I.H.RB.117.Banks
银行业务内嵌了复杂性，而复杂性本身就天然提供了更好的机会，让人犯错而不易被察觉，或者让人舞弊而不易被发现——那么你遇到的舞弊和错误就一定会比那种简单生意更多。
就像卖天然气的生意一定比卖沙子的生意发生更多爆炸一样：这些大型金融机构这种生意，按其本性就会有更多问题，而且永远如此。就算金融机构是政府持有，也同样如此【Appendix.3】。
1、系统性的风险
把钱借给别人就很难，把“不稳定的钱”借给别人是难上加难。
（1）结构性的脆弱
和保险业的浮存金不一样，银行存款不能长期锁定，再往深想一想更有问题，储户和放在银行里的存款天生不稳定，“人一旦害怕，就是真的害怕”，“任何资本金要求，都无法真正保护你免于一场“真实的挤兑”【Appendix.4】。
当出现急速变化、系统性风险的时候，存款端出现挤兑，贷款又不能同步收回，再叠加20倍的杠杆，存在清零的风险，这也是中央银行和存款保险制度存在的原因。
（2）能够被评估的确定性很差
金融公司的可评估性远不如一家有具体产品的公司，或者一家零售企业，银行的贷款质量就是一个很难评估的业务，计提的坏账准备金高一点、低一点，这是没有人能准确计算的数字，甚至“合理区间”都很难，使得报出“特定利润数字”非常容易。
更危险的是，这不仅是估算困难，还制造了一套扭曲的激励：管理层通过少提准备金可以虚报利润、拿到奖金，而代价由未来的股东承担。“随着时间推移，会看到各种舞弊，也会看到各种“大错”，年复一年，一个接一个”【Appendix.3】。
很多银行的问题并不是突然出现的，而是被长期隐藏的，其他行业的公司可以用现金是否紧张”更早地嗅到危险，但金融公司可能在账面上依然“钱很多”的时候，就已经越过了偿付能力的临界点，它们可以在周围仍然有大把钱流动的情况下，照样走到资不抵债的那一步【Appendix.2】。
2、制度性强制力
把“不稳定的钱”借给别人，是银行的第1个结构性矛盾；钱又是完全同质化的产品，这是第2个结构性矛盾，你无法让一笔钱（贷款）比竞争对手的"更好"。在这个行业里，唯一可见的竞争力指标就是利差与规模增速，而拉高利差的方式包括：（1）短借长贷但又不敢拉的太长，2023年的硅谷银行（SVB）是很好的例子；（2）放松贷款标准、追逐更高风险的借款人。
这是一个极具破坏力的机制：当同行在房地产泡沫或某类高收益资产上赚钱时，保守的银行在短期内"看起来落后"——客户流失、股东施压。更要命的是，"好的信贷标准"与"激进的信贷标准"在贷款发放的那一刻不容易区分，差异要等到信用周期的下行段才能暴露。
这就是为什么银行业的危机是反复上演的，巴菲特的评论中基本上把"制度性强制力（institutional imperative）"和银行业做了强关联【Appendix.1】——银行家们以旅鼠般的狂热相互复制错误的做法，然后经历旅鼠般的命运。
3、情绪结构和高回报、高增长
如果能排除系统性的、被清零的风险，又能找到对“制度性强制力”有敏感性的管理团队，扛过这两个风险，银行因为杠杆高一直是个比较赚钱的业务，ROA 只需要 1% 左右，ROE 就可能达到 15% 以上，巴菲特看中Wells Fargo的就是这个点【Appendix.1】。
Wells Fargo is big—it has $56 billion in assets—and has been earning more than 20% on equity and 1.25% on assets.
Wells Fargo 规模庞大——它拥有 560 亿美元的资产——且净资产收益率一直超过 20%，资产回报率为 1.25%。
DBS和JPM的ROE都在15%以上，并且能够将留存利润以相近的回报率进行再投资，能够长期做到这两点的商业模式并不多。但是综合型金融公司可被评估性的弱点始终存在，如果是差不多的管理水平，业务单一的专业型银行，比如，AMEX、Capital One可能会更好一些。
Appendix
Appendix.1.《1991-03-01 Warren Buffett's Letters to Berkshire Shareholders》
Lethargy bordering on sloth remains the cornerstone of our investment style: This year we neither bought nor sold a share of five of our six major holdings. The exception was Wells Fargo, a superbly-managed, high-return banking operation in which we increased our ownership to just under 10%, the most we can own without the approval of the Federal Reserve Board. About one-sixth of our position was bought in 1989, the rest in 1990.
近乎懒散的沉稳依然是我们投资风格的基石：今年，对于我们的六大重仓股中的五只，我们既没有买入也没有卖出过一股。唯一的例外是 Wells Fargo，这是一家管理卓越、高回报的银行业务，我们将其持股比例增加到了略低于 10% 的水平，这是在未经 Federal Reserve Board 批准的情况下我们所能持有的最高限额。我们持仓中约六分之一是在 1989 年买入的，其余部分是在 1990 年买入的。
The banking business is no favorite of ours. When assets are twenty times equity—a common ratio in this industry—mistakes that involve only a small portion of assets can destroy a major portion of equity. And mistakes have been the rule rather than the exception at many major banks. Most have resulted from a managerial failing that we described last year when discussing the “institutional imperative:” the tendency of executives to mindlessly imitate the behavior of their peers, no matter how foolish it may be to do so. In their lending, many bankers played follow-the-leader with lemming-like zeal; now they are experiencing a lemming-like fate.
银行业并非我们的心头好。当资产是净资产的 20 倍时——这是该行业常见的比例——仅涉及一小部分资产的错误就可能摧毁大部分净资产。而在许多大银行中，错误一直是常态而非例外。大多数错误源于我们去年在讨论“制度性强制力（institutional imperative）”时描述的一种管理失败：即高管们倾向于盲目模仿同行的行为，无论这样做有多么愚蠢。在贷款方面，许多银行家以旅鼠般的狂热玩着“跟随领导者”的游戏；现在，他们正经历着旅鼠般的命运。
Because leverage of 20:1 magnifies the effects of managerial strengths and weaknesses, we have no interest in purchasing shares of a poorly-managed bank at a “cheap” price. Instead, our only interest is in buying into well-managed banks at fair prices.
由于 20:1 的杠杆放大了管理优势和劣势的影响，我们对以“便宜”的价格购买管理不善的银行股票没有兴趣。相反，我们唯一的兴趣是以公平的价格买入管理良好的银行。
银行的存款是不能长期锁定的，这跟保险业的浮存金不一样，当出现急速变化、系统性风险的时候，存款端出现挤兑，贷款又不能同步收回，再叠加20倍的杠杆，存在清零的风险，这也是中央银行和存款保险制度存在的原因，参考：《2018-02-24 Warren Buffett's Letters to Berkshire Shareholders》。
Float will probably increase slowly for at least a few years. When we eventually experience a decline, it will be modest – at most 3% or so in any single year. Unlike bank deposits or life insurance policies containing surrender options, p/c float can’t be withdrawn. This means that p/c companies can’t experience massive “runs” in times of widespread financial stress, a characteristic of prime importance to Berkshire that we factor into our investment decisions.
至少在未来几年内，浮存金可能会缓慢增长。当我们最终经历浮存金下降时，降幅将是温和的——任何单一年份最多下降3%左右。与银行存款或包含退保选项的人寿保险保单不同，财产意外险的浮存金是不能被提取的。这意味着财产意外险公司在金融普遍承压时期不会经历大规模的“挤兑”，这一特性对伯克希尔至关重要，也是我们做投资决策时会考虑的因素。
Charlie and I never will operate Berkshire in a manner that depends on the kindness of strangers – or even that of friends who may be facing liquidity problems of their own. During the 2008-2009 crisis, we liked having Treasury Bills – loads of Treasury Bills – that protected us from having to rely on funding sources such as bank lines or commercial paper. We have intentionally constructed Berkshire in a manner that will allow it to comfortably withstand economic discontinuities, including such extremes as extended market closures.
