PSA: The market doesn’t care who wins

by quantize_me

View post on imgur.com

TL;DR: I compared how the price of $SPY compares with Biden/Trump’s odds of winning and found there was no correlation whatsoever. I will explain how I did this like you are retarded, because you are retarded.

Trades: Directional bets that aren’t on specific stocks/sectors are dumb. Trade condors or something if you want to play SPY.

We want to answer the question: Does the $SPY get uppies or downies when Biden’s odds of winning go up?

Look below to see how Kamala’s Biden’s (and inversely Trump’s) odds on PredictIt compare with the price of $SPY. If we assume the gambling machines of the market and PredictIt are reasonably efficient, then traders are betting on the same information at the same time. If there is a relationship between these time series we can pull it out. Moreover, since we don’t expect there to be any lag in information response we don’t have to do anything fancy like autocorrelation. (Note: there is some weirdness in that PredictIt data is polled at midnight ET and market data is at market close, 4pm ET).

preview.redd.it/zjotkdw3qxw51.png?width=1568&format=png&auto=webp&s=5eed405d96cd2a6d6549fb2f0485f972b1b36d16

To find this relationship, we can instead see how the daily changes are related to each other. Plotted below are those daily changes over the last ~4 months.

preview.redd.it/rls75cl4qxw51.png?width=1596&format=png&auto=webp&s=f3cac35ffd43834262e83cbfc39de4523c4c784c

Now, we can plot the changes in odds and the changes in $SPY price directly against each other to extract the line that shows how the two are related.

preview.redd.it/xpslxto5qxw51.png?width=1618&format=png&auto=webp&s=ac2838a853bf0a6ebbbb868e62a304435aa884ad

We are primarily funded by readers. Please subscribe and donate to support us!

Uh oh! The R squared is 0.003. R squared is a number that tells us how well the line fits the data. If it’s close to one, then the line is very good. If it’s close to zero, then the fit is very bad. This fit is very bad. There does not appear to be any relationship between which candidate is expected to win and the price of the S&P 500.

You might say “Look, the line points slightly down. Kamala Biden is bad for the market.” But then we can look at the uncertainty of that slope. The slope uncertainty is twice the reported slope. The slope reported by the fit estimate is not meaningful.

preview.redd.it/v2qbzen6qxw51.png?width=774&format=png&auto=webp&s=cca6e7c4cec35ccef90705672e2fbef5502dba74

Positions: Cash. Sorry.

Edit: I’ll take up to 5 requests, one per person to do the same analysis on individual tickers. A few examples below.

TSLA

GUSH

TAN

GEO This one’s the opposite of what I expected and with the least relative uncertainty

Disclaimer: This information is only for educational purposes. Do not make any investment decisions based on the information in this article. Do you own due diligence or consult your financial professional before making any investment decision.

Views:

Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.