As we’ve seen over the past few weeks, banks don’t behave like other companies. Analysing them can make even professionals queasy. I’ve been following them closely for over 25 years, as an analyst and investor. Here’s what I know:

1. Customer or Creditor?

The first thing to understand about banks is that they operate a unique financial structure. Other companies borrow from one group of stakeholders and provide services to another. For banks, these stakeholders are one and the same: their creditors are their customers.

Most customers don’t realise this and as long as their loan to the bank (traditionally called a deposit) falls within the scope of their national deposit insurance scheme, they don’t have to. As US depositors have been reminded these past few weeks, the first $250,000 they lend to their bank is fully insured. The bank could close down on Friday and their loan would be available for immediate redemption on Monday thanks to a backstop provided by the Federal Deposit Insurance Corporation (FDIC). Anything over $250,000, though, and they become an unsecured creditor (at least in theory).

Banks in some countries take this further, exploiting a captive pool of naive lenders to borrow on even looser terms. Up until around 2015, Italian banks would regularly sell subordinated bonds to their customers, offering them a more attractive rate than they could get on deposit. By October 2015, over half of the circulating bonds issued by banks were held by retail customers. This practice has now been reined in, but uninsured deposit issuance is still rife. 

For an analyst, this structure complicates things. No use looking at the ‘enterprise value’ of the business – the traditional barometer for measuring a business’ worth – because the debt is the business. No use looking at cash flow statements which disentangle cash generated in operating activities from cash generated in investing and financing activities because the operating activities are the financing activities.¹

Because it’s where the customers sit, if you want to understand a bank, you have to look at its balance sheet. As a bank analyst, I would often peer over in envy as colleagues covering other sectors would go on factory tours or visit flagship retail outlets. I’d stay at home and study bank balance sheets. 

To some extent the balance sheet is the business. The rest is there simply to feed it – the store a device to attract new funding. Bank executives may not want to admit it because it seems overly reductive, but those that forget do so at their peril.

Metro Bank is a case in point, discussed here a few months ago. Back in his home state of New Jersey, its founder, Vernon W. Hill II, “refined a business model patterned after the great retailers of America, i.e., Home Depot, WalMart, McDonald’s, Starbucks, rather than the typical bank or financial institution.”

On arriving in the UK, Hill designed Metro Bank “as a retail operation more akin to Apple than to the Bank of England.” After peaking at a market capitalisation of £3.5 billion in 2018, the company now flounders at a valuation of around £180 million. The stores turned out fine, the balance sheet not so much.²

In death, the balance sheet is all that’s left of a bank. When First Citizens Bank took over the remains of Silicon Valley Bank this week, it opened the casket to investors. Lying there was $72.1 billion of loans and $35.3 billion of cash, funded by $56.5 billion of deposits. Gone was all the talk about Silicon Valley Bank as “a networking tool” for its customers, all the talk about Silicon Valley Bank as a business. 

[More on the economics of this transaction for paid subscribers below]

2. Public or Private?

Banks have a licence to create money which confers on them a special status somewhere between private enterprise and public entity. Economists argue that commercial banks create money by making new loans. When a bank makes a loan, it credits the borrower’s bank account with a deposit the size of the loan. At that moment, new money is created. 

Bank analysts don’t quite see the world like this. In their view, banks need deposits in order to make loans. ‘Deposits before loans’ is a more useful model for an individual bank. Either way, banks operate a two-sided platform and maintaining an equilibrium is important. Banks struggle when one side of the ledger grows more quickly than the other. Silicon Valley Bank grew deposits more quickly than loans; Northern Rock grew loans more quickly than deposits. 

The privilege of creating money comes with responsibility. To ensure banks behave responsibly, they are heavily regulated. Getting a licence at all can be a challenge. Neobank Revolut has been trying to obtain one in the UK for several years to no avail; likewise The Narrow Bank in the US. Depending on the particular jurisdiction, a licence can grant its owner the right to raise insured deposits as well as give it access to central bank facilities.³

Once the licence is granted, banks are subject to very heavy oversight. They must retain a certain amount of capital linked to the size of their balance sheet and the riskiness of their assets. And they must run their balance sheets within clear liquidity parameters.