查理和我绝不会以依赖陌生人善意的方式来运营伯克希尔——甚至也不会依赖那些自身可能面临流动性问题的朋友。在2008-2009年的危机期间，我们乐于持有美国国库券——大量的国库券——这使我们不必依赖银行或商业票据等金融资源。我们有意将伯克希尔打造成能够从容抵御经济中断的模式，包括像长期市场关闭这样的极端情况。
把“不稳定的钱”借给别人，是银行的第1个结构性矛盾；钱又是完全同质化的产品，这是第2个结构性矛盾，你无法让一笔钱（贷款）比竞争对手的"更好"。在这个行业里，唯一可见的竞争力指标就是利差与规模增速，而拉高利差的方式包括：（1）短借长贷但又不敢拉的太长，2023年的硅谷银行（SVB）是很好的例子；（2）放松贷款标准、追逐更高风险的借款人。
这是一个极具破坏力的机制：当同行在房地产泡沫或某类高收益资产上赚钱时，保守的银行在短期内"看起来落后"——客户流失、股东施压。更要命的是，"好的信贷标准"与"激进的信贷标准"在贷款发放的那一刻不容易区分，差异要等到信用周期的下行段才能暴露。
巴菲特的评论中基本上把“制度性强制力（institutional imperative）”和银行业做了强关联，管理质量是银行业的关键因素。
With Wells Fargo, we think we have obtained the best managers in the business, Carl Reichardt and Paul Hazen. In many ways the combination of Carl and Paul reminds me of another—Tom Murphy and Dan Burke at Capital Cities/ABC. First, each pair is stronger than the sum of its parts because each partner understands, trusts and admires the other. Second, both managerial teams pay able people well, but abhor having a bigger head count than is needed. Third, both attack costs as vigorously when profits are at record levels as when they are under pressure. Finally, both stick with what they understand and let their abilities, not their egos, determine what they attempt. (Thomas J. Watson Sr. of IBM followed the same rule: “I’m no genius,” he said. “I’m smart in spots—but I stay around those spots.”)
对于 Wells Fargo，我们认为我们得到了业内最好的管理者 Carl Reichardt 和 Paul Hazen。在很多方面，Carl 和 Paul 的组合让我想起另一对搭档——Capital Cities/ABC 的 Tom Murphy 和 Dan Burke。首先，每一对组合都比其各部分之和更强大，因为每一位合伙人都理解、信任并钦佩对方。其次，这两个管理团队都给予能干的人丰厚的报酬，但都极其厌恶拥有超出需要的员工人数。第三，无论利润处于创纪录水平还是面临压力，两支团队都会同样积极地缩减成本。最后，两者都坚持自己理解的领域，让他们的能力而不是自我意识来决定他们的尝试。（IBM 的 Thomas J. Watson Sr. 也遵循同样的规则：“我不是天才，”他说，“我在某些方面很聪明——但我只待在那些领域周围。”）
检查的点：
NICO在1986-1999年让收入持续萎缩，因为定价纪律；Wells Fargo的Carl Reichardt在加州房地产泡沫期间拒绝跟风。这是Basic trust的直接证据——不需要同行认可来维持安全感。
Our purchases of Wells Fargo in 1990 were helped by a chaotic market in bank stocks. The disarray was appropriate: Month by month the foolish loan decisions of once well-regarded banks were put on public display. As one huge loss after another was unveiled—often on the heels of managerial assurances that all was well—investors understandably concluded that no bank’s numbers were to be trusted. Aided by their flight from bank stocks, we purchased our 10% interest in Wells Fargo for $290 million, less than five times after-tax earnings, and less than three times pre-tax earnings.
我们在 1990 年对 Wells Fargo 的购买得益于银行股市场的混乱。这种乱局是理所应当的：昔日备受推崇的银行那些愚蠢的贷款决策月复一月地被公之于众。随着一笔又一笔巨额亏损被揭露——而且往往紧随管理层关于“一切安好”的保证之后——投资者理所当然地认为，没有任何一家银行的数据是值得信赖的。借着他们逃离银行股的机会，我们以 2.9 亿美元买入了 Wells Fargo 10% 的权益，这还不到其税后收益的五倍，不到其税前收益的三倍。
Wells Fargo is big—it has $56 billion in assets—and has been earning more than 20% on equity and 1.25% on assets. Our purchase of one-tenth of the bank may be thought of as roughly equivalent to our buying 100% of a $5 billion bank with identical financial characteristics. But were we to make such a purchase, we would have to pay about twice the $290 million we paid for Wells Fargo. Moreover, that $5 billion bank, commanding a premium price, would present us with another problem: We would not be able to find a Carl Reichardt to run it. In recent years, Wells Fargo executives have been more avidly recruited than any others in the banking business; no one, however, has been able to hire the dean.
Wells Fargo 规模庞大——它拥有 560 亿美元的资产——且净资产收益率一直超过 20%，资产回报率为 1.25%。我们购买该银行十分之一的股份，可以大致等同于我们买下了一家具有相同财务特征、资产规模为 50 亿美元银行的 100% 股权。但如果我们真的去进行这样一笔收购，我们需要支付的金额大约是投资 Wells Fargo 所花的 2.9 亿美元的两倍。此外，那家开出溢价的高价银行还会给我们带来另一个问题：我们无法找到一个 Carl Reichardt 来管理它。近年来，Wells Fargo 的高管比银行业中任何其他人都更受猎头青睐；然而，没有人能够挖走这位“院长”。
Of course, ownership of a bank—or about any other business—is far from riskless. California banks face the specific risk of a major earthquake, which might wreak enough havoc on borrowers to in turn destroy the banks lending to them. A second risk is systemic—the possibility of a business contraction or financial panic so severe that it would endanger almost every highly-leveraged institution, no matter how intelligently run. Finally, the market’s major fear of the moment is that West Coast real estate values will tumble because of overbuilding and deliver huge losses to banks that have financed the expansion. Because it is a leading real estate lender, Wells Fargo is thought to be particularly vulnerable.
当然，拥有一家银行——或几乎任何其他业务——都绝非没有风险。California 的银行面临着大地震的特定风险，地震可能给借款人造成足够的破坏，进而摧毁向他们提供贷款的银行。第二个风险是系统性的——即可能发生极其严重的经济萎缩或金融恐慌，以至于几乎危及所有高杠杆机构，无论其经营得多么明智。最后，市场目前主要的担忧是，由于过度建设，West Coast 的房地产价值将会暴跌，并给资助了这一扩张的银行带来巨额损失。由于 Wells Fargo 是领先的房地产贷款人，它被认为特别脆弱。
None of these eventualities can be ruled out. The probability of the first two occurring, however, is low and even a meaningful drop in real estate values is unlikely to cause major problems for well-managed institutions. Consider some mathematics: Wells Fargo currently earns well over $1 billion pre-tax annually after expensing more than $300 million for loan losses. If 10% of all $48 billion of the bank’s loans—not just its real estate loans—were hit by problems in 1991, and these produced losses (including foregone interest) averaging 30% of principal, the company would roughly break even.
这些可能性都不能被排除。然而，前两种情况发生的概率很低，即使房地产价值出现实质性下跌，也不太可能给管理良好的机构带来重大问题。考虑一下数学计算：Wells Fargo 目前在扣除超过 3 亿美元的贷款损失支出后，每年税前收入仍远超 10 亿美元。如果 1991 年该银行全部 480 亿美元贷款中的 10%——不仅仅是房地产贷款——出现问题，且这些问题导致的损失（包括利息损失）平均占本金的 30%，该公司也大约能达到盈亏平衡。
1、特定地区的风险；（2）金融危机的风险；（3）特定业务的风险，巴菲特的结论是前两者概率低，第三个风险真实但可以用数学测算承受能力。
可以相信前几大银行是“大而不倒”的，任何政府都很承受大型银行倒闭的风险，中国的银行更不担心前两者的风险，特别是第（2）点，大银行很难发生挤兑，可能都没有实施操作的方法。
但是对于新加坡的银行，比如，DBS，巴菲特列举的三个风险要真实的多。2023年硅谷银行（SVB）的崩溃是个可以参考的例子，由于持有大量长期国债，利率上升导致资产端浮亏，引起储户担忧后存款大规模流出，银行被迫在市场最差的时候贱卖长期资产，两边同时崩塌。
SVB 给了一个现代版本：数字化提款速度可以把“挤兑”从几天压缩到几小时——官方材料披露，2023-03-09 单日流出约 420 亿美元存款，且第二天还有大量提款指令排队；同时 90%+ 存款为未保险存款，先天更容易在恐慌中迁移。这也解释了为什么美国事后动用了系统性风险例外来保护未保险存款人，而中国的银行可能采用了成本更低的策略。
A year like that—which we consider only a low-level possibility, not a likelihood—would not distress us. In fact, at Berkshire we would love to acquire businesses or invest in capital projects that produced no return for a year, but that could then be expected to earn 20% on growing equity. Nevertheless, fears of a California real estate disaster similar to that experienced in New England caused the price of Wells Fargo stock to fall almost 50% within a few months during 1990. Even though we had bought some shares at the prices prevailing before the fall, we welcomed the decline because it allowed us to pick up many more shares at the new, panic prices.
