節選自Bloomberg的文章:Banking Gets a Bit Narrower|Also Bed Bath stock, SVB’s bonds and the Chainsmokers.
作者:Matt Levine
翻譯:Notion AI

Emergent narrow banking|緊急狹義銀行業務

We have talked a few times recently about, like, the theory of banking. That theory goes roughly like this. Banks fund themselves with deposits, which are basically short term and safe: If you have $100 in a bank account, you expect that it will always be worth $100, and you expect to be able to withdraw five $20 bills any time you want. Meanwhile banks invest their money in assets (loans, bonds) which are basically long term and risky: Banks make loans to risky businesses that don’t have to be repaid for years.


This business has well-known problems. The main problems are:

  • The deposits are short-term but the assets are long-term: If all the depositors want their money back at once, it isn’t there, because it has been loaned out. This can lead to bank runs, panics, fire sales of assets, It’s a Wonderful Life, etc.
  • The deposits are safe but the assets are risky: If the bank makes bad loans and loses money, there won’t be enough money to pay back depositors, and bank deposits are meant to be money, so depositors really count on their bank deposits being safe.


  • 存款是短期的,但資產是長期的:如果所有存款人同時要求取回他們的錢,因為錢已經被借貸出去了,所以沒有足夠的錢,這可能導致銀行擠兌、恐慌、資產抛售,就像《生活真美好》中的情節一樣。
  • 存款是安全的,但資產是有風險的:如果銀行放出不良貸款並虧損,將沒有足夠的錢來償還存款人,而銀行存款本該是貨幣,所以存款人確實依賴他們的存款是安全的。

This business — “fractional reserve banking,” — is inherently risky and fragile. Everyone knows this, and there are standard methods to mitigate the risks. Banks have capital requirements (they are partly funded with equity, not deposits, so if the assets lose value the depositors still get their money back). They have liquidity requirements (some of their assets are short term and safe, so if depositors want money back there’s some money to give them). There is safety-and-soundness regulation and supervision (the government tries to prevent banks from making loans that will lose money). There is the lender of last resort (the central bank lends money to solvent banks that need cash, so if depositors want their money back the banks can get it). There is deposit insurance (the government promises that depositors will get their money back, making bank accounts safer and runs less likely). Still there is some unavoidable residue of fragility: The mismatch between the banks’ safe short term liabilities and their risky long term assets creates risk, and somebody — if not depositors then the government — has to bear that risk.


You might draw one of two conclusions from this:

  1. Oh well! Banking is good; there is social value in this fragility; it allows society to pool its safe capital to be able to take on risks. “Banking is a way for people collectively to make long-term, risky bets without noticing them, a way to pool risks so that everyone is safer and better-off,” I wrote last month. “Financial systems help us overcome a collective action problem,” Steve Randy Waldman wrote in 2011: “In a world of investment projects whose costs and risks are perfectly transparent, most individuals would be frightened. … A banking system is a superposition of fraud and genius that interposes itself between investors and entrepreneurs” to encourage investment and make everyone better off.
  2. No, this is bad, it’s antiquated, we should fix it. Let’s get rid of fractional reserve banking and do something else.


  1. 喔,好吧!銀行業是好的;這種脆弱性具有社會價值;它允許社會集中安全資本來承擔風險。我上個月寫道:“銀行業是人們集體進行長期、風險投注的一種方式,以一種不被注意到的方式存在,是一種將風險匯集起來,使每個人都更安全和更富有的方式。”“銀行系統幫助我們克服集體行動的問題。” Steve Randy Waldman在2011年寫道:“在一個成本和風險都是完全透明的投資世界裡,大多數個人都會感到害怕……銀行系統是一種欺詐和天才的疊加,它介入了投資者和企業家之間,鼓勵投資,使每個人都更富有。”
  2. 不,這是不好的,它已經過時了,我們應該進行修復。讓我們擺脫儲備金銀行業,并做些其他的事情。

Conceptually the main way to do that goes something like this:

  • Banks should take short-term safe deposits and invest them in short-term safe assets.
  • Long-term risky assets should be funded with long-term risky liabilities.


