(Bloomberg Opinion) -- Narrow banks.

The basic business of banks is to take deposits and make loans. The deposits have to be “risk-free”: A deposit in a bank account ought to be just “money”; depositors shouldn’t have to worry about it. The loans have to be risky: Lending money to people to start businesses or buy houses necessarily comes with the risk that they won’t pay you back. This is the magic, and the problem, of banking—that banks take risky loans and turn them into risk-free deposits—and there are lots of fraught imperfect ways to make it work. The banks diversify their loans and take collateral and monitor their borrowers; they have capital buffers to make sure they’ll still have money even if some loans go bad. Regulators monitor banks’ lending and liquidity and capital; central banks and deposit insurers backstop the banks. It mostly works most of the time, but it is controversial and conceptually imperfect. No private business can really turn risky loans into entirely risk-free deposits. The government (or the central bank) mostly can, at least if it prints its own currency, but using government resources to backstop lucrative private businesses has its own issues.

And so you sometimes see proposals to scrap this basic business of banking and replace it with something else. The something else is generally that banks should take risk-free deposits and invest them in really risk-free things (reserves at the central bank, or perhaps their own government’s securities), while someone else (mutual funds, lending corporations, whatever you want to call them) would be in the business of taking risky investments—from investors who want yield and don’t need money-like deposits—and using them to make loans. These ideas are often called “narrow banking,” or sometimes the “Chicago plan,” but there are other, milder flavors. We talked recently about a proposal, from Morgan Ricks, John Crawford and Lev Menand, in which the U.S. Federal Reserve would offer “FedAccounts” to anyone who wanted one. FedAccounts would be a flavor of optional narrow banking: If you want pure “money,” entirely safe deposits, you could just open a bank account at the Fed that would be entirely invested in (in fact, would consist directly of) reserves at the Fed. The Fed, in this proposal, would pay you the same interest rate that commercial banks receive on their Fed reserves, currently 1.95 percent, and also let you write checks and use ATMs and stuff. If you want a higher return, you could put your money in a money-market account or a bond mutual fund or, I suppose, a high-interest savings account at a regular bank, but if you just want a safe electronic money account the FedAccount would be fine. This proposal wouldn’t require narrow banking; regular banks could still offer deposit accounts and make loans and so forth. But it would certainly push in that direction, as the regular banks’ deposit accounts would be a lot less attractive with competition from the Fed, and they’d have to raise more of their money from equity and bonds and other non-deposit-like sources.

But that proposal is pretty radical, not only because it might upend the business model of existing banks, but also because the Fed would have to, like, open accounts and do customer service and issue ATM cards and do a lot of other administrative stuff to offer bank accounts to everyone.

But someone came up with a much simpler and amazing solution. It’s this:

  1. Start a bank;
  2. Take deposits;
  3. Invest 100 percent of those deposits in reserves at the Fed; and
  4. Pass the interest on to your depositors.

It is called TNB USA Inc. (for “The Narrow Bank”), it is run by the former head of research at the New York Fed, and it is simultaneously a dumb simple one-sentence idea and the most interesting bit of financial engineering that I’ve seen this year. Of course it is a very 2018 form of financial engineering, which is to say almost the opposite of what financial engineering looked like in, say, 2006. The traditional form of financial engineering is complex and alchemical, taking lots of dodgy ingredients and stirring them up into something more attractive. This is financial engineering by dynamite blast: You just get rid of all the apparatus of banking and blow a tunnel directly from savers to the ultimate backstop of the banking system.

Well, not quite directly; TNB wants to offer its accounts only to “the most financially secure institutions,” meaning apparently money-market funds and foreign central banks, and it does not plan to “provide any retail banking services for individuals.” It is a pure interest-rate arbitrage for money-market funds: Instead of effectively depositing money with the Fed through its reverse repo facility at 1.75 percent, funds could deposit their money with TNB and get the Fed’s higher rate on reserves, 1.95 percent. Presumably TNB would take a cut, but its expenses would be small: It would have only a relatively small number of very large customers, so there’s not a lot of administrative overhead, and it’s not doing any lending or investing work.

