[shawabawa3]

> It's said that no bank (even the best-managed) can withstand a fullscale bank run

I keep reading this but this should not be true

Any bank will hit liquidity issues on a full-on bank run, as not 100% of a banks assets will be marketable, but central banks will provide emergency liquidity in these situations

But banks should not hit insolvency issues like SVB did

[HWR_14]

The line is a little blurry. You can be underwater in mark-to-market terms on a bond portfolio, but be fine if you guess the timeline for redemption and can hold them to maturity. Banks estimate what percentage of their assets will be held to maturity. So needing extra liquidity causes insolvency.

[fsckboy]

> hold them to maturity

the PV of a bond is the risk adjusted, discounted (at current rates) FV of the bond (and interim coupons). The losses are real, they don't magically come back, they just magically appear to realign like your ETA does as you get close to your destination after losing hours in a traffic jam.

[HWR_14]

I mean, the bank lost money. However if there is not a bank run they make enough off the interest of their deposits to be solvent. It's not unlike an airline owing more on their fleet than it is worth. As long as they keep doing airline things, they'll service their debt and be fine. If they don't, then they'll go bankrupt.

[testrun]

This is wrong. Losses are only realised at redemption or sale.

[fsckboy]

the rules of accounting say losses are only recognized when realized (mostly because they don't want you playing the game in the other direction and exaggerating your worth), but the cashflows and market prices tell the true story, a different story than book value.

You don't buy bonds (or mortgage backed securities), you buy the future cashflows they represent. Let's say you have a billion dollars, and you buy future cashflows worth a billion at a 2% interest rate, then interest rates jump to 6%: you still own the same future cashflows, but those cashflows are not worth a billion any more, they're only worth a third of that (give or take). now you don't have the flexibility to sell those cashflows to finance a billion in other investment activities. Now your stock is not worth what it was, so if you sell stock you get less to finance investment opportunities. Will your bank be OK if you hold what you hold to maturity? you will get the cashflows, which if you have perfectly balanced cash flows-in with cash flows-out means you break even, but (a) you didn't do that and (b) that's not the business you are in, to sit in stasis for 20 years: you are much less valuable, your opportunities have dried up, and current customers do not want to do business with you any more. Sharks have to keep swimming or they suffocate (which may not be true for sharks, but it is true for banks) A bank is not a treasure chest waiting to be dug up, it's a McDonalds that needs people to order fries with that.

Roku put something like $500 million into the bank, that money gets invested by the bank overnight into very safe govt paper, but during the day it is turned back into cash assets of the bank, and it disappeared going into Roku's account and without being turned into govt paper again. It is not the case that Roku will get its $500 million back by sitting on that paper, there is no paper and the cash is gone.

[vatsalaggarwal]

I think what made it worse here is that depositors of SVB were undiversified... If you had a bank that held deposits from various kinds of people, it's less likely that all of them would want their money out at the same time unless there was some of kind of widespread change in sentiment towards the overall financial system.

[helsinkiandrew]

What made SVB even worse was that its customers were flush with VC cash and didn't need to borrow, so it looked elsewhere to make a return on the cash.

[dragonwriter]

> Any bank will hit liquidity issues on a full-on bank run, as not 100% of a banks assets will be marketable, but central banks will provide emergency liquidity in these situations

That was not the case in the US for banks with HTM assets until the backstop program announced by the Fed in the wake of the SVB collapse.

> But banks should not hit insolvency issues like SVB did

SVB’s liquidity issues turned into solvency issues because of the absence of a liquidity backstop.

[evancox100]

Treasuries are some of the most liquid instruments available, it is not a “liquidity” issue like “How do I line up buyers for all these weird, hard-to-price assets”.

And the measure put in place by the Fed is not a liquidity back stop, it is a value/solvency bailout, or kind of capital infusion. This is what SVB was trying to do on Wednesday, raise capital. That should tell you it is not a liquidity problem.

[s1artibartfast]

I think you may be confusing FDIC actions.

One of the actions was a program to loan against the full value of long term assets at term price instead of current market value.

Maybe we are just quibbling over the definition of liquidity. long term treasuries and MBS can be easily sold, but you may take a substantial loss in doing so instead of holding to maturity.

[evancox100]

Not confusing the two, and maybe it is quibbling but changes in market value have nothing to do with liquidity as that term is used in finance.

Just because you can't sell a bond at par doesn't mean it is an illiquid market.

[kyrra]

There's a certain amount of risk with all lending and investment that banks do with deposit funds. This can be home loans, government bonds, and other relatively safe products.

I don't think it should be up to the government to back these risks, because if banks think the government will always rescue them, they don't need to care as much about risky investments.

You could argue that it is depositor money, so they're not really saving the bank, they're saving customers. But if banks don't have to care about their risk profile, customers will deposit their money in whatever bank is offering the greatest interest rates, which will likely be those that are making the riskiest investments, which could lead to more bank failures with market swings.

[Ntrails]

> I don't think it should be up to the government to back these risks, because if banks think the government will always rescue them, they don't need to care as much about risky investments.

Counterpoint - if depositors had known that the bank could not lose their money, because the government will back it, there would be no run on the bank. Why bother? It's _safe_ by design. I do agree with the general point, and there are huge questions around capitalisation and marking with long term debt etc.

What's most interesting here to me is the inaction within SVB when they could have been fixing these problems for survivable losses early, but instead tried to ride the storm.

I'd love to read some of those meeting minutes...

[mywittyname]

This isn't really feasible. Bank executives will fill fight the regulations that stabilize their bank against market swings, and citizens will demand that the government do something when irresponsible bank executives lose their life savings. Banking regulations, like most regulations, are written in blood, but for some reason, the banking sector is the most adept at getting regulations overturned.

So the resulting system is a patchwork of solutions that force the government into the role of rescuer. It's just too difficult to get most regulations to stick long enough to prevent another banking crisis. Barring a Constitutional amendment to create a banking "tsar" with broad authority and who reports to no-one (i.e., non-political), the best solution we have is to have the government step in when banks inevitably fail.