What's Going on with First Republic Bank?

Following the failure of Silicon Valley Bank, a lot of other regionals have experienced depositor runs and serious pressure on their stock prices. But there's actually a lot of variation among regionals, and the solutions to SVB's problems don't necessarily fit the other regionals' problems, as the case of First Republic Bank shows.

SVB's problem appears to have been straightfoward:  it had short-term, flighty liabilities and long-term assets. Specifically, its liabilities were mainly deposits, and 86% of its deposits were uninsured, making them particularly flighty. When these deposits started flowing out, SVB needed liquidity to honor the withdrawals, and that meant liquidating or borrowing against its assets. Its assets consisted of primarily of securities (56% of assets) and loans (35% of assets). SVB had already in 2022 borrowed heavily from the Federal Home Loan Bank System against the loans (mainly residential mortgages)—$15 billion of advances at the end of 2022, so it's not clear how much more liquidity it could have gotten from its mortgages: it already held 11% of the capital stock of the Federal Home Loan Bank of SF (capital stock must be held equal to 3% of advances), more than any other bank at the end of 2022. Therefore SVB had to turn to its securities. Those securities were almost all being held to maturity (34% of total assets), meaning that they were not marked-to-market, so SVB had to realize the loss upon sale (or borrow with them at market-based valuations).

This is a profile that isn't matched by any other regional. Signature Bank, which also failed, had an even higher percentage of uninsured deposits (89%), but Signature's assets were overwhelming loans (66% of assets), and its HTM securities were limited (only 7% of assets). But after SVB's failure, the fast money was spooked, and Signature didn't have any way to come up with the necessary liquidity in time.

What about First Republic, which has been the bank under the most pressure since the failures of SVB and Signature?

First Republic has a lower (but still somewhat high) percentage of uninsured deposits (67%), but that's still high enough to raise real concerns about depositor flight. And on the asset side, First Republic's balance sheet is mainly loans (78% of assets), with securities being just 15% of assets. Of the securities, however, almost all are held-to-maturity. This suggests that First Republic might have more FHLB borrowing capacity than SVB. On the one hand, its FHLB advances were only $9 billion at the end of 2022. On the other hand, it already held 10% of the FHLB of SF's capital stock at the end of 2022, a figure exceeded only by SVB (11%). FHLBs require their members to hold capital stock in the banks in proportion to their borrowing, so this would suggest that First Republic was already a very heavy user of FHLB advances. One hopes that as a matter of basic risk management the FHLBSF would not let any bank go much over 10% of its capital stock. Unfortunately, we cannot really tell what First Republic's remaining FHLB borrowing capacity is because there are no public lending limits available for the FHLBs--it's very discretionary.

Even if the FHLB discount window is tapped out, there's always the traditional Fed discount window, which makes loans against Treasuries and agency securities, valued at market. That's not going to help First Republic, however, because First Republic's securities holdings aren't Treasuries or agencies. They're long-term munis! In other words, First Republic was really reaching for yield. 

Now the Federal Reserve Board announced a new lending program, called the Bank Term Funding Program. The program is basically discount window lending against collateral. The same type of collateral is eligible as with regular discount window lending:  Treasuries and agency securities. There's just one substantial twist: collateral in the BTFP will be valued at face value, rather than at market value. It's all being done at very low rates:  one-year overnight index swaps rate plus 10bps. There's no Bagehotesque punitive rate being charged. In other words, the BTFP is a bailout of banks that did a bad job managing their interest rate risk by hoarding up on long term Treasuries that they held to maturity.  It's a facility that would have been very helpful to SVB, but it's basically worthless for First Republic Bank, given the nature of First Republic's securities portfolio.

First Republic seems to have found some financing from private sources, but the point I want to close this post with is that while there's a lot of attention to the extension of deposit insurance to uninsured deposits, the Bank Term Funding Program bears some consideration. No one in the private market would lend against securities at face, rather than at market. But that's what the Fed's doing in order to enable banks that have held-to-maturity securities avoid loss realization. The Bank Term Funding Program is a lifeline for banks that failed at banking 101—managing interest rate risk. The whole nature of banking is that it involves balancing long-term assets and short-term liabilities. Firms that can't do that well probably shouldn't be in the banking business.

Comments

[SivaDancer]

Thanks, Adam. Always appreciate the good points you make.

One thought about the BTFP, however. You are clear that "no one in the private market would lend against securities at face, rather than at market."

But BTFP is only accepting Treasuries and agency securities as collateral; thus, the lender here is also the issuer of these securities, buying back their own debt.

Doesn't that reduce the moral hazard somewhat?

[Hudaf]

As was clearly identified by the FDIC in its quarterly updates, US banks unrealized mark to market (MTM) losses zoomed from negligible levels from negligible levels in Jan 22 to $ 620 billion in Dec 22. These are well over 25% of regulatory capital of $ 2.2 trillion - in many cases exceeding 80% of their capital making them extremely vulnerable and risk. Bank of America unbooked MTM losses in 31-Dec are likely to be over $100 billion - over 40% of it's capital.
The real moral hazard is not taking any action those who recklessly wreaked the US Banking system by inflicting huge losses on them - as not doing so, will allow others to indulge in similar reckless actions.