Discover more from The Workbench
We Want Banks to Do Two Mutually Contradictory Things
Banks make loans. And they provide us with a money system. But those are two very different economic functions, for which voters and regulators have completely incompatible expectations.
What are banks for, anyway? The bog-standard definition is they accept deposits and make loans.
To grotesquely simplify matters, banks take in money that people want to store, and then loan that money out to other people who need it. Banks attract those deposits by paying depositors an interest rate. And they charge interest to the people they loan money to. Banks then make their profit on the difference between the interest they pay and the interest they take in.
Banks also need to keep enough money in reserve so that anyone who wants to cash out their deposit can do so. Banks never actually have all their depositors’ money on hand, the idea being that only a fraction of depositors will want to cash out at any given moment. So banks are constantly managing a balancing act between money flowing in and money flowing out. Banks that do this well will profit and thrive; banks that do it poorly will fail and go under.
The trouble is that these two seemingly straightforward activities—taking in deposits and making loans—have evolved into two wildly different functions within our economy. And our social and political expectations for those two functions are mutually contradictory.
This is something I’ve wanted to dig into ever since Silicon Valley Bank collapsed back in March, setting off yet another (blessedly momentary, this time) crisis in the U.S. banking system. Even though the system ultimately “worked” in response to the blow-up, it still seemed to leave a thoroughly bad taste in everyone’s mouth. We keep running headlong into this problem without quite recognizing the nature of the wall we just crashed into.
Two Great Tastes That Don’t Taste Great Together
Bank loans provide a big portion of the financial capital that goes into starting up new businesses, or expanding existing ones. This makes bank loans central to the growth of the economy, innovation, the creation of jobs, and rising living standards.
Bank loans also involve necessary and unavoidable risks. Every loan finances some economic endeavor, and every loan is a bet that the endeavor will pay off; that it will result in real wealth creation and real social benefit, thus earning enough returns to pay back the loan plus interest. If a bank fails at those bets often enough, then it goes bust.
More to the point, banks must be allowed to fail and go bust. Bad bets use real resources (workers and materials) that need to be freed up for better bets. And the only way banks can figure out how to make better bets is through the iterative learning process of trial and error. Ultimately, that’s the whole raison d'être of market competition. If banks aren’t allowed to fail—if they’re bailed out and rescued from their follies—then they don’t learn. So there’s a moral objection to bailouts, in that they effectively reward people for foolhardiness and doing a bad job. But there’s also a coldblooded economic argument against bailing out failed banks, in that doing so short-circuits the market’s trial-and-error learning process. Resources will be put to poorer use, growth will slow down, and living standards will stagnate.
Thus, for bank loans to perform their economic function, we must tolerate the constant presence of both risk and instability—the ever-present churn of Schumpeter’s good ol’ creative destruction. And that reality runs headlong into the second function of banks in our economy: to take in deposits.
The bog-standard definition of a bank makes it sound like depositing your money is a gamble, same as if you were buying stock: you are betting that the bank you deposited with will make good bets. You’re taking responsibility for the possibility that all the money you deposited could go “poof.”
But tell any everyday worker, consumer, or business person that opening a plain-vanilla checking or savings account is equivalent to playing the stock market, and they’ll find the analogy bizarre and perverse. No one thinks they need to assess a bank’s stability, research its portfolio, or consider its reputation, just to open a damn checking account. We deposit our money in checking and savings accounts to avoid the kind of risk involved in playing the stock market.
In the modern context, “taking in deposits” really means “providing a money system.” We deposit money in banks to store it, but we also use our bank accounts to conduct the vast majority of our economic transactions. They’re the plumbing of commerce; the underlying payments infrastructure; the holding tanks and pipes by which we transmit money to one another, and store the money we haven’t used yet.
When it comes to that function, we do not want to tolerate risk and instability. We want risk and instability to be minimized. If our bank collapses and our deposits vanish, we don’t shrug and chalk it up to the facts of life. And we’d be furious with anyone who suggested we should. That’s not a risk we think we should “internalize.” We think the gambling should be left to actual investments and new business endeavors, and not mixed up with the mundane act of simply moving money around. Much like our actual public city pipes, we want the monetary plumbing that banks provide to just be there, functional and reliable, whenever we need it.
