Have is said to come from the PIE word that means to keep. (Habere in latin also comes from PIE for to take.) I have been thinking that, because some words such as street have been borrowed from Latin into later Proto-Germanic, considering the similarities, could have come from Latin habeo/habere?
I think the b could have changed into f and then a v. The 2nd and 3rd happens from PIE *p a lot.
English "have" is not cognate with Latin "habere" - even though they seem very close to each other.
English "have" is from PIE *kap- "to seize, take," cognate with Latin "capere." Grimm's law shifted PIE *k to proto-Germanic *x, which became English /h/.
The Latin habere's etymology is a bit less attested, but might come from PIE *ghabh-, which also means "to seize, take."
Could they? Sure, it's possible. "Yeet" could also be a borrowing from Proto-Western-Romance *jejt-, which survived unattested in one obscure English dialect for several centuries before becoming famous in the era of Vine. But it almost definitely isn't.
In other words, quite a lot of things are technically possible. It's very hard to prove a negative. It's technically possible that Japanese and Arapaho are Indo-European languages, and it would be extremely hard to show 100% conclusive proof that they aren't—even though that theory is ridiculous and there's no evidence to support it. That's just the nature of science.
This is why Occam's Razor is important: the simpler explanation is often the better one. It's somewhere between difficult and impossible to prove that Proto-Germanic *xabjaną existed before all contact with the Romans, because all our attestations of Germanic are post-Roman-contact. But we can say with good confidence that the PIE root *k-h₂p- existed, and that if this root's *-ye- present had survived into Germanic and gone through the usual sound changes, it would have come out looking exactly like *xabjaną. This is a simpler explanation that doesn't require us to assume any loanwords, so Occam's Razor says it's probably the better one.