像那样的一年——我们认为这只是一种极低的可能性，而非必然——并不会让我们感到困扰。事实上，在 Berkshire，我们非常乐意收购那些在一年内不产生回报、但随后预期能以不断增长的净资产获得 20% 回报的业务，或对其进行资本项目投资。尽管如此，由于担心 California 发生类似于 New England 经历过的房地产灾难，Wells Fargo 的股价在 1990 年的短短几个月内下跌了近 50%。即便我们在大跌前的价位已经买入了一些股份，我们依然对这种下跌表示欢迎，因为它使我们能够以全新的、惊恐的价格买入更多股份。
15%以上的ROE+可以不断投入新的资本，能够长期做到这两点估计不多。如果能找到一家银行，其管理层在过去的压力周期中展现出清晰的纪律——拒绝跟风扩张、主动收缩高风险业务、在同行贪婪时保守——那么这家银行长期15%+的ROE是可以持续的，而且每一元再投入的新资本都在以这个速率复利。
这种商业模式的内在价值增长率，等于ROE × 留存比率。如果DBS的ROE维持16%，留存60%利润，内在价值每年自然增长约9.6%，再加上分红，总回报不难超过12%。
Investors who expect to be ongoing buyers of investments throughout their lifetimes should adopt a similar attitude toward market fluctuations; instead many illogically become euphoric when stock prices rise and unhappy when they fall. They show no such confusion in their reaction to food prices: Knowing they are forever going to be buyers of food, they welcome falling prices and deplore price increases. (It’s the seller of food who doesn’t like declining prices.) Similarly, at the Buffalo News we would cheer lower prices for newsprint—even though it would mean marking down the value of the large inventory of newsprint we always keep on hand—because we know we are going to be perpetually buying the product.
那些预期一生中会持续购买投资资产的投资者，应当对市场波动采取类似的态：然而，许多人却不合逻辑地在股价上涨时感到欣喜若狂，在股价下跌时感到闷闷不乐。他们在面对食物价格时却不会表现出这种困惑：因为知道自己永远是食物的购买者，他们欢迎价格下跌，而对价格上涨感到不满。（不喜欢价格下跌的是食物的卖方。）同样地，在 Buffalo News，我们会为新闻纸价格的下跌而欢呼——即便这意味着我们必须减记手头始终持有的大量新闻纸库存的价值——因为我们知道我们将永远购买这种产品。
Identical reasoning guides our thinking about Berkshire’s investments. We will be buying businesses—or small parts of businesses, called stocks—year in, year out as long as I live (and longer, if Berkshire’s directors attend the seances I have scheduled). Given these intentions, declining prices for businesses benefit us, and rising prices hurt us.
同样的逻辑指导着我们对 Berkshire 投资的思考。只要我活着（如果 Berkshire 的董事们能参加我安排的降神会，我活得还会更久），我们年复一年地购买企业——或者称为“股票”的企业小部分所有权。基于这些意愿，企业价格的下跌对我们有利，而价格的上涨则对我们有害。
The most common cause of low prices is pessimism—some times pervasive, some times specific to a company or industry. We want to do business in such an environment, not because we like pessimism but because we like the prices it produces. It’s optimism that is the enemy of the rational buyer.
低价最常见的原因是悲观情绪——有时是普遍性的，有时是针对特定公司或行业的。我们希望在这种环境中开展业务，并不是因为我们喜欢悲观，而是因为我们喜欢它所产生的价格。乐观情绪才是理性买家的敌人。
None of this means, however, that a business or stock is an intelligent purchase simply because it is unpopular; a contrarian approach is just as foolish as a follow-the-crowd strategy. What’s required is thinking rather than polling. Unfortunately, Bertrand Russell’s observation about life in general applies with unusual force in the financial world: “Most men would rather die than think. Many do.”
然而，这并不意味着仅仅因为某项业务或股票不受欢迎，它就是一项明智的购买；逆向投资的方法与随大流的策略同样愚蠢。投资需要的是思考而非民意调查。遗憾的是，Bertrand Russell 对普通生活的观察在金融界有着异乎寻常的约束力：“大多数人宁愿死也不愿思考。事实上，很多人确实死了。”
Appendix.2.《2001-04-28 Berkshire Hathaway Annual Meeting》
46. Why Berkshire sold its Freddie Mac and Fannie Mae stakes
伯克希尔为何卖出 Freddie Mac 与 Fannie Mae 的持股
WARREN BUFFETT: Zone 3, please.
WARREN BUFFETT：请第 3 区提问。
AUDIENCE MEMBER: My name is Steve Sondheimer. I live in Chicago and I’m 14 years old. I’m a third generation shareholder and my question is, I noticed that you sold our position in Freddie Mac. What risks do you see in that industry?
观众：我叫 Steve Sondheimer，住在芝加哥，我今年 14 岁。我是第三代股东。我的问题是：我注意到你们卖掉了我们在 Freddie Mac 的持仓。你们在这个行业里看到了哪些风险？
WARREN BUFFETT: Are you Joe’s granddaughter?
WARREN BUFFETT：你是 Joe 的孙女吗？
AUDIENCE MEMBER: Yeah.
观众：是的。
WARREN BUFFETT: Oh, good. We have an amazing number of second and third and even fourth generation shareholders, which I’m delighted with. I mean, I don’t think lots of companies — big companies on the stock exchange — are in that position. It is true, we sold the Freddie Mac stock last year. And there were certain aspects of the business that we felt less comfortable with as they unfolded — and Fannie Mae, too. And the consequences of what we saw may not hurt the companies, I mean, at all. But they made us less comfortable than we were earlier, when, actually, those practices or activities didn’t exist. We did not — I would stress — we did not sell because we were worried about more government regulation of Freddie and Fannie. If anything, just the opposite, so — It was not — it was not — Wall Street occasionally will react negatively to the prospect of more government regulation and the stocks will react sometimes short-term for that reason. But that was not our reason. We were — we felt the risk profile had changed somewhat. Charlie?
WARREN BUFFETT：哦，那太好了。我们有数量惊人的第二代、第三代，甚至第四代股东，我对此非常高兴。我想在交易所里，并没有多少大型公司能拥有这样的股东结构。确实，我们去年卖掉了 Freddie Mac 的股票。随着一些业务做法逐步展开，我们对其中某些方面越来越不舒服——Fannie Mae 也是如此。我们看到的那些变化，其后果未必会伤害这些公司——甚至可能完全不会。但相较于更早以前，当那些做法或活动还不存在时，它们确实让我们更难安心。我要强调：我们卖出并不是因为担心 Freddie 和 Fannie 会面临更多政府监管。若真要说，反而可能恰恰相反，所以——不是——不是——华尔街有时候会对“监管可能加强”作出负面反应，相关股票也会因此在短期内波动。但那不是我们的理由。我们只是觉得它们的风险画像发生了一些变化。Charlie？
CHARLIE MUNGER: Yeah, but that may be a peculiarity of ours. We are especially prone to get uncomfortable around financial institutions.
CHARLIE MUNGER：是的，但这也许是我们自己的一个“怪癖”。我们对金融机构尤其容易感到不舒服。
WARREN BUFFETT: We’re quite sensitive to —
WARREN BUFFETT：我们对——
CHARLIE MUNGER: Yeah.
CHARLIE MUNGER：对。
WARREN BUFFETT: — risk in — whether it’s in banks, insurance companies or in what they call GSEs here, in the case of Freddie and Fannie. We feel there’s so much about a financial institution that you don’t know by looking at just figures, that if anything bothers us a little bit, we’re never sure whether it’s an iceberg situation or not. And that doesn’t mean it is an iceberg situation, in the least, at banks or insurance companies that we pass. But we have seen enough of what happens with financial institutions that push one way or another, that if we get some feeling that that’s going on, we just figure we’ll never see it until it’s too late anyway. So we bid adieu without — and wish them the best — without any implication that they’re doing anything wrong. It’s just that we can’t be 100 percent sure of the fact they’re doing things that we like. And when we get to that situation, it’s different than buying into a company with a product or something, or a retail operation.
WARREN BUFFETT：——金融机构里的风险非常敏感——不管是银行、保险公司，还是这里所谓的 GSE（就 Freddie 和 Fannie 这种）。我们认为，金融机构有太多东西，光看数字是看不出来的；因此，只要有一点点地方让我们不舒服，我们就永远无法确定那是不是“冰山一角”。这并不意味着我们否决过的那些银行或保险公司就一定存在“冰山问题”——一点也不意味着。但我们见过足够多的案例：金融机构只要在某个方向上用力过头，最后会发生什么；所以一旦我们有一种感觉，觉得它可能正在发生，我们就会判断：反正我们也不可能在还来得及之前看清楚，等我们真正看清楚时往往已经太晚了。于是我们就挥手告别——祝他们好运——并不暗示他们做错了什么。只是我们无法百分之百确定：他们做的事情完全符合我们的偏好。到了这种状态，我们的处理方式就不同于去买一家有具体产品的公司，或者一家零售企业。
You could spot troubles usually fairly early in those businesses. You spot troubles in financial institutions late. It’s just the nature of the beast. Charlie?
对于那些“实体生意”，你通常能相当早地就发现麻烦。但金融机构的问题，你往往发现得很晚。这就是这个“物种”的天性。Charlie？
CHARLIE MUNGER: Yeah. Financial institutions tend to make us nervous when they’re trying to do well. (Laughter) That sounds paradoxical, but that’s the way it is.
