I.H.142.Right Business.Matt Levine.Banks

银行对储户有不可调和的矛盾，既是客户又是“原材料”。

1、《2024-11-18 Matt Levine.SRTs》

SRTs

Another theme of this column is that banks don’t have any money. Oh, not literally; I’m partly kidding about this one too. But I have argued that, between stricter bank capital regulation and the rise of giant alternative asset managers with huge pots of money to put to work, banks are just less likely to make and hold loans than they used to be. In the olden days, a fairly standard model was “you go to a bank for a loan, and the bank gives you the loan, and if you don’t pay it back then the bank loses money.”

本专栏的另一个主题是银行没有钱。哦，不是字面意义上的；我也有点开玩笑。但我认为，由于更严格的银行资本监管以及拥有巨额资金可供运作的巨型另类资产管理公司的崛起，银行如今不太可能像过去那样发放和持有贷款。过去，一个相当标准的模式是“你去银行贷款，银行给你贷款，如果你不还钱，银行就亏钱。”

In modern banking, it is becoming more common for the model to be “you go to a bank for a loan, and the bank goes to an investor to get the money,3 and the bank lends you the money, and if you don’t pay it back then the investor loses the money.” The bank, in this model, is increasingly a sales network; it is in the business of having a lot of bankers who cover a lot of companies and can originate loans, but its balance sheet is precious, and it will try to find someone else to actually put up the money.

在现代银行业中，“你向银行贷款，银行向投资者筹集资金，然后银行把钱借给你，如果你不还钱，投资者就会亏钱”这种模式变得越来越普遍。在这种模式下，银行越来越像一个销售网络；它的业务是拥有大量银行家，他们覆盖许多公司并能发起贷款，但其资产负债表十分宝贵，因此会尽量寻找其他人来实际提供资金。

银行和储户的目标本来就不一致，跟投资者的目标也不一致。

In some ways this model is obviously safer than the traditional model: Banks are largely funded with short-term leverage from depositors, so they are at risk of bank runs. Private credit firms and other new alternative lenders are largely funded with long-term locked-up equity capital from big investors (insurance companies, endowments) with long time horizons, so they can more safely bear the credit risk of the banks’ loans. It is a move toward narrower banking.

在某些方面，这种模式显然比传统模式更安全：银行主要依赖存款人的短期杠杆融资，因此面临银行挤兑的风险。私人信贷公司和其他新型替代贷款机构主要依赖大投资者（保险公司、捐赠基金）提供的长期封闭股本资金，这些投资者的投资期限较长，因此可以更安全地承担银行贷款的信用风险。这是一种向窄银行模式发展的趋势。

On the other hand, there is an obvious problem with this model, one that you might remember from 2008. It is moral hazard. If a bank (1) makes loans and then (2) loses money if they don’t get paid back, it will try very hard to make loans that will get paid back. If a bank (1) makes loans and then (2) sells them to someone else, who will lose money if they don’t get paid back, it will try less hard. It will still try! It is good for the bank’s long-term business to make good loans, so it can keep selling them to willing buyers. It’s just less important for the loans to all be good, if the buyers take the losses.

另一方面，这种模式存在一个明显的问题，你可能还记得 2008 年的情况。这个问题就是道德风险。如果一家银行（1）发放贷款，然后（2）如果贷款无法偿还就会亏损，那么它会非常努力地确保贷款能够被偿还。如果一家银行（1）发放贷款，然后（2）将其出售给其他人，而如果贷款无法偿还，亏损由其他人承担，那么它的努力程度就会降低。它仍然会努力！为了银行的长期业务，它仍然需要发放优质贷款，以便能够继续将其出售给愿意购买的买家。只是，如果买家承担损失，那么所有贷款都必须是优质贷款的必要性就会降低。

Here are Bloomberg’s Carmen Arroyo, Laura Noonan and Sridhar Natarajan on SRTs4:

以下是彭博社的 Carmen Arroyo、Laura Noonan 和 Sridhar Natarajan 关于 SRTs4 的报道：

A campaign by US watchdogs to strengthen banks’ balance sheets is spurring lenders to explore creative ways to reduce risks on their books. Increasingly, they’re turning to significant risk transfers — a bit of Wall Street alchemy that shifts the first losses on loans for things like cars, commercial properties and corporate operations — to investors such as [EJF Capital co-founder Neal] Wilson. If the loans perform well, he can reap a tidy profit.

