To The Moon

Several banks have failed, collapsed, or been bought in the last few weeks. Do not make the mistake of thinking this is just like 2008.

Banks need to maintain solvency: assets have to be greater than liabilities. This has been the issue in both crises, hence people drawing many parallels.

A big difference is that banks in 2023 are not a house of cards. Last time the banks started to fail, it was largely because of exposure to highly overvalued mortgage derivatives that were never worth what they were sold for and contained excessive risk. Banks like Lehman Brothers had high exposure to these “toxic” assets, so they failed.

In 2008, markets woke up and realized that many assets owned by the banks are worth a lot less than previously thought. This is partially due to widespread mortgage fraud, overcomplexity of exotic derivatives like synthetic CDOs, and improper risk assessment and credit ratings. Mortgage instruments were held as low-risk capital — commonly held because of their attractive yields to add to a banks’ assets.

These assets’ valuations plummeted (assets thus go down), so suddenly banks ran into their worst nighmare: liabilities > assets. Banks that were playing things risky, holding less capital relative to liabilities, thus suffered. This created liquidity crunches, attempts at huge borrowing, and a whole cascade of effects. The point is that banks were holding coal instead of gold.

2023 is completely different. There is no “coal”, even if it may seem like it.

There are two main issues: interest rate risk and insured deposit risk.

I will use Silicon Valley Bank (SVB) to illustrate the issue.

Interest rate risk

The amount of deposits at SVB skyrocketed during 2021. Deposits are liabilities to the bank because they are what the bank owes: they need to pay clients interest and also owe their clients their deposit amount.

Banks balance this by using those deposits to generate assets. They can lend out money received through deposits, then collect interest on their loanbook. Borrowers owe the bank the loan back plus interest, becoming the bank’s assets.

Where SVB got into trouble is the other thing that banks do with deposit money. They can’t loan out all deposits, because people still need to withdraw funds. So banks keep some cash, and they hold “capital” in the form of assets of varying liquidity that can be sold as needed.

So SVB got a lot of new deposits, even faster than they could lend, and they decided to buy up some interest-bearing assets. They bought long-dated mortgage backed securities (these were safe, not toxic, so don’t worry) and T-bills. This was a pretty big mistake.

The Fed was increasing interest rates in 2021. And they continued to raise interest rates throughout 2022. When rates go up, existing bonds resell at low prices because their yields are suddenly unattractive relative to bonds that can be bought at the newly high rates.

But SVB had just bought a huge amount of bonds that were yielding really low interest rates. As long as SVB didn’t sell these bonds, they would be fine even if they just collect their low interest rate. If SVB had to sell the bonds, they would sell them at a big discount, because investors can now get bonds with way more attractive interest rates.

So SVB had a big issue. Interest rates moved against their favor, and they moved historically quickly. This made SVB vulnerable, but it wasn’t the nail in the coffin. As long as they kept their deposits and didn’t have to sell, they would be fine.

Insured Deposit Risk

The problem is that people withdrew, and they withdrew quickly.

A lot of the accounts at SVB were held by startups and venture capital funds. The FDIC only insures up to $250,000 of an entity’s deposits at a bank. These tech players’ accounts were flush with cash — a lot more than just $250k. So if SVB were to theoretically collapse, people would be wiped out and lose most of their account value.

There was an absurd amount of deposits past the FDIC insured limit at SVB.

People began to realize the issue. SVB had an issue, it had assets that could easily be marked down, and it had a lot of deposits that people had incentives to withdraw.

SVB experienced a run on its bank. People saw it was at risk, and they didn’t want to lose their uninsured deposits. In a matter of hours, clients withdrew billions.

This was because SVB had acknowledged its interest rate risk and sold billions worth of bonds at a discount, stepping into the trap it had accidentally set up for itself. When depositors saw SVB lose asset value, they realized there could be more to come, and many decided to flee before others could.

This was different than 2008

This was a more classic run on the bank, in the era and style of modern banking. It is not the tale of widespread unseen risk or artificially propped up assets. It is the repercussions of idiosyncratic trends in interest rates, accounting rules, and in the case of SVB, you could say suffering from success.

The way I choose to look at it is that the Great Financial Crisis of 2008 was very inorganic and artificial. It was smoke and mirrors, a house of cards.

This time around, the whole landscape seems more natural. Banks played by the rules but some couldn’t weather the storm.

The bailout situation is also notably different. There is no Congress-led liquidity injection into the banks like TARP (2008) was.

The FDIC has stepped in to scrap SVB, but the FDIC isn’t taxpayer funded. Credit Suisse was bought by UBS, which was a deal aided by the Swiss government but still a market acquisition.

There has been no government action to prop up a bank beyond using regulatory strength or tools of the financial system. I think of bailouts as the government bailing out banks by buying a stake or giving them the cash they need to survive. That isn’t happening.

A lot could change, and more could come to light in the next few months. Perhaps we will see more classic bailouts. But right now, this crisis is its own beast.