CHARLIE MUNGER：对。金融机构在“努力把业绩做得很好”的时候，反而最容易让我们紧张。（笑声）听起来很矛盾，但现实就是这样。
WARREN BUFFETT: Financial institutions don’t get in trouble by running out of cash in most cases. Other businesses, you can spot that way. But a financial institution can go beyond the point — and we had banks 10 years ago that did that, en masse — but they can go beyond the point of solvency even while they still have plenty of money around.
WARREN BUFFETT：金融机构大多数情况下并不是因为“现金用光”而陷入麻烦。其他行业的公司，你可以用“现金是否紧张”更早地嗅到危险。但金融机构可能在账面上依然“钱很多”的时候，就已经越过了偿付能力的临界点——我们十年前就见过银行成群结队地这么干。它们可以在周围仍然有大把钱流动的情况下，照样走到资不抵债的那一步。
Appendix.3.《2005-04-30 Berkshire Hathaway Annual Meeting》
41. Don’t blame the ratings agencies
不要怪评级机构
WARREN BUFFETT: Number 4.
AUDIENCE MEMBER: Thank you. Bill Ackman from New York, New York. Four of the handful of triple-A rated companies — AIG, Fannie Mae, Freddie Mac and MBIA — are under formal investigation for accounting shenanigans and are in the process of restating their financials. Like Charlie said before, I think of a triple-A rated company as an exemplar, a company that should behave with the highest accounting and ethical standards. My questions this leads me to are, how can investors comfortably invest in any financial service company when even — when a decent percentage of the triple-A rated companies have false and misleading financials? And I guess the follow-up question is, why don’t the rating agencies do some independent due diligence from an accounting standpoint so that they can help serve as a watch on this issue?
AUDIENCE MEMBER：谢谢。Bill Ackman，来自 New York, New York。有少数几家 AAA 评级公司——AIG、Fannie Mae、Freddie Mac 和 MBIA——其中有四家正因会计造假嫌疑而接受正式调查，并且正在重述财务报表。正如 Charlie 之前说的，我认为一家 AAA 评级公司应当是行业典范，应当以最高的会计与道德标准行事。由此我有两个问题：第一，当相当比例的 AAA 评级金融公司都存在虚假且误导性的财务数据时，投资者怎么还能安心投资任何金融服务公司？第二，评级机构为什么不从会计角度做一些独立的尽职调查，从而在这个问题上发挥“看门人”的作用？
WARREN BUFFETT: Well, financial companies are more difficult to analyze than many companies. I mean, the — it is more — if you take the insurance business, you know, the biggest single element that is very difficult to evaluate, even if you own the company, is the loss and loss adjustment expense reserve. And that has a huge impact on reported earnings of any given period. And the shorter the period, the more the impact can be from just small changes in assumptions. You know, we carry, we’ll say 45 billion of loss reserves. But, you know, if I had to bet my life on whether 45 billion turned out to be a little over or a little under, I mean, it’d be a — I’d think a long time. And you could just as easily have a figure of 45 1/2 billion or 44 1/2 billion. And if you were concerned about reporting given earnings in a given period, that would be an easy game to play. In a bank, you know, it basically is whether the loans are any good. And I’ve been on the boards of banks. And that’s — you know, I’ve gotten surprises.
WARREN BUFFETT：金融公司比很多其他公司更难分析。也就是说——拿保险业举例，最难评估、即便你拥有这家公司也很难评估的最大单一项目，就是“赔付及理赔费用准备金（loss and loss adjustment expense reserve）”。而这会对任何一个期间的已报告利润产生巨大影响。期间越短，仅仅因为假设做了小幅变动，对利润的影响就可能非常大。比如我们账上大概有 450 亿美元的损失准备金。但如果要我拿命赌：450 亿最终会略高一点还是略低一点，我得想很久。你完全可能报 455 亿，也可能报 445 亿。而如果你关心的是在某个期间“报出特定利润数字”，那就太容易玩这个游戏了。银行这边，本质上就是贷款到底好不好。我也在银行董事会待过——我确实遇到过“意外”。很难判断。
It’s tough to tell. It’s — financial companies — if you’re analyzing something like WD-40, you know, or See’s Candy, or our brick business, or whatever, you know, they may have good or bad prospects but you’re not likely to be fooling yourself much about what’s going on currently. But with financial institutions, it’s much tougher. Then you add — throw in derivatives on top of it, and, you know, it’s — no one probably knows, you know, perfectly, what some of the — or even within a reasonable range — the exact condition of some of the biggest, you know, banks in the world. And — but that brings you back to the due diligence question of the agencies. You had very high-grade, very smart — financially smart — people on the boards of both Freddie and Fannie. And yet, you know, one was five billion and one was apparently nine billion. Those are big numbers. And I don’t think those people were negligent. And it’s just, it’s very, very tough to know precisely what’s going on in a financial institution.
这事很难看清。金融公司——如果你分析的是 WD-40，或者 See’s Candy，或者我们的砖厂之类——它们的前景可能好也可能坏，但你不太可能在“当前正在发生什么”这件事上把自己骗得太厉害。但金融机构要难得多。再把衍生品叠加上去，你知道——可能没人能非常准确地知道，甚至很难在一个“合理区间”内知道，世界上一些最大银行的真实状况到底如何。但这又把你带回到评级机构“尽职调查”的问题上。Freddie 和 Fannie 的董事会里都有非常高水平、非常聪明、在金融上很精明的人。然而，你看，最后一个是 50 亿，另一个据说是 90 亿（重述/缺口）。这不是小数字。我不认为那些人是失职的。只是要精确知道一家金融机构里到底发生了什么，真的、真的非常难。
Charlie and I were directors of Salomon [Brothers]. And Charlie was on the audit committee. And I forget the size of a few of those things that you found. But, you know, what wasn’t found — and that doesn’t mean that people below are crooks or anything like that. It just means that it’s very tough with thousands and thousands and thousands of complicated transactions, sometimes involving — the computations involving — multiple variables, it can be very hard to figure out where things stand at any given moment. And, of course, when the numbers get huge on both sides, and you get small changes in these huge numbers, they have this incredible effect on quarterly or yearly figures because it all comes lumped in — those adjustments — come lumped in a short period of time. So I just think you have to accept the fact that insurance, banking, finance companies — we’ve seen all kinds of finance company — both frauds and just big mistakes over time — of just one after another over the years. And the — it’s just a more dangerous field to analyze. It doesn’t mean you can’t make money in it.
Charlie 和我都曾是 Salomon [Brothers] 的董事。Charlie 还在审计委员会里。我已经记不清你当时发现的那些问题里，有几个的规模到底有多大了。但你知道，仍然有很多东西是没被发现的——这并不意味着下面的人就是骗子之类的。只是说：当你面对成千上万、再成千上万笔复杂交易时，其中有些计算还牵涉到多变量、多参数，想在某一个时点准确搞清楚“现在到底是什么状况”，确实非常难。而且，一旦双方的数字都极其庞大，你在这些巨大数字上做出一些很小的调整，就会对季度或年度报表产生惊人的影响——因为这些调整会被集中在很短的期间里一次性反映出来。所以我认为你必须接受这样一个事实：保险、银行、金融公司——我们见过各种各样的金融公司——随着时间推移，你会看到各种舞弊，也会看到各种“大错”，年复一年，一个接一个。这一行的分析本质上就更危险。这并不意味着你赚不到钱。
We’ve made a lot of money on it. But it’s difficult. Now, obviously a GEICO, where you’re insuring pretty much the same thing — auto drivers — and you get — your statistics are much more valid in something like that than they will be if you’re taking something that — like asbestos liability — you’re subject to far greater errors in estimation. Doesn’t mean that people aren’t operating in good faith. But, you know, I would take — just take the asbestos estimates of the 20 largest insurance companies. I will bet you they’re way off, but I don’t know in which direction. And that’s sort of the nature of financial companies. I wouldn’t fault the rating agencies in terms of not being able to dig into the financials and find things that — You know, all of the companies that you’ve talked about have had big name auditors. And our auditors at Berkshire, how many hours did they spend last year? You know, I don’t know whether — what it would be, probably 60, 70,000 hours.
我们就在这类公司上赚过很多钱。但确实很难。很明显，如果是 GEICO 这种业务，你承保的东西相对同质——基本都是汽车驾驶员——那你得到的统计数据要比你去估算类似“石棉责任（asbestos liability）”这种东西更可靠得多；后者的估计误差会大很多。这也不代表相关人员就不是善意经营。但你想想：仅仅看美国最大的 20 家保险公司的石棉责任估计，我敢打赌它们会偏离得很厉害，只是我不知道偏离方向是哪边。这就是金融公司的某种“天性”。所以我不会因为评级机构“挖不出财务报表里的一些问题”就去责怪它们。你提到的这些公司，全都有大牌审计师。我们 Berkshire 的审计师，去年花了多少小时？我不知道具体——可能是 6 万到 7 万小时。
And I’m sure at other — you know, if you take major banks, they’re spending more than that. But, you know, can they be certain of the numbers? I doubt it. Charlie?
