Building the ‘Musk Proximity Metric’

​​“[T]he way finance works now is that things are valuable not based on their cash flows but on their proximity to Elon Musk”

-Matt Levine describing the “Elon Markets Hypothesis”

In order to improve the efficiency of “Elon Proximity Pricing”, we shouldn’t just be looking at how much Musk has acknowledged a particular publicly traded company, we also need to attempt to judge the likelihood that a company receives future acknowledgment.

While there are a number of qualitative ways to attempt this, I tried to implement a quantitative method by examining the degrees of separation between different companies’ Twitter accounts and @elonmusk.

Looking at Elon’s profile, the only publicly traded company he follows is the one he is CEO (or Technoking) of - Tesla. Using the Twitter API, and the list of publicly traded companies Twitter handles we compiled for our Twitter dashboard, we can parse through the following lists of the accounts that Musk is following in order to see which companies are two degrees removed from Elon on Twitter.

We found 168 of these companies — here is a graph of them broken down by sector:

Among the accounts within two degrees of connection, it makes sense that there would be some variation in connectedness. In order to attempt to quantify this, I decided to look at the number of 2-degree connections between Musk and these publicly traded companies. I began by filtering down to companies with more than one path, which cut the list down to 44.

Looking at how many paths to Musk each of these 2nd-degree connections has doesn’t produce a very satisfying result. Companies like $TWTR and $NYT are near the top, but that’s likely more a product of their general popularity than it is due to any particular closeness to Musk. It’s clear that we’ll want to develop a more advanced metric than just the number of 2nd degree paths to Musk.

A naive approach is dividing the number of paths to Musk by the companies’ number of followers, but that also yields a rather unsatisfying result (shown below) as companies with large follower counts seem to be too heavily penalized. I’m not really willing to accept any result that doesn’t have Tesla near the top.

After tinkering around a bit, I came up with the (very subjective) “Elon Proximity Metric”:

(2nd-degree paths to Elon) / Log(Followers)⁵

This gives the following as the 15 closest publicly traded companies to Musk on Twitter:

Note that “Elon Proximity” is measured solely based on his nearness to the company through Twitter follows, and does not necessarily represent how much Elon has discussed the company or how closely related their business is to his.

Unfortunately, we didn’t have any historical data to produce a bias-free backtest of a strategy based upon our new metric, but we’ll be tracking the performance of a portfolio comprised of the these top 15 stocks going forward at quiverquant.com/sources/elon.