It’s fine, you know, it’s a trade, but if it works the really interesting extension is to do it for retail. The overhead would be higher—you’d need at least a website, a customer service department, ATM cards—but the opportunity is intriguing. You could offer retail depositors a checking account at a bank that is immune to financial crises, runs and bad investments, a checking account that is fully backed by the Fed.

Also, weirdly, you might be able to pay them more interest than regular banks offer. “The FDIC recently reported that jumbo deposits—$100,000 or more—on average earned eight basis points (0.08 percent) from savings accounts, and five basis points (0.05 percent) from checking accounts, while the Federal Reserve Banks pay 195 basis points (1.95 percent) as the IOER rate,” notes TNB. If banks are mostly investing in things that are riskier (and so higher-return) than Fed deposits, they should be able to pay interest that is higher than the rate on Fed deposits. But of course if you’re a regular bank doing that, you have to hire a lot of bankers to evaluate those loans and investments, and risk managers to check up on them, and some of the investments will go bad and you’ll lose money, and you probably have branches, and the whole thing is just big and expensive. The TNB model is much leaner, and it is quite possible that it could pay a higher rate.

But so far it doesn’t work. TNB is in the news because it got a provisional banking charter in Connecticut, set itself up as a bank, and went to the Fed asking to open a reserve account, but the Fed said no. So TNB is suing the Fed, arguing that the Fed’s rules require it to open an account for any qualified bank, and that it is a qualified bank. Here is TNB’s complaint; here are an analysis from John Cochrane and an article about it in the Wall Street Journal. From the complaint it seems like TNB is mystified about why the Fed said no, but it apparently goes all the way to the top: The New York Fed declined the account “reportedly at the specific direction of the Board’s Chairman,” Jerome Powell. The Fed has not said what its objection is.

Cochrane has a few theories: “The Fed may worry about controlling the size of its balance sheet -- how many reserves banks have at the Fed, and how many treasuries the Fed correspondingly buys,” or it may worry “that narrow banks are a threat to financial stability, not a guarantor of it as I have described, because people will run to narrow banks away from repo and other short term financing in times of stress.” Or, more darkly, the explanation may be “that the Fed, which is a banker's bank, protects the profits of the big banks system against competition.”

I think those are good explanations, but the way I would put it is: People have been proposing “narrow banking,” in some form or another, to the Fed for many years. Since the financial crisis, these proposals have become more popular, more detailed and more forceful. And the Fed has not decided to scrap the traditional banking system, to tell the big banks to stop taking deposits to make loans and instead do something radically different. I think there are good economic reasons that the Fed would not want to do that (principally, there is surely some social benefit to transforming money from savers seeking risk-free return into risk capital for productive investments), and there are certainly good institutional path-dependency reasons: If you were designing a banking system from a blank slate, you might not want an entity with the particular size and mix of businesses of JPMorgan Chase & Co., but given that JPMorgan exists you want to be very careful about accidentally setting it on fire.

Given that reluctance to completely scrap the banking system and start over, people have proposed more incremental, optional steps: FedAccounts, or now TNB. In these plans, nothing changes in bank regulation; banks can keep doing exactly what they’re doing now. But their competitive landscape changes: Instead of just competing with other banks, or other non-government-backed sources of money-like claims (money-market funds, etc.), the banks would now have to compete with interest-bearing accounts issued directly (or, with TNB, somewhat indirectly) by the Fed. But that is tough competition! If you can get paid 1.95 percent on perfectly safe deposits, getting 0.08 percent on imperfectly safe deposits is not very appealing. Narrow banking by competition might end up in the same place as narrow banking by regulation: If banks can no longer attract deposits—or, with TNB, if they can no longer attract short-term funding from money-market funds who go to TNB instead—then they will effectively be forced to transform into something else. (Cochrane uses the term “equity-financed banks”: They would have the assets of traditional giant banks—loans, securities, etc.—but their liabilities wouldn’t include anything deposit-like; they’d just be funded by equity and long-term debt.) If you think that that would be a bad thing—if you think that pulling deposits from big banks would precipitate a crisis, or at least reduce lending—then you won’t want to do it accidentally. You won’t want to open a Fed account for a tiny weird bank whose business model might bring the whole system down. No matter how interesting it is.

