3 ms·
I'm so glad that this was posted. I, too, have recently been thinking about how the FDIC works given the lockout of funds at Celsius. This is an incredible read
by Shank 4y ago
I'm so glad that this was posted. I, too, have recently been thinking about how the FDIC works given the lockout of funds at Celsius. This is an incredible read.
I've also dived into this more, however. It turns out the FDIC maintains a list of bank failures and summaries [1][0], respectively. These are awesome to read. There have been zero failures in 2022 so far, zero in 2021, and four in 2020.
Most failures in the last few years have resulted in assets being sold to acquiring banks. The most recent failure where the FDIC mailed checks to people (this is what I imagine most people think of when they see the FDIC insurance policy) was First NBC Bank in Louisiana in 2017 [3]. Some of the assets were acquired by Whitney Bank, but a large number of assets were not. The bank failed on April 28th of that year, and the FDIC mailed checks on May 1st [2], which is pretty dang good, in my opinion.
I think what I appreciate the most about this article and the recent failures is just how much attention FDIC devotes to keeping services available for customers, even when the situation is changing. Scanning over a few of the bank failure summaries, they all have a similar pattern: bank closes, all ATMs stay online, checks clear as soon as reasonably possible, and branches remain open. In the First NBC Bank failure, the worst "downtime" was experienced by people who had accounts backing other more complex financial products that didn't have immediacy attached (certificates of deposit, which are predicated on being held for a long time, for example). Everything else was "zero downtime", which is really stellar.
Seeing other bank failures on the list, particularly in 2017, that I never heard about, also reminds me of the fact that this is stuff that matters more frequently than I would have expected (i.e., outside of financial crises, like 2007/2008).
[0]: https://www.fdic.gov/bank/historical/bank/ https://www.fdic.gov/bank/historical/bank/
[1]: https://www.fdic.gov/resources/resolutions/bank-failures/failed-bank-list/ https://www.fdic.gov/resources/resolutions/bank-failures/fai...
[2]: https://www.fdic.gov/resources/resolutions/bank-failures/failed-bank-list/firstnbc.html https://www.fdic.gov/resources/resolutions/bank-failures/fai...
[3]: https://www.fdic.gov/bank/historical/bank/bfb2017.html https://www.fdic.gov/bank/historical/bank/bfb2017.html
- DSingularity 4y agoI wonder where cryptocurrency will end up. Eventually people will want entities like FDIC to step in and soften the blow. If Celsius auctioned its assets would we be where we are?
- nradov 4y agoI certainly don't want the federal government to get involved in providing insurance to cryptocurrency "investors". It would be a total waste of resources. Let them take some hard blows. If they lose all of their money that's a totally acceptable outcome.
- kerblang 4y agoTrue but I think the crypto companies should be able to buy the same insurance as gambling casinos, since... they are gambling casinos
- jaywalk 4y agoCrypto companies are more like the gamblers than the casinos. And nobody is going to give a gambler insurance against losses at a casino.
- dmurray 4y ago"Insurance" in blackjack is a reasonably popular bet, and gives the house more edge than the regular game. Some people hedge their cryptocurrency portfolios with put options, which is a form of insurance. You can package almost any financial instrument and sell it to someone. Perhaps you mean the state shouldn't subsidise gamblers with free insurance - I think most people would agree with that.
- jaywalk 4y agoInsurance in Blackjack is a losing bet for the player, and nobody with even beginner-level experience takes it. I'm not sure how anybody can insure against something like Celsius. The "investments" just went poof.
- jsmith45 4y agoThe FDIC operates on a few constraints. One is that is has two roles, both as insurer, and as receiver. The second is that the law requires it to deal with failed banks in the manner that incurs the least costs on its fund. It should be quite unsurprising that if an offer to buy 100% of the banks assets and liabilities comes in, this is likely to get accepted, since it likely to have the lowest possible costs to the FDIC itself. No costs in insurance reimbursing, obviously, but also fairly minimal receivership costs, as they don't have left over assets to sell, don't have to deal with creditors for any debts that transfer with the sale, leave a relatively small set of creditors (including the shareholders) to deal with. Whole bank purchase offers are not exactly rare, but offers to purchase only some of the assets while taking on some of the liabilities (often including accounts either in full, or up to the insured limit) are probably even more common. This is obviously more expensive to the FDIC, since they may end up with more unacquired liability (i.e. more creditors to deal with), assets to sell, etc. The FDIC does move fast, like disturbingly so, but their requirement to minimize their own expenses basically requires it. They generally already have the acquisition offers in hand when they first show up, and for their expected plan of action, have already talked with the relevant state regulators to have them ready to issue any approvals needed. Even things like issuing checks if no-one acquires the accounts, taking any more time than necessary can accrue additional costs. For things that don't actually cost the FDIC itself any money, like negotiating terms of acquisition, they can afford to be extremely consumer friendly. Requirements on the acquirer for things like keeping all branches open for at least x months, no sweat off the FDIC's back.