That's only true if you force all shorts to be covered at once without a chain of trades. That's not how it happens.

Person A covers their short by buying a share from Person B and returning to Person C. Person D then buys that share from Person C and returns to Person E to cover their short. That's 2 short shares covered with a single underlying share and no failure to deliver.

Yes, the SEC does track failure to deliver, but >100% short interest does not mean there is naked shorting nor does it imply there will be failure to deliver.

"Naked shorting is the illegal practice of short selling shares that have not been affirmatively determined to exist"

- https://www.investopedia.com/terms/n/nakedshorting.asp

> that extra 40% will fail to deliver, because those shares don't exist

This is not true because all shorts don't have to be covered simultaneously.

> The SEC actually keeps a list of trading companies with high rates of failure to deliver as a means of detecting naked shorting

Funnily enough, $GME had very high failure to deliver rates in December [0] but this is not necessarily due to the short interest.

[0] https://www.reddit.com/r/wallstreetbets/comments/l97ykd/the_...