美国监管机构加强银行资产负债表的行动正在促使贷款机构探索创造性的方法来降低账面风险。它们越来越多地转向重大风险转移——一种华尔街的“炼金术”，将汽车、商业地产和企业运营等贷款的首批损失转移给投资者，如[EJF Capital 联合创始人 Neal] Wilson。如果这些贷款表现良好，他可以获得可观的利润。

监管并没有增加社会的整体效益。

SRTs in the US are relatively new and very hot:

SRT 在美国相对较新且非常热门

That’s creating a combination of eager investors, too few deals to satiate their demand and participants on both sides who are new to SRTs — stoking concerns about the possibility that the budding market may soon be polluted with deals in which buyers accepted loose terms. Veterans say it’s important, for example, to scrutinize the fine print, such as replenishment clauses that could allow a bank to jam in lower-quality loans. …

这导致了一个组合：热切的投资者、过少的交易无法满足他们的需求，以及双方参与者对 SRTs 都不够熟悉——引发了人们对这一新兴市场可能很快被买家接受宽松条款的交易所污染的担忧。资深人士表示，例如，仔细审查细则至关重要，比如补充条款可能允许银行塞入质量较低的贷款。…

Ideally, buyers and sellers engage in a back-and-forth over which loans to include in pools, especially in corporate debt and commercial real estate portfolios. That can help set up deals that may be renewed again and again, even if the economy shifts, until the debts are repaid.

理想情况下，买卖双方就应包含在哪些贷款池中进行来回协商，特别是在公司债务和商业房地产投资组合中。这有助于达成可以一再续约的交易，即使经济发生变化，直到债务偿还完毕。

But the influx of new investors has also meant more are willing to take a “quantitative approach” to SRTs and focus their efforts on blind pools, said Diameter co-founder Scott Goodwin. That’s not something Diameter wants to do. It’s still sweating over the quality of the underlying collateral before getting the transaction over the finish line, he said.

但新投资者的涌入也意味着更多人愿意对 SRTs 采取“量化方法”，并将精力集中在盲池上，Diameter 联合创始人斯科特·古德温表示。这并不是 Diameter 想要做的事情。他表示，公司仍在努力评估基础抵押品的质量，以确保交易顺利完成。

If your model is roughly “banks have the sales force to make loans, and investors have the money to take the risk,” who is in charge of underwriting the loans? Who gets to decide which loans are a good credit risk? One obvious answer is “the banks, since they are the ones dealing with the borrowers”; the other obvious answer is “the investors, since they are the ones on the hook for bad credit decisions.” A reasonable answer — Diameter’s answer — is “both of them,” but a plausible, “quantitative” answer is “neither of them.”

如果你的模型大致是“银行拥有发放贷款的销售团队，而投资者拥有承担风险的资金”，那么谁负责贷款的承销？谁来决定哪些贷款是良好的信用风险？一个显而易见的答案是“银行，因为他们直接与借款人打交道”；另一个显而易见的答案是“投资者，因为他们要承担信用决策失误的后果。” 一个合理的答案——Diameter 的答案——是“两者都是”，但一个看似合理的“量化”答案是“两者都不是。”

2、《2025-01-15 Matt Levine.Capital One High-Yield Account Wasn’t》

I’m sorry but the whole theory of banking, the basic core of how banking works, is that a lot of people don’t pay attention to the interest rate on their bank accounts, so banks can pay them below-market rates. This is not a minor peccadillo of nasty bankers; this is not a little malfeasance at the periphery of a basically customer-centric industry; this is not “oh those sneaky bankers, trying to keep deposit rates low even as the Fed raises rates.” This is what banking is. “Banks can keep deposit rates low even as the Fed raises rates” is why there is banking.

对不起，但整个银行理论，银行运作的基本核心，是很多人不关注他们银行账户的利率，所以银行可以支付低于市场的利率。这不是可恶银行家的小过失；这不是一个基本以客户为中心的行业边缘的小不当行为；这不是“哦，那些狡猾的银行家，试图在美联储加息时保持存款利率低”。这就是银行业。“银行可以在美联储加息时保持存款利率低”是银行存在的原因。

I am exaggerating? A little? But we have talked about this theme a lot, because in 2023 there was a mini-crisis at US regional banks. The quick story of the crisis was that the Federal Reserve raised interest rates rapidly, and a lot of depositors took their money out of US regional banks and put it elsewhere. (Other banks, Treasury bills, money market funds, etc.) The banks had had a lot of deposits on which they paid roughly zero interest, which provided them funding that was cheap and that they thought was pretty stable. But it turned out not to be stable: The deposits left, and the banks mostly had to replace their deposits with more expensive money, undermining their business model and occasionally leading to bank failures.