As important, regulators are constantly checking up on banks (or at least they should be). The Federal Reserve’s budget for banking supervision and regulation was $1.7 billion in 2022. In addition, the FDIC budgeted $1.1 billion for supervision and consumer protection, and the Office of the Comptroller of Currency (OCC) spent $1.1 billion. Between them, these three organisations employ over 30,000 staff. A team of 20 took over day-to-day supervision of Silicon Valley Bank in the second half of 2021.

All of this lies in the normal course of business for a bank. What is sometimes overlooked, because it is utilised so infrequently, is the executive power that authorities retain over banks. In some countries, where state owned banks dominate the market, intervention is explicit. During the global financial crisis, state-owned banks in India and Brazil carried on extending credit even when their private sector peers pulled back. 

But even when a bank is notionally private, the state can exercise direct influence over its operations.

This became clear during the pandemic, when regulatory authorities in the US and Europe unilaterally shut down banks’ dividend and share buyback programs.

It became clear when sanctions were declared on Russia at the outset of the invasion in Ukraine that states exercise foreign policy through their banking sector. “In Treasury, we realised that private-sector actors – most importantly, the banks – could drive the isolation of rogue entities more effectively than governments,” writes former US Treasury official, Juan Zarate, in his book, Treasury’s War. 

And it became clear again in Switzerland recently, when authorities decided Credit Suisse was no longer viable as a private entity. 

Wherever you sit in a bank’s capital stack, make no mistake: a higher authority sits above you.

3. Growth is … not good

Most companies thrive on growth. “If you’re not growing, you’re dying,” they say. For investors, growth is a key input in the valuation process. 

But if your job is to create money, growth is not all that hard. And if the cost of generating growth is deferred, because the blowback from mispricing credit isn’t apparent until further down the line, it makes growth even easier to manufacture.

Think about the mortgage market. The average life of a mortgage in the UK is seven years. An imprudent banker could originate as much volume as he likes right now; the cost may only become apparent later. The principle stretches into other areas of a bank’s business as well. When Deutsche Bank set up a special unit to ring-fence unloved assets a few years ago, the average life of a portfolio of interest rate derivatives it included was eight years  – they’d been put on three CEOs earlier.

Warren Buffett talks about a similar dynamic in insurance. In his 1985 shareholder letter he told how an advertisement placed in an insurance weekly garnered US$50 million of premiums. “Hold the applause,” he cautioned. “It’s all long-tail business and it will be at least five years before we know whether this marketing success was also an underwriting success.”

For these reasons, a focus on revenue growth for balance sheet businesses makes little sense, notwithstanding what some fintech companies would have you believe. 

Unusual asset growth has long been recognised as a red flag among bank analysts and regulators. Since the global financial crisis, regulators track how growth rates diverge from long-term trends. The Silicon Valley Bank collapse highlights how deposit growth can be just as unstable. If a balance sheet is a two-sided platform, it doesn’t matter where the growth comes from to make it vulnerable. 

The corollary to this is that, unlike in other industries, competition is not necessarily that good either – or at least it comes with a trade-off against financial stability. 

This is something I wrote about in my Bloomberg column this week:

There are many case studies of how competition can lead bankers astray. The classic is Chuck Prince’s acknowledgement that “as long as the music is playing, you’ve got to get up and dance.” Fifteen months later, his bank, Citigroup Inc. needed a government bailout. More recently, Silicon Valley Bank executives may have been incentivised to take more risk because their compensation was directly linked to the return on equity they generated “relative to peers.”

In a tradeoff between financial stability and competition, authorities will mostly choose stability. Last week, the chair of the Swiss financial regulator Finma brushed off antitrust concerns when she prodded UBS Group AG to take over a Credit Suisse Group AG that was dangerously close to collapse. “We can override monopolies and mergers as an institution and this is something we made use of,” she declared at a press conference.

It’s a playbook UK regulators used during the global financial crisis in 2008. When HBOS was shepherded into a merger with Lloyds-TSB, putting a third of UK personal current accounts into the hands of one institution, competition authorities baulked. They were overruled by a higher authority: “The public interest of ensuring the stability of the UK financial system outweighed competition concerns,” said Secretary of State Peter Mandelson.