  • 銀行應該接受短期安全存款,並將它們投資在短期安全資產上。
  • 長期風險資產應該通過長期風險負債來獲得資金。

This is loosely speaking called the “Chicago plan,” or “narrow banking.” Banks would just hold cash for their depositors, so the cash would always be there and banks would not have any risk of runs or credit losses. In modern banking, the banks would not hold “cash” in the sense of $20 bills in a vault; rather, they’d take deposits and turn around and deposit the money as reserves at the Federal Reserve. The Fed pays interest on reserves, so the banks could earn enough money to pay for salaries and branches and still pay some interest to their customers. (Though it’s not clear why a narrow bank would need branches; it would be a lot cheaper to operate a narrow bank than a full bank, so presumably it would need less of an interest margin.)


Meanwhile, risky loans would come from people who intend to make risky loans: You could have, say, a loan fund that raises money from investors, locks up their money for 10 years, uses the money to make 10-year loans, and then pays back the investors whatever the fund gets back on the loans. 4 If the fund makes bad loans, the investors lose money; they bear the risk knowingly and directly, unlike bank depositors who sort of don’t know where their money is going.


Perhaps this system would make borrowing more expensive, as the current system of “fraud and genius” disguises risks in the financial system and so makes it more willing to take risks cheaply. But perhaps it would be less fragile; it would cost society a little more most years but a lot less some years.


People sometimes talk about imposing this sort of system by legislation or regulation; this had a wave of popularity after 2008, though that has waned in recent years. I have written before that the Fed seems to have no interest in narrow banking — it kind of hates it and has declined to approve a bank that tried to do exactly this model — and I don’t really expect fractional reserve banking to end anytime soon.


But I do want to point out that there are indications that the markets are moving toward narrow banking all on their own. We talked last month about money market funds and the Fed’s reverse repo program. You can put your money in a money market fund that will just park it directly at the Fed, though when a money market fund parks cash at the Fed that is called “reverse repo” rather than “reserves.” A money market fund that just parks its money at the Fed has essentially no risk: All of its assets are short-term (they are parked overnight) and safe (its debtor is the Fed), so your money is safer there than it is at a bank. It has few expenses: It doesn’t need a branch network or a lot of lending officers to take in money online and park it at the Fed, so it can pass on most of the interest that the Fed pays it directly to its customers. And the Fed does pay it a lot of interest — reverse repos pay something like 4.8% — so it can end up paying depositors more interest for taking less risk than a bank can.


n 2021, when interest rates were zero-ish everywhere and when there hadn’t been a big bank failure in a while, nobody had all that much incentive to move their money from a bank account paying 0% interest to a money market fund paying barely more. But in 2023, when people were worried about their banks’ safety anyway, moving from a bank account paying 0% interest to a safer narrow money market fund paying 4.8% seems pretty attractive. And so people do. The Wall Street Journal reports:


Main Street banks such as Citizens Financial Group Inc. and First Horizon Corp. said in recent first-quarter earnings reports they are having a tougher time hanging onto customer money in a world where the Federal Reserve has aggressively raised interest rates. To keep those depositors around, some lenders are paying more on savings accounts and turning to products like certificates of deposit.

Though profits rose at many banks in the first quarter, the deposit declines signal a fundamental change in their business. Deposits were plentiful in the era of superlow rates because customers had little incentive to move their money elsewhere. Banks grew to rely on them as a cheap source of funding that they could use to make loans or buy bonds and other securities. …

In recent months, with the Fed continuing to raise rates, consumers and businesses at banks across the industry started moving their money from bank deposits to higher-yield offerings such as money-market funds and Treasurys. Customers who used to be satisfied keeping their money in accounts that paid no interest decided they no longer were.

The collapse last month of Silicon Valley Bank and Signature Bank accelerated the shift at some firms. Both banks were felled by panicky customers pulling out their money after concerns started spreading about the banks’ health. Customers at those banks and some other smaller firms across the U.S. started moving their money to the largest banks, betting they were safer even in crisis. …

Even the biggest banks, which enjoyed a windfall of deposits from smaller-bank customers, had to pay more for deposits.

JPMorgan Chase & Co. said last week it picked up about $50 billion in new deposits following March’s bank failures, but overall deposits were up only $37 billion from the end of 2022, indicating they would have dropped if not for the industry turmoil. JPMorgan executives cautioned that the new deposits might not stick.