而你看那些大型银行，它们花在审计上的时间可能还更多。但你说审计师能不能“确定”这些数字？我怀疑做不到。Charlie？
CHARLIE MUNGER: Yeah. Warren is obviously correct that where you’ve got complexity, which by its very nature provides better opportunities to be mistaken and not have it come to notice, or to be fraudulent and have it not be found out, you’re going to get more fraud and mistakes than you are if you’re selling a business where you shovel sand out of the river and sell it by the truckload. And just as a business that sells natural gas is going to have more explosions than a business that sells sand, a business like these major financial institutions, by its nature, is going to have way more problems. And that will always be true. And it’s true when the financial institutions are owned by governments. In fact, some of the worst financial reporting in the world is done by governments and governments — institutions like government banks in China, et cetera. So, if you don’t like the lack of perfect accounting in financial institutions, you’re in the wrong world.
CHARLIE MUNGER：是的。Warren 显然说得对：当你面对复杂性时——复杂性本身就天然提供了更好的机会，让人犯错而不易被察觉，或者让人舞弊而不易被发现——那么你遇到的舞弊和错误就一定会比那种简单生意更多。就像卖天然气的生意一定比卖沙子的生意发生更多爆炸一样：这些大型金融机构这种生意，按其本性就会有更多问题，而且永远如此。就算金融机构是政府持有，也同样如此。事实上，世界上最糟糕的一些财务报告就是政府做出来的——比如中国的政府银行之类的机构。所以，如果你无法接受金融机构“不可能做到完美会计”，那你就是活在一个错误的世界里。
The third aspect is the way he looks at business models. The best way I can describe this is that it’s as if you and I see an animal, and he sees its DNA. He isn’t interested in whether the animal is furry; all he sees is whether it can run and how well it will reproduce, which are the two key elements that determine whether its species will thrive.
第三个方面是他看待商业模型的方式。我可以这样描述：就好像你我看到一只动物，而他看到的是它的 DNA。他对动物是否有毛发并不感兴趣；他只关注它是否能跑以及它的繁殖能力，这两个关键因素决定了它的物种是否能够繁荣。
WARREN BUFFETT:
It’s the nature of the national institutions both life insurance companies and banks. The, no capital requirements protect you against a real run. I mean, if your liabilities all are payable virtually that day or I should say virtually all your liabilities are payable that day, you can’t run a financial institution and be prepared for that. And that’s why we’ve got the Fed and the FDIC. I mean, you can’t stand, if you’re a life company or a bank, you can’t stand around, you can be the most soundly capitalized firm in town but if I hire, if there were no FDIC and the Fed and you had a bank capitalized with 10% of capital and I had one with 5% of capital and I hired 50 people to go over and start standing right in front of your bank, you’re the guy that’s going to fail first. Then when get through with you they’re going to come over to my bank too, that’s why we don’t do that sort of thing because you can’t contain the fire over on the other guy’s bank. But you can’t, you can’t stand a run. So you need the Federal Reserve and the FDIC. And even with Northern Rock, the UK government and came and said we guarantee everything they still had lines. I mean, when people are scared they’re scared. And there is no reason to leave, I mean, if you see if it’s uninsured and you see a line at a bank where you’ve got your money, get in line. (laughs). You know, buy a place from the guy that’s first in line if necessary, you know. And even if you’re at another bank, get in line there and take your mad money and put it under a mattress. You can always put it back a week later as long as there no penalties, why in the world, you know. That’s why we got a Fed and an FDIC and I think it’s one of the, you know, I think the FDIC and Social Security were the two most important things that came out of the ’30s. I mean, the system needed an FDIC.
Warren Buffett：
这其实是全国性机构的“天性”——不管是寿险公司还是银行。任何资本金要求，都无法真正保护你免于一场“真实的挤兑”（a real run）。我的意思是：如果你的负债几乎都是“当天就要兑付”的（virtually that day），或者我换句话说，几乎所有负债都要在当天兑付，那你就不可能经营一家金融机构，还能为这种情况做好准备。也正因如此，我们才需要 Fed（美联储）和 FDIC（联邦存款保险公司）。你扛不住挤兑。假设没有 FDIC 和 Fed：即使你是镇上资本最稳健的银行——比如你资本金 10%，而我的银行只有 5%——如果我雇 50 个人去你的银行门口排队站着，你也会先倒下。等他们把你弄倒了，接着就会来弄倒我——这就是为什么我们不会允许那种事情发生，因为你无法把“火”控制在隔壁那家银行。所以你扛不住挤兑，你就需要美联储和 FDIC。就算像 Northern Rock 那样，英国政府站出来说“我们担保一切”，银行门口照样排长队。人一旦害怕，就是真的害怕。而且如果你的存款没有保险，你看到你存钱的那家银行门口排队——那就去排队（笑）。必要的话，你可以从排在最前面那个人手里“买个位置”（买号）。甚至就算你是在另一家银行，你也应该去那家银行排队，把你的“疯钱”（mad money）取出来塞到床垫下面。只要没有罚金，你一周后再把钱存回去也行——那你为什么不这么做呢？正因如此我们需要 Fed 和 FDIC。我认为 FDIC 和 Social Security（社会保障）是 1930 年代带来的两项最重要的东西。这个体系需要 FDIC。
相比之下，保险浮存金有结构性的优势。
Appendix.5.《2014-10-07 Warren Buffett.Fortune MPW》
Let’s go from that to an entirely different industry — the big banks of the United States — and the question of whether they are good businesses and the question of what’s happened to them in the last few years. Are they as good a business as they were a few years ago?
我们从这里转到一个完全不同的行业——美国的大银行——来谈谈它们算不算好生意，以及过去几年它们发生了什么变化。它们还像几年前那样是门好生意吗？
No. Banks earn on assets. They don’t earn on net worth. You calculate it eventually as to what they earn on equity or net worth, but assets are the earning factors. And they’ve changed the rules so that you have to have more net worth per dollar of assets. And, obviously, if you have more net worth per dollar of assets and you’re earning a constant amount on assets, your earnings on net worth go down. They were ungodly profitable — the better ones were — back 15 years ago or 10 years ago even when they had high ratios of assets to net worth. And some of them have even cheated in terms of having even more assets than the regulators would have allowed. You had these SIVs5 as they were called — Citigroup had a whole bunch of them — so there were off balance sheet ways of controlling more assets.
不是。银行是靠资产赚钱的，不是靠净资产赚钱的。你最终当然会把它算成“赚了多少股东权益回报”或“净资产回报”，但真正产生收益的是资产。而规则已经变了：每 1 美元资产需要配更多的净资产。显然，如果每 1 美元资产对应的净资产更多，而你在资产端赚到的收益保持不变，那么你在净资产上的收益率就会下降。15 年前，甚至 10 年前，当资产与净资产的比例很高时，银行（尤其是更好的那些）利润高得离谱。而且有些银行甚至还“作弊”，让自己的资产规模超过监管原本允许的水平。那时有一种叫做 SIVs[5]]的结构——Citigroup 就搞了一大堆——这就提供了通过表外方式去控制更多资产的途径。
But all of that sort of thing has been terminated and now they’ve got much lower limits as to the assets-to-net-worth ratio. To some extent, the bigger the bank, the lower that ratio can be. So what was a very profitable business has been turned into a good business if executed well.
但这一类做法都已经被终止了，现在资产对净资产的比例上限被压得低得多。在某种程度上，银行越大，这个比例能允许的上限反而越低。所以，原本“非常赚钱”的生意，如今如果执行得好，就变成了一门“不错的生意”。
It’s a pretty simple business. You get your money cheap. Very cheap. Wells Fargo6 will have a trillion dollars roughly — or close to that — of depositors’ money and it’s probably costing around 10 basis points. Most people would think that if you can get a trillion dollars of money and pay a tenth of 1% for it, [you] would find some way to do something profitable. (Laughs)
这门生意其实很简单：你以很低的成本拿到资金。非常低。Wells Fargo[6]]大概会有一万亿美元左右——或接近这个数——的储户资金，而成本可能只有大约 10 个基点。大多数人会觉得，如果你能拿到一万亿美元的资金、只付 0.1% 的成本，[你] 总能想办法把它做成一门赚钱的生意。（笑）
But the banks have always gotten in trouble on the asset side. They’ve never gotten in trouble on the liability side, basically. They really haven’t even gotten in trouble too much on the expense side, but they go crazy occasionally on the asset side. What they do is they start copying what their dumb competitors are doing. That happens in every business, but it’s particularly virulent in the banking business. John Stumpf7 once said, “I don’t know why we keep looking for new ways to lose money when the old ones were working so well.” (Laughs) But they do — and they copycat. That’s a great danger in any business.