US authorities are unusually squeamish about the trade-off. Partly, it reflects a respect for private markets but mostly it’s because their smaller banks harness significant lobbying power. More members of Congress have a small bank among their top 25 donors than a large bank, according to data compiled by Byrne Hobart.

The US is not necessarily making the wrong choice – its economy is more complex than others and its companies have more diverse financing needs. But it is a choice. As Thomas Sowell said, “There are no solutions. There are only trade-offs.”

4. Confidence is king

The absolute foundation of banking is confidence. Depositors entrust their money to banks in the confidence that they will get it back according to the terms of their deposit. 

As Matt Levine of Bloomberg’s Money Stuff writes (emphasis mine):

“The bank doesn’t just put your dollars in a box and wait for you to take them out; the bank uses its depositors’ money to make loans or buy bonds, and just keeps a little bit around for people who need cash. If everyone asked for their money back tomorrow, the bank wouldn’t have it. But everyone is confident that, if they ask for their money back tomorrow, the bank will have it. So they mostly don’t ask for it, so when they do, the bank does have it. The widespread belief that banks have the money is what makes it true.”

The history of banking is replete with innovations designed to cement that belief. Before the advent of deposit insurance, banks would market the reputation of their directors and their years in business to cultivate confidence. In the financial crisis of 1907, banks took out newspaper ads, detailing their financial condition.

Since then, banks have grown more complex, making any disclosure of their financial condition less intelligible to most consumers. Instead, they rely on credit ratings, capital ratios, liquidity ratios and supervisory assessment.

The problem is that these are all proxies for confidence rather than confidence itself. 

Credit Suisse ticked off all the proxies. When it collapsed, it did so with a regulatory capital ratio of 14.1%, ahead of its 13% target, and a liquidity coverage ratio of 144%. Once central bank support was factored in, that liquidity coverage ratio jumped to 190%, signalling that the bank had sufficient liquidity to meet almost 60 days of stressed deposit outflows. Without confidence, though, the bank was toast. 

It’s very difficult to restore confidence once it’s gone. One thing not to do is put out a press release saying your liquidity is strong. You’d think people would have learned after Bear Stearns, but no. When the proxies cease to work, saying it ain’t so won’t help either. 

5. Nobody knows anything

The dirty secret among bank analysts is that it’s quite hard for an outsider to discern what’s going on inside a bank. Former bank analyst Terry Smith knows it. “I think it is precisely because I understand banks that I never invest in their shares,” he wrote in the Financial Times last week.⁴

It’s only after the fact it becomes apparent what questions to ask.

For example, it turns out that Silicon Valley Bank’s top ten customers had $13.3 billion in deposits between them out of a total deposit base of $165 billion at the end of February 2023. That’s quite some concentration and it wasn’t disclosed. Since the bank has collapsed, we’ve also found out that regulators regarded the bank “not well-managed” as long ago as last summer. More unsavoury details will undoubtedly emerge when the regulators publish the findings of their inquiry in April. 

But historic inquiries reveal similar unknowns at other banks.

At Wells Fargo, management used to speak about “needs-based” cross selling. “Our cross-sell focus starts with customer needs,” said the group’s head of community banking in 2010. In contrast to these public statements, she implemented a volume-based sales model in which employees were directed to sell large volumes of products to existing customers. The cross-selling scandal was a huge reputational blow to Wells Fargo, leading to fines, regulatory censure and management turnover.⁵

At Credit Suisse, the inner workings of the prime brokerage business were laid bare only after the losses incurred on Archegos exposures came to light. The special report it commissioned into the episode is littered with examples of poor management practice. My favourite is that “neither of the co-heads of Prime Services believed he was specifically responsible for supervising CS’s relationship with Prime Financing clients in the United States—including Archegos.” 

Near-miss and indeed crash investigations are as informative in banking as they are for airlines and investors are wise to read them. The only problem is that the number of things that can go wrong is vast and so banks will find new ways to fail. One helpful heuristic is that banks are unlikely to be upended by the same factors that drove the last crisis. 

Paid subscribers please read on for my comments on the First Citizens acquisition of Silicon Valley Bank, the disintermediation of banks by money market funds, and the future of bank liquidity requirements. Also, a programming note. If you are not already a paid subscriber, please consider signing up...