Edit: Here's the graphs for the people who can't read:

Calls

Puts

​

Intro

Hey everyone! So some background – I make indexes/strategies that sound really fucking WallStreetBets-y but have a little bit of sound analysis behind them. I'm the guy who built the Inverse Cramer ETF, as well as the South Park investing strategy. Tl;dr – these meme strategies seem to work (I'm working on turning these into real ETFs, because they've gotten considerable attention)

What I wanted to see next, based on some WallStreetBets comments, was whether I could find either an equity or an options strategy based on the JPow's speeches that seemed to have worked historically. Here's what I did:

The Data

I took all the data I could find on Powell's speech dates from C-Span. Ended up being quite a few hundred videos. I wasn't looking for any bullish/bearish content from the speech. The signal I was looking to test for with this strategy was – can we glean any market information from the time of JPow's speeches, and/or do market participants act differently shortly after his speech?

The Strategy

Fuck JPow right? So anyways, I tracked the S&P with 100% of the portfolio. Whenever JPow spoke, I reallocated 15% of the portfolio into daily puts (at the specific time of the speech). Used the same purchase price as strike price, so I benefit from a directional bet rather than a magnitude one.

I also did this with strategy with calls for fun. First time I've done an options strategy, so any feedback here is welcome!

Execution

In execution, it looked like this:

• Took all the dates that JPow was speaking
• Used the adjusted close/open to factor in dividends and splits
• If you've followed along my posts before, I always start off with a naive algorithm as a sanity check. If I'm able to get some sort of signal without doing much work on structuring the data, that makes me happy because there might be some kind of strong signal that you don't have to do statistical magic for
  • The naive algorithms used equities. What if we bought the SPY the day of JPow's speech, then sold either at end of day, or at open the next day
  • This time, the naive algorithm failed. Becaue you're mostly out of the S&P, you don't get all the upside of the last 4 years. In hindsight, the naive algorithm (and using equities) was pretty fucking stupid. You'd have to have the thesis that the S&P has the biggest jumps only on days JPow speaks, and is otherwise flat or in a bear market

Anyways, then I tried a daily options strategy. When JPow's speech is released on C-Span, I bought an option the next hour, and exercised by the end of day.

The Performance

Naive 1: Just buying on speech date

Naive 2: Buy the day after, sell end of day

Again, the naive ones didn't do BAD – they just weren't exciting, and were really what you'd expect. Staying in cash most of the time and not making any big bets on direction means you're going to stay relatively flat (therefore beat probably 95% of WSB)

PUT Portfolio

This was the strategy that reallocated 15% into puts right at the start of JPow's speech (or start of the next trading day), and exercised/let expire at the end of the closest trading day.

CALL Portfolio

This portfolio did worse than the PUT portfolio when you look at the end of our trading period vs the start, but it did have some really, really good moments in 2020.

Conclusion

So, yeah – JPow doesn't fucking have predictive power. Was a pretty stupid idea in hindsight. What we did learn here is that JPow's speech dates are very much a confounding factor – he speaks during times of crisis, meaning following them is a play on trading volatility.

The best strategy based on the results here seems to be buying puts on a daily basis when there are inklings of panic and volatility, and then buying calls when you're sure people are panicked, scared, and volatile. You'll lose your bet most days, but when you're right, you're right big. Thoughts? Any other conclusions to be made?