像Citizens Financial Group 和 First Horizon Corp這樣的城鎮銀行在最近的第一季度財報中表示,在美聯儲積極上調利率的世界中,他們更難留住客戶的資金。為了留住這些存款人,一些貸款人正在為儲蓄賬戶支付更多的費用,並轉向像存單這樣的產品。



上個月硅谷銀行和Signature Bank的倒閉加速了一些公司的轉變。在擔心銀行健康狀況開始擴散後,兩家銀行都被驚慌失措的客戶抽出了他們的資金。這些銀行和美國其他一些較小的公司的客戶開始將他們的資金轉移到最大的銀行,打賭即使在危機中他們也是更安全的。



Then there is the other side of the Chicago plan: If banks don’t do lending, lending needs to come from funds whose investors lock up their money and consciously take on the risk of making loans. Those funds might have to charge more for loans than banks do, because their investors, unlike bank depositors, are knowingly locking up their money and taking risks and so want to be compensated.


And of course there is a lot of that in modern finance, loans that come from funds rather than from banks. Leveraged buyouts, for instance, used to be financed with a combination of junk bonds (bought by funds, insurers, etc.) and bank loans (bought by banks). In recent years, the “bank loans” have increasingly been owned by hedge funds, collateralized loan obligations, etc.; banks syndicate them but increasingly they end up on someone else’s balance sheet. And more recently a lot of leveraged buyouts have increasingly been financed by private credit, deals with loan funds (often sponsored by private equity firms) that cut out banks entirely.


It is not just leveraged buyouts, though. Apollo Global Management Inc. bought a lot of Credit Suisse Group AG’s structured finance and weird lending businesses, before Credit Suisse disappeared, forming a new lending business funded by investors and its insurance company. Last month it gave PacWest Bancorp a $1.4 billion lending facility, and I wrote: “It used to be that banks funded themselves with deposits and used the money to make loans to private equity firms. Now banks fund themselves with loans from private equity firms.”

不僅僅是槓桿收購。在Credit Suisse消失之前,Apollo Global Management Inc. 收購了Credit Suisse Group AG 的大量結構化融資和奇怪的貸款業務,組建了一個由投資者和保險公司資助的新貸款業務。上個月,它向PacWest Bancorp提供了14億美元的貸款便利,我寫道:「曾經,銀行用存款為自己融資,並用這些資金向私募股權公司發放貸款。現在,銀行通過從私募股權公司獲得貸款來為自己融資。」

Or here is a Bloomberg News story about Blackstone Inc.’s earnings:

[Blackstone President Jon] Gray said that the tumult following the collapse of three US regional lenders last month has created investment opportunities — even after Blackstone backed a firm’s losing bid for Silicon Valley Bank. Blackstone has been talking to small banks about stepping in to lend alongside them as more look to slim down their balance sheets

“Because of the focus on liquidity they may want to find partners,” Gray said.

這是彭博新聞報道的有關黑石集團(Blackstone Inc.)財報的信息:

[黑石集團總裁Jon] Gray 表示,上個月三家美國地區銀行破產後的動蕩局面為投資提供了機會(黑石支持的一家公司對倒閉的硅谷銀行進行出價未成功)。但是,黑石一直在與小銀行交流,希望與他們一起貸款,因為越來越多的銀行希望縮減其資產負債表。

「由於關注流動性,他們可能希望尋找合作夥伴,」Gray 說。

Small banks don’t want to make loans “because of the focus on liquidity”: Their deposits are more risky, so their banking needs to be narrower. So they turn to funds run by Blackstone, which raises long-term money from investors who want to take credit risk, to make their loans.


Basically in 2023 the market has created a lot of money-market funds that look like the deposit-taking side of narrow banking, and a lot of credit funds that look like the lending side of narrow banking. I don’t want to overstate this, and there are still a lot of big universal banks that are attracting deposits and making lots of loans. But they have some competition from narrower banking.


One very speculative thing that I might say here is that the magic — the opacity, the “superposition of fraud and genius” — of traditional fractional reserve banking is maybe a bit harder to do in a world that pays attention to it. It is relatively easy to find a bank’s balance sheet online, and if one person reads that balance sheet and notices that the bank is insolvent, it is relatively easy for her to tweet that and have it go viral and cause a bank run. 5 “Game’s the same, just got more fierce,” the vice chairman of the Federal Deposit Insurance Corp. said last week, about bank runs in the age of social media. If banking relies on a certain opacity, modern communications technology and social media might just make that opacity harder to achieve. 6 If it’s too easy to see what the magician is doing, the magic doesn’t work anymore.



Published by