Credit ratings.

After the financial crisis, the big credit-rating firms—mainly S&P Global Inc. and Moody’s Investor Service—were criticized ( and sued) for being too generous in rating structured bonds. The obvious problem was that banks paid a lot of money to the agencies for structured-product ratings, and the ratings were a competitive business, and if your ratings weren’t high enough then the banks would just go to another agency and you’d lose business. The “issuer-pays” model created incentives for the raters to be generous. A less obvious but also very important problem was that investors wanted high ratings too: Many investors wanted to buy high-yielding (and, thus, risky) securities while telling their boards or regulators or customers that everything they bought was AAA; by giving out AAA ratings generously the agencies were making investors happy. (Of course ultimate investors, regulators, etc. were arguably deceived by this, but the agencies were hearing directly from the banks and investors, not those more-removed stakeholders.)

The only real constraint on the ratings—besides I guess fear of eventual lawsuits, which happened—was the agencies’ own internal commitment to the sacred eternal meaning of the AAA rating. But that commitment was … kinda weak. For one thing there was money on the line. But also there is no sacred eternal meaning of a AAA rating; it is just a series of letters. And in fact the same rating corresponds to different default rates in different asset classes: Municipal and sovereign bonds seem to be rated more harshly than corporate bonds, while collateralized debt obligations seem to have been rated more generously. An “AAA” rating arguably conveys not so much a specific numerical probability of default as it does “risk-free enough to be considered risk-free in this particular market.”

Anyway this is great:

S&P Global Inc. is developing a custom credit-rating scale for China that will likely mean more triple-A’s, worrying investors about inflated grades.

The credit-rating company, which is setting up an independent business in China, recently said it will tailor a system for rating bonds from businesses, local governments and other issuers there.

Why would it do that? “We believe that considering the size, dimensions and extent of diversification of China’s domestic capital market, there needs to be a set of special rating standards and rating methodology that fit the local situation,” says S&P, meaninglessly, but the straightforward answer is competition:

In China’s $11 trillion bond market—which has been open only to domestic credit-rating companies—roughly 17% of bonds are triple-A and 76% double-A, according to Wind Info.

And investor demand:

There are also practical reasons for an inflated scale in China, where many financial institutions, such as banks and insurance companies, as well as some asset managers, are restricted from buying bonds below a certain rating level.

If you come into a market to rate its bonds, and there are already established conventions about what the ratings mean, and those conventions differ from the conventions that you follow in other markets, what do you do? You can stick up for the sanctity of your conventions, and say that AAA means AAA and you’ll give this Chinese bond a BBB even though its local raters give it an AAA. But, one, this will lose you business: What AAA issuer will pay you for a BBB rating? And, two, it doesn’t really make sense. They are just letters, just conventions; there is no objective eternal meaning of a AAA rating. If the Chinese convention is that a bond that would be BBB in the U.S. is AAA in China, then calling it BBB would be confusing or misleading to Chinese investors. Of course calling it AAA would be confusing or misleading to global investors. The—reasonable, marketing-friendly, but awkward—compromise is to call it AAA with an asterisk, to adopt a China-specific convention and explain in a footnote that it is not comparable to your other conventions. 


There are apparently a lot of people who really want to day-trade stocks or currencies or cryptocurrencies with enormous leverage, and a lot of unscrupulous (legal, semi-legal or illegal) online brokerages that will let them do it to their heart’s content. Those people will mostly lose all their money, though I guess they’ll have a good time doing it. But it seems a little wasteful, if you are one of those unscrupulous brokerages, to take those people’s money and lose it for them. Sure you’ll end up pocketing a lot of it, in the form of commissions and fees on all the trading they do, but a lot of it will just be frittered away into the market and end up in the hands of smarter traders. From your perspective, it is a leakage, an inefficiency. Your dumb client comes to you with money and asks you, in effect, to help him lose it. You can do that; that is your specialty. But why not just take it instead?