我有些夸张？有一点？但我们已经多次讨论过这个主题，因为在 2023 年，美国地区银行发生了一次小型危机。危机的简要情况是，美联储迅速提高了利率，许多储户将他们的钱从美国地区银行取出并存到其他地方。（其他银行、国库券、货币市场基金等）这些银行拥有大量存款，几乎不支付利息，这为它们提供了廉价且被认为相当稳定的资金。但事实证明这并不稳定：存款流失，银行大多不得不以更高成本的资金替换存款，削弱了它们的商业模式，偶尔导致银行倒闭。

My explanation of this crisis was that banks, and bank regulators, had grown accustomed to a model in which depositors basically don’t pay attention to interest rates. The 2023 crisis was a surprise because deposits turned out to be less stable than everyone thought. Why deposits turned out to be unstable is an interesting question, and we have discussed some speculative explanations. “Online banking makes it easier to move your deposits from one bank to another, and social media makes news travel faster” is a popular set of explanations. (“Game’s the same, just got more fierce,” as the vice chairman of the Federal Deposit Insurance Corp. put it.) In traditional theory, depositors don’t pay attention to interest rates on bank accounts because it is hard to pay attention to that, and hard to move your money even if you do; in modern online life those things are easier, so depositors might pay more attention.

我对这场危机的解释是，银行和银行监管机构已经习惯了一种模式，即储户基本上不关注利率。2023 年的危机令人惊讶，因为存款比大家想象的要不稳定。为什么存款会变得不稳定是一个有趣的问题，我们已经讨论了一些推测性的解释。“网上银行使得将存款从一个银行转移到另一个银行变得更容易，社交媒体使得新闻传播得更快”是一套流行的解释。（“游戏规则没变，只是变得更激烈了，”正如联邦存款保险公司副主席所说。）在传统理论中，储户不关注银行账户的利率，因为关注这些很难，即使关注了也很难转移资金；在现代在线生活中，这些事情变得更容易，所以储户可能会更加关注。

Here is a recent paper by Itamar Drechsler, Alexi Savov, Philipp Schnabl and Olivier Wang on “Deposit Franchise Runs,” which begins with an explanation of the basic model of banking:

这是 Itamar Drechsler、Alexi Savov、Philipp Schnabl 和 Olivier Wang 最近发表的一篇关于“存款特许经营挤兑”的论文，开篇解释了银行业的基本模型：

In the year leading up to March 2023, the Federal Reserve raised short-term interest rates by nearly 5%. Long-term rates rose by 2.5%. The sharp increase had a big impact on the value of bank assets. Banks held $17 trillion of long-term loans and securities with an average duration of about four years. The implied loss on these assets was thus 4 × 0.025 × $17 = $1.7 trillion. Bank equity stood at $2.2 trillion, hence these losses had the potential to wipe out most of the capital of the banking sector.

截至 2023 年 3 月的一年中，美联储将短期利率提高了近 5%。长期利率上升了 2.5%。这一急剧增长对银行资产的价值产生了重大影响。银行持有 17 万亿美元的长期贷款和证券，平均期限约为四年。因此，这些资产的隐含损失为 4 × 0.025 × $17 = $1.7 万亿美元。银行股本为 2.2 万亿美元，因此这些损失可能会消耗掉银行业大部分的资本。

Strikingly, bank investors were unconcerned. The value of bank stocks … held up well and tracked the overall stock market throughout 2022 and early 2023. Bank profits, as captured by their net interest margins, also held up and even rose slightly despite the large maturity mismatch. Evidently, banks had another source of profits supporting their value.

