> Even so, what's the point of all this money splitting?
It eliminates the most proximal cause to the SVB collapse: strongly correlated behavior by large account holders. If banks have many account holders that are diverse, it's hard enough to get a panic that causes illiquidity.
The duration risk would have still been an issue, but a far smaller issue without the need for panicked liquidation. (Even with a healthier balance sheet, most banks would struggle to endure a bank run like SVB).
> Let account holders get on with more useful work.
If you have a massive amount of money, there's a little bit more work to do than dumping it in a random bank.
Ages ago, when my firm had a $15M bank balance, it was close to trivial to work with a vendor to spread the balance among 20 banks. We were still not fully insured, but we had significant insurance and limited exposure to any individual bank failing. 20x the risk of failure, but 1/30th the cost should one happen. We could have also bought some treasuries directly.
> Apply those increased premiums retroactively to all the banks that failed recently, if feasible.
How do you charge a bank that failed and doesn't have the money to pay its depositors money?
I'm in favor of a small increase in the insurance coverage. Maybe $500k or $750k. But I don't think it's good to make bank accounts equivalent in safety to treasuries up to any amount but more liquid.