但银行历来出问题都出在资产端。基本上，它们从来不是在负债端出问题。甚至在费用端，它们也很少出大问题；可它们偶尔会在资产端“发疯”。它们做的事就是开始抄那些愚蠢竞争对手的做法。这种现象在每个行业都会发生，但在银行业尤其严重、尤其具有传染性。John Stumpf[7] 曾经说过：“我不知道我们为什么总在寻找新的亏钱方式，而旧的方式明明一直很好用。”（笑）但他们确实会这么干——而且会跟风模仿。这在任何行业里都是一个巨大的危险。
I warn our managers against it all the time. If anybody comes to me and says, “We want to do this because the other guy is doing it,” I say, “Go back to square one and come up with a better reason.” But human nature is such that you do want to do what others are doing.
我一直在提醒我们的经理们要防这件事。只要有人来跟我说：“我们想这么做，因为别人也在这么做。”我就会说：“回到起点重新想，找一个更好的理由。”但人性就是这样——你就是会想去做别人正在做的事。
同质化竞争，长期以来的Attention都是盯着别人，这种心理上的惯性非常麻烦。
One time, I was at a directors meeting where a leading property casualty insurance manager — a very well-known guy — was making a presentation to buy a life insurance company. He was going through all these kind of silly reasons why they should do it and he realized that the crowd was kind of catching onto the fact that his reasons weren’t too good, so finally he just threw up his hands and said, “All the other kids have one!” Basically, there’s a lot of business decisions made because all the other kids have one. (Laughs)
有一次，我在一场董事会会议上，一位顶尖的财产险（property casualty）保险经理——非常有名的人——正在做一个收购寿险公司的提案。他罗列了一堆有点可笑的理由，解释为什么他们应该这么做；他也意识到大家开始看出来这些理由并不怎么站得住脚，所以最后他干脆一摊手，说：“别的小孩都有一个！”基本上，很多商业决策就是因为“别的小孩都有一个”。（笑）
On the other hand, there are a great number of banks in the world. If you take the 50 largest banks in the world, we wouldn’t even think about probably 45 of them. Wouldn’t you say that, Charlie?
另一方面，世界上有大量的银行。如果你拿出全球前50大银行，我们可能不会考虑其中的45家。你觉得呢，查理？
Appendix.7.《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。”
Appendix.8.《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。
存贷款期限错配始终是个风险，存款端是短期的，那么贷款端最好也是短期的。
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。
Crypto bank charters
加密银行特许经营权
A theme that I have written about a lot around here is that US banking seems to be getting narrower. Traditionally, banks are in the business of (1) taking deposits and (2) making loans, and that is a famously risky combination: If all the depositors want their money back, the bank doesn’t have it, and there can be a destructive bank run. There are various ways to mitigate this problem — liquidity regulation, deposit insurance, the central bank as a lender of last resort — but it keeps popping up.
我在这里写过很多次的一个主题是，美国银行业似乎正在变得越来越狭窄。传统上，银行的业务是（1）吸收存款和（2）发放贷款，而这是一种众所周知的高风险组合：如果所有存款人都想取回他们的钱，银行却没有足够的资金，就可能发生破坏性的银行挤兑。对此问题有多种缓解方法——流动性监管、存款保险、中央银行作为最后贷款人——但这个问题仍然不断出现。
There is one theoretical way to get rid of the problem entirely: You could separate the business of taking deposits from the business of making loans. Some companies — call them “narrow banks” — could be in the business of taking deposits and just holding onto them, in literal piles of cash or more plausibly in the form of electronic reserves at the Federal Reserve. Other companies — call them “loan companies” — could be in the business of making loans, but not using deposits as funding. Those companies could raise funds from investors who knowingly take the risk of making loans and lock up their money for the long term. Deposits would be safe, loan funds would not be subject to runs, and the central risk would be removed from banking.
有一种理论上的方法可以彻底解决这个问题：你可以将吸收存款的业务与发放贷款的业务分开。一些公司——称之为“狭义银行”——可以专注于吸收存款并仅仅持有这些存款，可能是以现金堆积的形式，或者更可能是以联邦储备银行的电子准备金形式存在。其他公司——称之为“贷款公司”——可以专注于发放贷款，但不使用存款作为资金来源。这些公司可以从投资者那里筹集资金，投资者明知贷款存在风险并愿意长期锁定资金。存款将是安全的，贷款资金不会面临挤兑，银行业的核心风险将被消除。
This is a very old idea that has gotten a lot of interest from academics, but it is controversial. It is possible that the central risk of banking — taking safe deposits and using them to make risky loans — might be desirable, that separating the businesses might be bad. I often quote a 2011 Steve Randy Waldman post at Interfluidity:
这是一个非常古老的想法，受到了许多学者的关注，但也存在争议。银行业的核心风险——拿安全的存款去发放高风险贷款——可能是有益的，分开这两种业务可能是不利的。我经常引用 2011 年 Steve Randy Waldman 在 Interfluidity 上的一篇文章：
One purpose of a financial system is to ensure that we are, in general, in a high-investment dynamic rather than a low-investment stasis. In the context of an investment boom, individuals can be persuaded to take direct stakes in transparently risky projects. But absent such a boom, risk-averse individuals will rationally abstain. Each project in isolation will be deemed risky and unlikely to succeed. Savers will prefer low risk projects with modest but certain returns, like storing goods and commodities. Even taking stakes in a diversified basket of risky projects will be unattractive, unless an investor believes that many other investors will simultaneously do the same. …
金融体系的一个目的，是确保我们总体上处于高投资动态状态，而非低投资停滞状态。在投资繁荣的背景下，个人可以被说服直接参与透明且有风险的项目。但在缺乏这种繁荣的情况下，风险厌恶的个人会理性地选择回避。单个项目孤立来看会被认为风险大且不太可能成功。储蓄者会更倾向于低风险、回报适中但确定的项目，比如储存货物和商品。即使是投资于多样化的高风险项目组合，也不会有吸引力，除非投资者相信许多其他投资者会同时这样做。…
This is a core problem that finance in general and banks in particular have evolved to solve. A banking system is a superposition of fraud and genius that interposes itself between investors and entrepreneurs.
这是金融总体，尤其是银行发展起来解决的核心问题。银行系统是欺诈与天才的叠加体，它介于投资者和创业者之间。
If you tell people “put your money in the bank, where you can get it out anytime, guaranteed by the government,” (1) they will do it and (2) the bank can use the money to make risky loans to entrepreneurs and young home-buyers. If you tell people “put your money in a fund to make risky loans to entrepreneurs and young home-buyers,” they might just not do it; credit and the economy would contract. Banks take money from savers who don’t want to take risks, and use it to fund risky projects, a pro-social sleight of hand. Waldman:
如果你告诉人们“把钱存进银行，随时可以取出，且由政府担保”，(1) 他们会照做，(2) 银行可以用这些钱向创业者和年轻购房者发放高风险贷款。如果你告诉人们“把钱投入一个基金，用来向创业者和年轻购房者发放高风险贷款”，他们可能根本不会这么做；信贷和经济都会收缩。银行从不愿冒险的储户那里拿钱，用来资助高风险项目，这是一种利社会的巧妙手法。沃尔德曼说：
The analogy I would choose is finance as placebo. Financial systems are sugar pills by which we collectively embolden ourselves to bear economic risk. As with any good placebo, we must never understand that it is just a bit of sugar.
我选择的比喻是金融如同安慰剂。金融系统是我们集体鼓励自己承担经济风险的“糖丸”。就像任何好的安慰剂一样，我们绝不能明白它只是点糖。
That’s what I thought a few years ago. It also seems to be roughly what the US Federal Reserve thinks. In 2018, an entity called TNB USA Inc. tried to open a narrow bank. It applied for an account at the Fed, where it would just park its depositors’ money. It wouldn’t make any loans or investments; all the money would just sit at the Fed, which pays interest on reserves. The Fed dragged its feet on opening an account for TNB, and explained its reasoning 2019, arguing that “narrowly focused depository institutions” like TNB, which “hold a very large proportion of their assets in the form of balances at Reserve Banks,” could “have the potential to complicate the implementation of monetary policy” and “disrupt financial intermediation in ways that are hard to anticipate.”
几年前我也是这么想的。美国联邦储备委员会似乎也大致持相同看法。2018 年，一个名为 TNB USA Inc.的实体试图开设一家狭义银行。它申请在联储开设账户，只是将存款人的资金存放在那里。它不会发放任何贷款或进行投资；所有资金都将存放在联储，联储会对准备金支付利息。联储在为 TNB 开设账户的问题上拖延，并在 2019 年解释了其理由，认为“像 TNB 这样的‘狭义存款机构’，其‘大部分资产以联储银行余额的形式持有’，可能‘会对货币政策的实施造成复杂影响’并且‘以难以预料的方式扰乱金融中介’。”
Again, in 2018 and 2019, this struck me as largely correct: If everyone could have totally safe, Fed-backed, interest-paying deposits, surely there would be less demand for bank deposits, fewer loans and less financial intermediation.