Here is a Securities and Exchange Commission enforcement action charging “two Michigan men with fraud for their roles in a fake accounts scheme perpetrated by a phony day-trading firm, Nonko Trading.” But they weren’t completely phony. The SEC alleges that they “both had extensive experience in day-trading operations.” They had enough familiarity with, and contempt for, their customer base to know that the customers were just going to lose all their money, so they allegedly took it from them first:

To attract day-traders, Chamroonrat, through Nonko, offered terms that were not available at any SEC-registered broker-dealer in the United States, including low trading commissions (typically at or below $0.006 per share), a minimum deposit of only $2,500 (and occasionally lower), as well as leverage (or margin) of 20:1 (that is, purporting to give traders the ability to trade $20 of total capital for each dollar deposited). Such low account balances and high leverage ratios are prohibited for many day traders in the United States under FINRA’s rules.

Observing that many of Nonko’s day-trading customers often lost money in the market, the Nonko team decided to take advantage of this pattern by secretly providing some of Nonko’s customers with training accounts instead of live ones and simply pocketing those customers’ deposits. To minimize the risk of detection, Goldman, Eikenberry, Chamroonrat, and Chamroonrat’s associates Avnon, Armon, and later Plumer targeted traders who appeared inexperienced or unsophisticated, or had a history of trading losses. The Nonko team reasoned that such traders were likely to place more losing trades and thus unlikely to ask to withdraw funds from their accounts.

It’s basically “The Producers,” but for day-trading: No one will notice that you stole their money if they were going to lose it anyway. I … look. If the SEC’s allegations are true, this is illegal. It is bad. It’s definite willful fraud. You shouldn’t do it. But in my line of work, I often read about people making incredibly dumb trading and investing decisions, and I constantly think “man, you should have just sent me all your money instead, I would have made better use of it.” Not that I’d have invested it better for you, I mean, just that I’d have spent it on myself, and that would have been a better use of your money than losing it in whatever crypto scheme you got caught up in. So I find myself sympathizing with these guys. They knew that their customers were just going to lose their money, so they allegedly decided to spend it on themselves instead.

Things happen.

Justice Department Nearing Antitrust Approval of Health Mergers Combining CVS-Aetna, Cigna-Express Scripts. CBS in Talks With Redstone Family’s National Amusements To Settle Litigation. FBI Investigating American Express Foreign-Exchange Pricing. Bitcoin Falls Off a Cliff Again as Cryptocurrency Slump Deepens. Tesla's $1.8 Billion of Junk Bonds Hit Lowest Since 2017 Sale. The story of a house: how private equity swooped in after the subprime crisis. How to Make Bankers Try Harder to Avoid Going Bust. Near Everest’s Slopes, a Helicopter Rescue Fraud Preys on Trekkers. Bitcoin Bodice Rippers Are Romance’s Most Sizzling New Subgenre. Gwyneth Paltrow's Goop Pays $145,000 Over Jade Vaginal-Egg Claims. 

If you'd like to get Money Stuff in handy email form, right in your inbox, please subscribe at this link. Thanks!

Like Money Stuff? Subscribe to Bloomberg All Access and get much, much more. You'll receive our unmatched global news coverage and two in-depth daily newsletters, The Bloomberg Open and The Bloomberg Close.

To contact the author of this story: Matt Levine at mlevine51@bloomberg.net

To contact the editor responsible for this story: James Greiff at jgreiff@bloomberg.net

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

Matt Levine is a Bloomberg Opinion columnist covering finance. He was an editor of Dealbreaker, an investment banker at Goldman Sachs, a mergers and acquisitions lawyer at Wachtell, Lipton, Rosen & Katz, and a clerk for the U.S. Court of Appeals for the 3rd Circuit.

©2018 Bloomberg L.P.