The average is probably horribly misrepresented. There are a ton of super long holders, which hold for decades, and a crazy high number of HFTs running that hold for microseconds. I bet nobody is actually close to the average, 22 seconds is extremely slow for software trading and extremely fast for a human. If you're using the holding duration as a proxy for people caring about companies, I think you are probably including a massive amount of HFT timing which it's clear isn't representing company loyalty. Also, what if an HFT buys and sells the same stock 1M times. The average time would likely be very short, but the "loyalty" could be 100%. Bad metric.

Very, very few people hold stocks for multiple years, so they have no reason to care about the company's long term future:

"There are different ways of slicing it, but Reuters calculations based on New York stock exchange data show the average holding period for U.S. shares was 5-1/2 months in June, versus 8-1/2 months at end-2019."

> Codetermination in Germany

For some reason, at first, I read this as "Code termination" (not co-determination) and thought it was a mistranslation of some German compound noun, and I was confused.

TTS on Mac also reads it as "code termination"