同样，在 2018 年和 2019 年，我认为这在很大程度上是正确的：如果每个人都能拥有完全安全、由联储支持且支付利息的存款，肯定会减少对银行存款的需求，贷款减少，金融中介也会减少。
金融中介的冒险精神有非常重要的价值。
But since then it does seem that banking has gotten narrower. On the deposit side, you kind of can get totally safe deposits. Treasury money market funds don’t quite park their cash at the Fed and earn interest, but they get pretty close, investing in short-term Treasury securities and reverse repos at the Fed. If you don’t like the risk of depositing money at a bank — and after the Silicon Valley Bank run in 2023, many people don’t — you can get much closer to narrow banking via money market funds. And on the lending side, the enormous boom in private credit proves that in fact there is a ton of demand from investors who want to knowingly take credit risk. You can get lots of loans — risky leveraged buyout loans, but also consumer loans and mortgages — without the placebo of banking.
但从那时起，银行业务似乎变得更狭窄了。在存款方面，你可以获得完全安全的存款。国债货币市场基金虽然不能将现金完全存放在美联储并赚取利息，但它们非常接近，通过投资短期国债证券和美联储的逆回购操作实现。如果你不喜欢将钱存入银行的风险——在 2023 年硅谷银行挤兑事件之后，很多人都不喜欢——你可以通过货币市场基金更接近狭义银行业务。在贷款方面，私人信贷的巨大繁荣证明，实际上有大量投资者愿意有意识地承担信用风险。你可以获得大量贷款——风险较高的杠杆收购贷款，也有消费贷款和抵押贷款——而无需银行的“安慰剂”。
So banking is getting narrower, though it is not yet narrow in any absolute sense. Private credit is still much smaller than the banking system, and private credit firms themselves often borrow from banks to increase their ability to make loans. The Fed’s 2019 objections to narrow banking still apply; the Fed formally turned down TNB’s account application only last year.
因此，银行业务确实变得更狭窄了，尽管在任何绝对意义上还不算狭窄。私人信贷仍远小于银行系统，私人信贷公司本身也经常从银行借款以增强其放贷能力。美联储 2019 年对狭义银行业务的反对意见仍然适用；美联储直到去年才正式拒绝了 TNB 的账户申请。
But narrow banking pops up elsewhere. One important modern form is stablecoins. A stablecoin is a crypto form of banking: You deposit dollars with a stablecoin issuer, it gives you back tokens entitling you to get your dollars back, and meanwhile it does whatever it wants with the dollars. In the unregulated early days of crypto, “whatever it wants” could be quite spicy indeed, but these days stablecoins are a big business and there is something of a norm of parking the deposits in very safe short-term dollar-denominated assets, ideally Treasury bills or reverse repos or a BlackRock money market fund. If you launched a new stablecoin today and said “we will take your dollars and use them to make loans to emerging crypto entrepreneurs,” you’d have a hard time competing with the big incumbent stablecoins that say “we will take your dollars and use them to buy Treasury bills.” (Especially if, like most stablecoins, you didn’t pay interest.)
但狭义银行业务在其他地方也出现。一种重要的现代形式是稳定币。稳定币是一种加密形式的银行业务：你将美元存入稳定币发行方，他们会给你返还代币，赋予你取回美元的权利，同时他们会用这些美元做任何他们想做的事情。在加密货币早期不受监管的时代，“任何他们想做的事情”可能相当激进，但如今稳定币已成为一项大业务，且普遍的做法是将存款停放在非常安全的短期美元计价资产中，理想情况下是国库券、逆回购或贝莱德货币市场基金。如果你今天推出一个新的稳定币，并说“我们将拿你的美元去向新兴的加密创业者发放贷款”，你将很难与那些说“我们将拿你的美元去购买国库券”的大型现有稳定币竞争。（尤其是如果你像大多数稳定币一样不支付利息的话。）
One flavor of this is “central bank digital currency,” which means roughly “a stablecoin run by the Fed and backed directly by Fed deposits.” And my impression is that the Fed does not like the idea of a CBDC, in part for these sorts of financial-intermediation concerns. More generally, my impression is that the Fed does not want a world in which stablecoins are (1) a dominant form of digital money and (2) narrow banking. That is a world in which the alchemy of banking is harder to achieve.
其中一种形式是“中央银行数字货币”，大致意思是“由美联储运营并直接由美联储存款支持的稳定币”。我的印象是，美联储不喜欢 CBDC 的想法，部分原因是出于对金融中介的担忧。更广泛地说，我的印象是，美联储不希望出现这样一个世界：稳定币成为（1）主导的数字货币形式，且（2）是狭义银行业务。那将是一个银行炼金术更难实现的世界。
On the other hand, do you want a world in which stablecoins are (1) a dominant form of digital money and (2) not narrow banking? Broad banking? Where the stablecoin issuers do take dollars and use them to make loans to emerging crypto entrepreneurs? What about a world where the stablecoin issuers take dollars and use them to make, you know, leveraged buyout loans, or consumer loans, or home mortgages? Maybe that’s fine? Maybe that’s just banking? Maybe we can just recreate the whole banking system but in crypto?
另一方面，你想要一个稳定币成为（1）主导的数字货币形式且（2）不是狭义银行业务的世界吗？广义银行业务？在这个世界里，稳定币发行者会拿美元去向新兴的加密创业者发放贷款？那如果稳定币发行者拿美元去发杠杆收购贷款、消费贷款或住房抵押贷款呢？也许这没问题？也许这就是银行业务？也许我们可以在加密领域重新创造整个银行系统？
Anyway the Wall Street Journal reports:
无论如何，《华尔街日报》报道：
A host of crypto firms including Circle and BitGo plan to apply for bank charters or licenses, according to people familiar with the matter. Crypto exchange Coinbase Global and stablecoin company Paxos are considering similar moves, other people said.
据知情人士透露，包括 Circle 和 BitGo 在内的一批加密公司计划申请银行特许或许可证。其他人士表示，加密交易所 Coinbase Global 和稳定币公司 Paxos 也在考虑类似的举措。
That comes as the Trump administration moves to incorporate crypto into mainstream finance and Congress advances a pair of bills that would establish a regulatory framework for stablecoins, which let people easily trade in and out of more volatile cryptocurrencies. The legislation would require stablecoin issuers to have charters or licenses from regulators.
这正值特朗普政府推动将加密货币纳入主流金融体系之际，国会也在推进两项法案，旨在为稳定币建立监管框架，稳定币使人们能够轻松地在更为波动的加密货币之间进行交易。该立法将要求稳定币发行商获得监管机构颁发的特许或许可证。
Some crypto firms are interested in national trust or industrial bank charters that would enable them to operate more like traditional lenders, such as by taking deposits and making loans. Others are after relatively narrow licenses that would allow them to issue a stablecoin. …
一些加密公司对国家信托或工业银行特许感兴趣，这将使它们能够更像传统贷款机构一样运营，例如接受存款和发放贷款。其他公司则寻求相对狭窄的许可证，以允许它们发行稳定币。…
Any crypto firm that obtains a bank charter would become subject to stricter regulatory oversight.
任何获得银行特许的加密公司都将受到更严格的监管监督。
The obvious story here is that any crypto firm that obtains a bank charter would be subject to stricter oversight of its know-your-customer and anti-money-laundering programs, its capital, its information security, etc., all the stuff that crypto firms sometimes get in trouble for.
显而易见的故事是，任何获得银行执照的加密公司都将受到更严格的客户身份识别和反洗钱程序、资本、信息安全等方面的监管，所有这些都是加密公司有时会遇到麻烦的地方。
But the weirder story here is: If you are a crypto firm that wants a bank charter, do you have to do lending? If you are a stablecoin issuer whose entire business is (1) taking deposits and (2) parking those deposits in Treasury bills, do bank regulators want to give you a charter? Aren’t bank charters for banks that do financial intermediation? Will there be pressure for stablecoin issuers to become real banks, to take deposits and use them to fund loans? Will crypto companies that want to become banks have to be regular banks, not narrow ones?
但更奇怪的故事是：如果你是一家想要银行执照的加密公司，你必须进行贷款业务吗？如果你是一家稳定币发行商，其整个业务是（1）接受存款和（2）将这些存款投资于国债，银行监管机构会愿意给你发放执照吗？银行执照难道不是给那些从事金融中介业务的银行的吗？是否会有压力促使稳定币发行商成为真正的银行，接受存款并用这些存款发放贷款？想成为银行的加密公司是否必须成为普通银行，而不是狭义银行？
This is a little funny because crypto is to some extent a reaction against traditional banking. Satoshi Nakamoto invented Bitcoin in part in disgust at the 2008 financial crisis. And cryptocurrencies are distinct from dollars in that a dollar fundamentally is a debt obligation of a bank, while a Bitcoin fundamentally isn’t; crypto is a form of non-debt money, a way to have money and payments and bank accounts without traditional banking. I wrote in 2023 that crypto “to some extent started as a backlash to fractional reserve banking and the shadow banking crisis of 2008, and by last year had matured to the point that it recreated both fractional reserve banking (but without regulation!) and a 2008-style shadow banking crisis.” But that is just the beginning of the story. Perhaps the future is that crypto will fully recreate normal banking, including bank regulation.