令人惊讶的是，银行投资者并不担心。银行股票的价值表现良好，并在整个 2022 年和 2023 年初与整体股市保持一致。银行利润，如其净利息差所反映的，也保持稳定，甚至略有上升，尽管存在较大的期限错配。显然，银行还有其他利润来源支撑其价值。

That source of profits was deposits. Deposits are special, as reflected in their rates, which are low and insensitive to market rates. This “low beta” makes deposits much more profitable for banks when interest rates rise. Historically, deposit betas have been about 0.4. As a result, when the Fed raised rates banks went from earning 0% to a (1 − 0.4) × 0.05 = 3% spread on deposits. Deposits stood at $17 trillion, so deposit profits rose by $510 billion per year. …

利润的来源是存款。存款是特殊的，这反映在它们的利率上，这些利率较低且对市场利率不敏感。这种“低贝塔”使得当利率上升时，存款对银行来说更加有利可图。历史上，存款贝塔值约为 0.4。因此，当美联储加息时，银行从赚取 0%变为在存款上获得(1 − 0.4) × 0.05 = 3%的利差。存款总额为 17 万亿美元，因此存款利润每年增加 5100 亿美元。…

The bank has deposit market power, which allows it to pay a deposit rate that is low and insensitive to the market interest rate, i.e. its deposit beta is less than one. The bank thus earns a deposit spread that increases with the market rate. This makes its deposit franchise valuable.

银行拥有存款市场势力，这使其能够支付低且对市场利率不敏感的存款利率，即其存款贝塔系数小于一。因此，银行赚取的存款利差随着市场利率的上升而增加。这使得其存款特许经营权具有价值。

When interest rates go up, a bank’s long-term assets (loans, bonds) lose value, which could make the bank insolvent on a mark-to-market basis. (Which actually happened to Silicon Valley Bank in 2023.) Fortunately, the bank has another asset, “the deposit franchise,” whose value goes up when interest rates go up. “The deposit franchise” is just depositors not paying attention to interest rates. That’s what it is! The bank can “pay a deposit rate that is low and insensitive to the market rate.” “Deposit franchise” is a nicer way to put that, and it gestures at the reasons that this funding advantage exists: There’s a “franchise,” the bank has built up customer relationships and brand loyalty, it has branches and tellers and Little League sponsorships, so people leave their money in the bank without fighting for every last basis point of interest.1

当利率上升时，银行的长期资产（贷款、债券）会贬值，这可能会使银行在市值基础上资不抵债。（这实际上发生在 2023 年的硅谷银行。）幸运的是，银行还有另一项资产，“存款特许权”，其价值在利率上升时会上升。“存款特许权”就是存款人不关注利率。这就是它的本质！银行可以“支付一个低且对市场利率不敏感的存款利率。”“存款特许权”是一个更好听的说法，并且暗示了这种资金优势存在的原因：有一个“特许权”，银行建立了客户关系和品牌忠诚度，它有分行、柜员和小联盟赞助，所以人们把钱留在银行里，而不是为每一个基点的利息而争斗。

Again, the paper is titled “Deposit Franchise Runs,” and the point is that this theory is not as true as it used to be. “In this paper,” say Drechsler et al., “we argue that the deposit franchise is a runnable asset, and that this creates fragility when interest rates rise.” (“A new type of bank run,” they call it.) Silicon Valley Bank’s depositors did take a lot of their money out; the value of its deposit franchise evaporated. At least some banks had depositors who were systematically more sensitive to interest rates and their financial condition than bankers (and regulators) would have predicted, causing problems for those banks.

这篇论文题为《存款特许权挤兑》，其观点是这一理论不再像过去那样真实。Drechsler 等人在论文中表示：“我们认为存款特许权是一种可挤兑的资产，这在利率上升时会造成脆弱性。”（他们称之为“一种新型银行挤兑”。）硅谷银行的储户确实提取了大量资金，其存款特许权的价值蒸发了。至少有些银行的储户对利率和其财务状况的敏感性比银行家（和监管者）预测的要高，这给这些银行带来了问题。

小银行的风险更大。

But most banks did fine, and big banks generally did — as traditional theory predicts — have rising profits as rates went up. As rates went up, those banks were able to earn more interest on their assets (by putting money into investments that paid market interest rates) without paying much more interest on their liabilities (by not raising rates on their deposits). Their deposit franchises held up.

