘TWTR’ |
2009-10-21 | ‘WWE’ |
2009-11-04 | ‘HOG’ |
2010-03-24 | ‘TWTR’ |
2010-10-13 | ‘TWTR’ |
2010-11-03 | ‘NKE’ |
2010-11-10 | ‘NKE’ |
2010-11-17 | ‘PGR’ |
2011-04-27 | ‘AAPL’, ‘BBY’ |
2011-05-18 | ‘FDX’ |
2011-11-02 | ‘RRGB’ |
2012-10-03 | ‘WMT’ |
2012-10-10 | ‘AMZN’, ‘UPS’ |
2012-10-31 | ‘NKE’ |
2012-11-07 | ‘DIS’ |
2013-11-13 | ‘SNE’ |
2013-11-20 | ‘MSFT’, ‘SNE’ |
2013-12-04 | ‘MSFT’, ‘RRGB’ |
2014-10-15 | ‘LYFT’, ‘TSLA’, ‘UBER’ |
2014-11-05 | ‘TWTR’ |
2014-11-12 | ‘BBY’ |
2014-12-03 | ‘TWTR’ |
2014-12-10 | ‘TWTR’ |
2015-10-14 | ‘YELP’ |
2016-09-21 | ‘TWTR’ |
2016-09-28 | ‘TWTR’ |
2017-09-13 | ‘TWTR’ |
2017-10-11 | ‘NFLX’ |
2017-10-25 | ‘ROST’ |
2017-11-08 | ‘BYND’ |
2017-11-29 | ‘NFLX’ |
2018-12-05 | ‘AMZN’ |
2018-12-12 | ‘AMZN’ |
2019-10-02 | ‘AAPL’, ‘DIS’ |
2020-09-30 | ‘BBW’, ‘VIAC’ |
The Efficiency
South Park destroys. Not what I used to be anticipating in any respect. A rise of about 1100% over 7 years (together with the current downturn), in comparison with near 200% for SPY and 300% for QQQ. This bakes in a couple of 50% CAGR for the previous 7 years, with a Sharpe of about 1.4. Remember, although, that this appears to basically be a “purchase $BBW and $VIAC in the course of the pandemic” technique that causes this a lot outperformance – in any other case, the outcomes would doubtless be just a few proportion factors of outperformance, primarily based on the South Park picks beating, however basically monitoring the opposite indexes.
Dialogue/Conclusion
No fucking clue lmao – I’d by no means make investments primarily based on South Park knowledge. Possibly that is why I am poor although. On a extra severe word, the way in which many hedge funds get their supposed “edge” is by paying some huge cash for different datasets like this. Some others that I’ve heard of are car parking zone fill knowledge, firm mentions on Instagram posts weighted by follower depend, and I would positively assume WallStreetBets sentiment evaluation monitoring. Every other different datasets you have been inquisitive about?