Intro

I'm aware of how WallStreetBets-y this post sounds, and to be fair I did post something similar there earlier today. But bear with me – I tried to be more sober with this analysis.

I made this Inverse Cramer Index, which is beating the S&P by *20% YTD – I also made this post asking folks what I should do next, and the highest upvoted comment was a portfolio that invested in stocks based on South Park mentions. So I built it.

The Question

Is a portfolio that invests in stocks when South Park mentions them a viable investment strategy?

The Data

I originally used the South Park Fandom site to try and brute force scrape every episode script and compare to every company name in the NYSE/NASDAQ database. This picked up a lot of shit though (and it also didn't pick up a lot of stuff that it should have, like a mention of Campbell's Soup in the 2015 episode Red Man's Greed), so had to figure out a different way to get all that data. I got creative – found out that Wikipedia has a summary for every episode, and links to the company names, so I built out a script that pulled tickers by scraping for these company names.

The Strategy

I tried a bunch of things. Because the signals we get from the show are only buy signals, we have to create the sell signals ourselves based on some sort of unbiased metric. First, I tried making a few investment windows, aka selling all buy picks after x days since the last buy mention – I tried with 30 days, 60 days, 120 days, and 360 days.

Most investment windows were on par or slightly outperformed the S&P. The iteration that worked best was selling completely once the next episode mention occurs, so that you're always fully invested in the most recent episode's picks. This ends up exposing you pretty strongly to South Park as a stockpicker aka signal producer, which I guess is exactly what you want. If I were doing this in an actual portfolio like the Inverse Cramer, I'd do the investment windows, albeit maybe a larger window, greater than 365 days, just because of tax reasons.

The Rebalances

The Performance

https://i.imgur.com/waNPQAp.png

South Park destroys. Not what I was expecting at all. An increase of about 1100% over 7 years (including the recent downturn), compared to close to 200% for SPY and 300% for QQQ. This bakes in about a 50% CAGR for the past 7 years, with a Sharpe of about 1.4. Keep in mind, though, that this seems to essentially be a “buy $BBW and $VIAC during the pandemic” strategy that causes this much outperformance – otherwise, the results would likely be only a few percentage points of outperformance, based on the South Park picks beating, but essentially tracking the other indexes.

Discussion/Conclusion

No fucking clue lmao – I would never invest based on South Park data. Maybe that's why I'm poor though. On a more serious note, the way many hedge funds get their supposed “edge” is by paying a lot of money for alternative datasets like this. Some others that I've heard of are parking lot fill data, company mentions on Instagram posts weighted by follower count, and I'd definitely assume WallStreetBets sentiment analysis tracking. Any other alternative datasets you've been curious about?