但大多数银行表现良好，大型银行通常也表现良好——正如传统理论所预测的那样——随着利率上升，利润也在上升。随着利率上升，这些银行能够通过将资金投入支付市场利率的投资中来赚取更多的资产利息，而无需在负债上支付更多的利息（因为它们没有提高存款利率）。它们的存款业务保持稳定。

Anyway: 无论如何：

[Yesterday], the Consumer Financial Protection Bureau (CFPB) sued Capital One, N.A., and its parent holding company, Capital One Financial Corp., for cheating millions of consumers out of more than $2 billion in interest. The CFPB alleges that Capital One promised consumers that its flagship “360 Savings” account provided one of the nation’s “best” and “highest” interest rates, but the bank froze the interest rate at a low level while rates rose nationwide. Around the same time, Capital One created a virtually identical product, “360 Performance Savings,” that differed from 360 Savings only in that it paid out substantially more in interest—at one point more than 14 times the 360 Savings rate. Capital One did not specifically notify 360 Savings accountholders about the new product, and instead worked to keep them in the dark about these better-paying accounts. The CFPB alleges that Capital One obscured the new product from its 360 Savings accountholders and cost millions of consumers more than $2 billion in lost interest payments.

[昨天]，消费者金融保护局 (CFPB) 起诉 Capital One, N.A. 及其母公司 Capital One Financial Corp.，指控其欺骗数百万消费者，导致超过 20 亿美元的利息损失。CFPB 指控 Capital One 向消费者承诺其旗舰产品“360 Savings”账户提供全国“最佳”和“最高”的利率，但银行在全国利率上升时将利率冻结在较低水平。大约在同一时间，Capital One 创建了一个几乎相同的产品“360 Performance Savings”，与 360 Savings 的区别仅在于其支付的利息大幅增加——一度超过 360 Savings 利率的 14 倍。Capital One 并未专门通知 360 Savings 账户持有人有关新产品的信息，而是努力让他们对这些收益更高的账户保持不知情。CFPB 指控 Capital One 向其 360 Savings 账户持有人隐瞒新产品，导致数百万消费者损失超过 20 亿美元的利息收入。

I love it.2 “Capital One did not specifically notify 360 Savings accountholders about the new product.” Imagine! Imagine Capital One calling up all of its depositors earning 0.3% and saying “hey just FYI, we have another product that is basically identical but pays you 4.25%, want to switch?”

我喜欢它。2 “Capital One 并没有特别通知 360 储蓄账户持有人关于新产品的消息。” 想象一下！想象一下 Capital One 给所有赚取 0.3%利息的储户打电话说：“嘿，仅供参考，我们有另一个基本相同但支付 4.25%利息的产品，想要转换吗？”

Oh, fine, that is not actually that hard to imagine. That’s really the minimum you’d expect from an adviser who is looking out for you, just straightforwardly good customer service.

哦，好吧，那其实并不难想象。这真的是你对一个为你着想的顾问的最低期望，只是简单的良好客户服务。

But it’s bad banking! Capital One is not an adviser who is looking out for its depositors! It’s a bank! Its relationship with depositors is “we want to pay you a deposit rate that is low and insensitive to the market rate.” That is the very heart of the relationship. It’s not very nice, for you, if you’re the depositor. But that’s banking, man. If you are a depositor at a bank, you are not so much “a customer” as you are “raw material.” You are valuable to the extent you don’t pay attention. If they went out of their way to pay you a market rate on your deposits, they would not be a bank.

但这就是糟糕的银行业务！Capital One 不是为储户着想的顾问！它是一家银行！它与储户的关系是“我们想给你一个低于市场利率且对市场利率不敏感的存款利率。”这就是关系的核心。如果你是储户，这对你来说不太好。但这就是银行业，伙计。如果你是银行的储户，你与其说是“客户”，不如说是“原材料。”你有价值的程度在于你不注意。如果他们特意给你支付市场利率的存款，他们就不是银行了。

I.H.142.Right Business.Matt Levine.Banks

I.H.142.Right Business.Matt Levine.Banks

1、《2024-11-18 Matt Levine.SRTs》

2、《2025-01-15 Matt Levine.Capital One High-Yield Account Wasn’t》

热门主题

Related Articles

I.H.111.Right Business.Warren Buffett.Private Equity

I.H.138.Right Business.Warren Buffett.Floats

I.H.116.Right Business.Warren Buffett.Pharma

I.H.301.Right Price.Warren Buffett.Mr.Market

I.H.100.Warren Buffett.Right Business