That works in internet logic but it's not true. At least, it's not what I've been told inside YC and I don't believe anyone was lying. What it's actually related to is Garry taking over YC and deciding to refocus YC on its original roots. (I specifically asked about that today during an all-hands btw.)
At a minimum, it takes longer than 3 days to make a big change like this. I realize that's not a persuasive argument though, since there is endless supply of explanations one can come up with (what if they knew about SVB beforehand!)
With all respect, those interpretations differ only in spin. "Let's refocus on our roots" and "Oh shit, the gravy train dried up so we can't run our late stage cyclic money printer anymore; at least we still dominate the seed landscape" describe the same actions.
And as for assuming good faith that this was a fait accompli already in progress... I'm sorry, but I can't get past "We're doing this on the first business day following the biggest upset to our industry in a decade and a half. But that's just a coincidence!"
I see both of your points and I get how compelling the second one is, but what can I tell you? It was definitely already in progress and it has to do with long-term leadership changes inside the organization and thinking about what YC is really supposed to be. Keep in mind that Garry's previous years at YC were before the late-stage business got started. Why they made the announcement yesterday, I have no idea—perhaps someone enjoys Hacker News drama!
FWIW I don't think the difference is spin, I think it's the opposite—from my perspective (I'm not privy to the inside discussions) this change is not a reaction to short-term business climate, it's a vision thing. Most people underestimate the power of having just one thing to focus on and knowing exactly what it is. It's a cliché everyone pays lip service to but almost no one actually does. It was a huge insight for me that running HN is all about focusing on one thing*—it's a kind of superpower that turns a surprising number of hard calls into easy calls.
You can't know about a bank run beforehand. It dies when it dies, it's kind of a surprise that way.
(Being insolvent isn't actually enough to kill a bank as long as the run doesn't start. Of course, here the bank dies when the regulator decides it dies anyway.)
At a minimum, it takes longer than 3 days to make a big change like this. I realize that's not a persuasive argument though, since there is endless supply of explanations one can come up with (what if they knew about SVB beforehand!)