这有点好笑，因为加密货币在某种程度上是对传统银行业的反应。中本聪发明比特币，部分原因是对 2008 年金融危机的反感。加密货币与美元不同，美元本质上是银行的债务义务，而比特币本质上不是；加密货币是一种非债务货币，是一种无需传统银行即可拥有货币、支付和银行账户的方式。我在 2023 年写道，加密货币“在某种程度上起初是对 2008 年部分准备金银行和影子银行危机的反弹，到去年已经成熟到重新创造了部分准备金银行（但没有监管！）和 2008 年式的影子银行危机。”但这只是故事的开始。也许未来加密货币将完全重塑正常银行体系，包括银行监管。
加密货币技术上还存在技术上的障碍，更有可能成为小众的方案。
Appendix.10.《2025-10-09 Matt Levine.Crypto narrow banking》
Crypto narrow banking
加密货币的“窄式银行业”
A theme around here is that the funding model of banking, though common, is risky. Banks take money from depositors who can ask for it back at any time, and they use that money to make long-term loans. There is a duration mismatch: If the depositors all do ask for their money back at once, the bank won’t have it, and disaster will ensue. This is a well-understood problem, and there are important mechanisms — central banks as lenders of last resort, deposit insurance, liquidity and capital regulation, etc. — to mitigate it.
这里的一条主线是：银行业的融资模式虽常见，但很危险。银行吸收可随时支取的存款，再用这些钱发放长期贷款，形成期限错配：若存款人同时来取钱，银行拿不出来，灾难就会发生。这个问题众所周知，体系内也有多种重要缓释机制——央行充当最后贷款人、存款保险、流动性与资本监管，等等。
But even so, in recent years, we have seen a bit of a retreat from this traditional “maturity transformation” in banking, and a bit of a move toward narrow banking. Some depositors don’t put their money in traditional banks, but instead put it in “narrow banks” (like money-market funds and stablecoins) that park it in safe short-dated government-backed instruments (Fed reserves, Treasury bills). Meanwhile risky long-term loans are increasingly made by private credit funds, which get a lot of their money from, paradigmatically, life insurance companies. The idea is that life insurance companies — unlike banks — have long-term, stable funding: They get money from customers and give it back in the form of annuities (with fixed payments over some term) or death benefits (hopefully many years later). So life insurers are natural sources of long-term loans.
即便如此，近几年我们看到传统的“期限转换”有所收缩，向“窄式银行业”略有转向。有些存款人不再把钱放在传统银行，而是放在“窄式银行”（如货币市场基金和stablecoins稳定币），这些机构把资金停放在安全的、短久期、政府背书的工具上（如Fed准备金、Treasury bills）。与此同时，风险更高的长期贷款日益由私募信贷基金提供，而这类基金的大量资金来源于典型的机构——人寿保险公司。逻辑在于，人寿保险公司的资金期限长且稳定，与银行不同：它们从客户处收取保费，未来以annuities（按一定期限固定支付的年金）或身故给付的形式返还（理想情况下在多年之后）。因此，寿险公司是长期贷款的“天然供给者”。
事情一直在向前推进，寿险不太了解，巴菲特的评论中认为寿险资金有提前支取的风险。
Crypto relearns a lot of the lessons of traditional finance, including this one. In 2022 there was a crypto banking crisis: There were a lot of crypto shadow banks (lending platforms and exchanges) that took crypto from depositors who could ask for it back at any time and used it to make often quite long-term loans. The loans went bad, the depositors asked for their crypto back, and various crypto platforms — Celsius, Voyager, Genesis, FTX — blew up. This is a well-understood problem, but people conveniently forgot it during the boom, and crypto had no mechanisms — no lender of last resort, no deposit insurance, no particular discipline about capital or liquidity — to mitigate it.
加密行业在不断重学传统金融的诸多教训，这一点也不例外。2022年出现了一场“加密银行业”危机：大量加密“影子银行”（借贷平台与交易所）吸收可随时赎回的加密资产，再将其投向期限往往相当长的贷款。贷款坏掉后，存款人要赎回，加密平台——Celsius、Voyager、Genesis、FTX等——相继爆雷。这个问题本就人尽皆知，但在繁荣期被选择性遗忘；加密领域也缺乏相应机制——没有最后贷款人、没有存款保险、在资本与流动性上也缺乏纪律——来加以缓冲。
The obvious solution is that crypto life insurers should make the crypto loans? Bloomberg’s Muyao Shen reports:
显而易见的解决方案是：让“加密寿险公司”来做加密贷款？据Bloomberg记者Muyao Shen报道：
Apollo, Northwestern Mutual, Pantera Capital and Stillmark are joining Bain Capital and crypto investor Haun Ventures to back the Bitcoin life insurance firm Meanwhile in a $82 million funding round.
Apollo、Northwestern Mutual、Pantera Capital与Stillmark，携手Bain Capital和加密投资机构Haun Ventures，参与对Bitcoin寿险公司Meanwhile的8200万美元融资。
The Bermuda-regulated insurer, which claims to be the first life insurer to offer products entirely denominated in crypto, began to provide policies in 2023. The firm invests policyholders’ premiums by lending Bitcoin to large, regulated financial institutions. Zac Townsend, a co-founder and chief executive officer of Meanwhile, said the firm is now “one of the largest lenders of Bitcoin in the world at duration,” or over longer periods of time.
这家受百慕大监管的保险公司自称是首家完全以加密货币计价提供产品的人寿险公司，已于2023年开始承保。公司通过将Bitcoin出借给大型、受监管的金融机构来投资保单持有人的保费。Meanwhile联合创始人兼首席执行官Zac Townsend表示，公司如今已是“全球在长期久期维度上最大的Bitcoin出借方之一”，也就是以更长期限开展放贷。
“We are not running a hedge fund or trading desk or worried about the price of Bitcoin day-to day or week-to-week or month-to-month,” Townsend said in an interview. “We engage on this side of the business in institutional B2B, private credit.”
Townsend在采访中表示：“我们不是在运营对冲基金或交易台，也不为Bitcoin的日、周、月度价格波动操心。我们的这一侧业务定位于面向机构的B2B、private credit（私募信贷）。”
I mean. Yes. Also there’s a tax trade:
嗯，是的。另有一笔税务上的“安排”：
Meanwhile’s life insurance products also provide tax advantages for Bitcoin holders. After two years, policyholders can borrow up to 90% of their policy’s Bitcoin value in a tax-free manner, according to the firm’s website. The borrowed Bitcoin will adopt a new cost basis, allowing the policyholders to sell the Bitcoin without triggering capital gains taxes.
Meanwhile的人寿保险产品还为Bitcoin持有人提供税收优势。根据公司网站，满两年后，保单持有人可在免税的情况下借出其保单Bitcoin价值的最高90%。被借出的Bitcoin将采用新的cost basis（计税成本基础），从而使保单持有人在出售这些Bitcoin时不触发资本利得税。
热门主题
Recent Articles
2025-05-03 Berkshire Hathaway 股东大会：30条批注
以下是 2025 年 Berkshire Hathaway 股东大会的30条批注，记录了一些值得反复推敲的东西。 1. 日本投资：用事实看文化接受度 【3.Interest rates in Japan】 Warren Buffett: "Japan's record has been extraordinary. My guess is that Tim Cook would tell you that iPhone sales there are about as great as any ...2025-05-03 Warren Buffett.Marmon
Refer To：《2025-05-03 Berkshire Hathaway Annual Meeting》。 7.Portillo’s Hot Dogs Portillo的热狗 [00:49:50] Sean Siegel: Hello, I’m Sean Siegel from Chicago, IL. Thank you for investing your time, and to the executive committee for putting on this meeting, ...1936 John Maynard Keynes.The State of Long-Term Expectation
Refer To：《1936 John Maynard Keynes.The State of Long-Term Expectation》。 I We have seen in the previous chapter that the scale of investment depends on the relation between the rate of interest and the schedule of the marginal efficiency of capital ...1991-03-01 Warren Buffett.Wells Fargo
Refer To：《1991-03-01 Warren Buffett's Letters to Berkshire Shareholders》。 Lethargy bordering on sloth remains the cornerstone of our investment style: This year we neither bought nor sold a share of five of our six major holdings. The exception was ...1991-03-01 Warren Buffett.Measuring Insurance Performance
Refer To：《1991-03-01 Warren Buffett's Letters to Berkshire Shareholders》。 Measuring Insurance Performance 衡量保险业绩 In the previous section I mentioned “float,” the funds of others that insurers, in the conduct of their business, temporarily hold. ...