Good bankers know the industry they serve. They know its idiosyncrasies, its cast of characters, its business cycles. They help their clients through good years and bad — and are rewarded with deep loyalty and fat deposit accounts.

Why it matters: The industry that bankers know differs from bank to bank — but it's never banking. A wine-industry relationship banker at Silicon Valley Bank has to know much more about malolactic fermentation than about Tier 1 capital adequacy ratios.

• Even at the CEO level, someone like SVB's Greg Becker found himself on the board of the San Francisco Fed not because he was an expert in banking, but because he had extraordinary visibility into California's technology industry.
• With hindsight, that was an egregious conflict, given that the San Francisco Fed was SVB's primary regulator. All the same, bankers really do have an unrivaled level of insight into what's going on with companies that are too small to have gone public — including Silicon Valley startups.

The big picture: Capital markets and investment banks are very good at providing banking services to big public companies — but only about 15% of American jobs are at big public companies.

• Commercial banks therefore provide a crucial and irreplaceable service — and, as is natural in all industries, they tend to specialize.
• Silicon Valley Bank banked Silicon Valley, as its name implied, as well as the California wine industry. Signature Bank banked Broadway. Silvergate banked crypto.
• Such relationships take time to build and nurture, which is one reason the federal government is keen to sell SVB and Signature as going concerns — and is a reason it's not trying to rescue Silvergate, which operated in an industry that authorities broadly want to be cut off from the regulated financial system.

Between the lines: In countries dominated by three or four massive universal banks, there's less idiosyncratic industry risk when a single mid-sized bank fails. But the U.S. has made a policy decision that it doesn't want its biggest banks to get any bigger than they already are — which means many more smaller banks with a broad range of specialties.

Where it stands: Risk management is a cost center for banks, not a profit center. The way to rise up the ranks at a commercial bank is to make money for your employer — and the way to do that is to know your customers and their needs really well.

• When commercial banks fail, outsiders are often shocked at how badly they handled things like interest rate risk or deposit concentration risk. But those jobs are done only at the top of the org chart — and the people who rise that far tend to do so not because of their understanding of how banking works, but because of their understanding of other industries entirely.

The bottom line: There's an obvious financial-stability reason for regulators to want to minimize bank failures. But there's also a grassroots economic reason. Regional banks like SVB and Signature are deeply embedded in local industry, and without them many industries would find themselves in deep trouble.

Here's a riddle for you. Shares in First Republic Bank closed at $31.21 on Monday, and opened at $49.69 on Tuesday — an overnight spike of 59%. There was no news about the bank over the course of the intervening hours. So why did the stock move so much?

• The answer: The stock moved so much because there was no news about the bank over the course of the intervening hours. But also, once banks start to get into trouble, the share price naturally becomes incredibly volatile, just because their balance sheet leverage is so enormous at that point.

Between the lines: There are a lot of possible futures for any given company, and a stock price can be thought of as a kind of weighted average of all of them, discounted by a similarly probabilistic range of discount rates.

• At times like these, certain banks start to become the subject of worries that they might get taken over by the FDIC overnight and ultimately sold to a new owner.
• When that happens, depositors (account holders) are generally unaffected, while shareholders are generally wiped out. The bank lives on — do a blood test on JPMorgan and you'll find WaMu DNA in there somewhere — but that's no solace to its former owners.
• In other words: It doesn't matter how valuable Signature Bank or SVB end up being to their future owners, whoever that might be. All that value accrues to the new shareholders, while the old shareholders remain zeroed out.

How it works: When markets closed on Monday, there was some non-negligible chance that federal authorities would look at First Republic's deposit outflows and take it over before markets opened on Tuesday.

• That possibility was reflected in its Monday share price — so when Tuesday morning arrived and nothing had happened, the stock price, no longer reflecting the probability of a Monday-evening failure, went up.

The big picture: Banks are some of the most levered businesses in America — they have vastly more debt than equity. That's just how fractional-reserve banking works.

• Debt is normally a combination of total deposits, bonds outstanding, and various other instruments like repos; equity is normally gauged using measures like Tier 1 capital.
• Take a typical boring bank like PNC Bank. Its capital is about $43.5 billion, while its deposits alone — all of which are unsecured liabilities of the bank — are ten times that sum.

The bottom line: Because equity value is such a small sliver of a bank's overall capital stack, bank shares tend to be particularly volatile, especially when market capitalization falls well below book value.

With T-bills yielding 5%, it's hardly any wonder there's been a huge surge of enthusiasm for them, even when the main website for purchasing them feels like it's stuck in 1998.

Between the lines: The above chart shows the number of people buying T-bills on TreasuryDirect through February, before the banking crisis.

• In March, of course, millions of Americans with deposit accounts holding more than $250,000 realized that (a) a lot of their money was uninsured; and (b) they could get much better interest rates just by buying T-bills.

So here's the contest. The number of people buying T-bills through TreasuryDirect was 189,823 in February.

• What do you think it's going to be in March? Reply to [email protected] with your best guess, and the person who comes closest* will win a signed copy of my forthcoming book.
• The winner will be announced on either April 15 or April 22, depending on when Treasury releases the March data.

The spike in the number of individuals directly holding Treasury bills makes it more likely that the government will find its way to raising the debt ceiling.

Why it matters: The biggest reason the debt ceiling is such a big political problem is that most Americans don't care about it, and most of those who do think that its existence is a good thing and that it shouldn't be raised.

• Holders of T-bills, by contrast, are acutely aware of the necessity that the government pays its debts in full and on time.

Between the lines: The kind of people who are engaged enough to buy T-bills also tend to be voters with relatively unfettered access to their local member of Congress. After all, very few people buy T-bills unless they're first maxed out on their $10,000 allocation of I-bonds.

The bottom line: Raising the debt ceiling is still not going to be easy. But the current flight to quality out of bank deposits and into Treasury securities definitely helps.

Bab Mansour, in Morocco, is by far the grandest of the 20-odd gates surrounding the city of Meknes.

• Designed by and named after the architect, Mansour Laalej, it was completed in 1732.
• Its green and white tiles are now faded with age; its marble columns date to ancient Rome and were probably relocated from the nearby site of Volubilis.