So, on the one hand, we have loans, for which risk and instability are necessities. On the other hand, we have deposits and the money system, from which risk and instability should be purged. We expect banks to perform both functions while meeting both sets of expectations, and we expect government policy towards banks to honor both sets of expectations.
Needless to say, this is asking for trouble: Every policy attempt to make one function better will make the other function worse. And unless our policy and regulatory designs are exceptionally smart, efforts to split the baby and balance between the two expectations will simply make the banking system more incoherent, contradictory, and dysfunctional.
Stabilizing the Money System Puts the Banks on Easy Street
The fact that today, you really can treat your checking and savings accounts as infrastructure—rather than an investment gamble—is not the “natural” result of market forces. It’s the result of massive and proactive policymaking by the U.S. federal government, which has knit the for-profit, private banking industry into a stable and coherent whole. (Sort of.)
Under the Lincoln administration, we passed a series of National Bank Acts that sought to organize the country’s banks into a unified framework, laid down some regulatory guardrails for banks’ finances, and established a federal chartering system to make sure new banks had sufficient financial resources to succeed. In 1913, we created the Federal Reserve: the country’s central bank, and also the “lender of last resort.” Through a program called the discount window, the Fed loans banks U.S. dollars—in exchange for collateral—to keep them afloat when no one else will.
In 1933, we created the Federal Deposit Insurance Corporation (FDIC), which charges banks a fee to fill an insurance pool, which then guarantees everyone’s deposits up to $250,000. Not only does this protect the vast majority of everyday depositors from losing their money when a bank fails, it also protects the banks from failing in the first place, by discouraging bank runs.1 (Unfortunately for Silicon Valley Bank, it catered to Big Tech venture capital and startups, and thus the vast majority of its deposits were over the $250,000 threshold, leaving it unprotected.) The FDIC also uses that money to take ownership of a failing bank, stabilize its finances, and then sell off its assets to another bank: for depositors, their money stays where it is, and only the brand name on the bank building changes.
But all these policies are also implicit subsidies: they allow banks to take risks in their loan-making that they would not have contemplated otherwise. And they allow the investors and executives who own and operate the banks to walk away with even bigger profits. The price of stability and coherence for the money system is that we dilute the possibility and consequences of failure for bank loans.
Of course, legislators and regulators are aware of this. So they try to design these policies to “mimic” market forces, and to limit the implicit subsidies.
When the Fed gives banks loans through the discount window, for instance, it does so at interest rates that are less attractive than what banks could get from their fellow private banks under “normal” circumstances. And when the FDIC takes a bank into receivership, all the bank’s shareholders and unsecured creditors lose all the money they had invested in the enterprise. As you might recall, this is precisely what happened to the shareholders and unsecured creditors of Silicon Valley Bank (SVB, henceforth) and Signature Bank, when those two institutions went belly up. Those investors were punished for their folly, in the way “market logic” demands.
Yet at the same time, another bank, First Citizens, was able to buy up much of SVB’s operations and assets from the FDIC—to the benefit of the executives and shareholders at First Citizens—which it likely couldn’t have done had SVB been allowed to collapse the old-fashioned way. And while the FDIC may “only” insure deposits up to $250,000, even that safety net creates a set of circumstances that encourage banks as an aggregate population to take risks they otherwise might not. “In some cases, the risks won’t pay off,” as economist Scott Sumner put it at the time. “But that doesn’t mean executives don’t have an incentive to take excessive risks.”
As for the Federal Reserve, the key thing to realize there is that, while it may lend to banks on comparatively unattractive terms, it isn’t seeking a profit and cannot go bankrupt and collapse in the ways private banks can. Thus, it can keep lending to a troubled bank through the discount window long after private lenders would’ve abandoned it. Despite the harsher interest rates it offers, the fact that the Fed’s discount window exists at all changes the banks’ incentives for risk-taking, compared to what they’d be if the Fed didn’t exist.
When Drawing Arbitrary Lines, You Never Know Where To Draw Them
What was even more striking about the SVB crisis was that federal regulators decided to use their emergency powers to dramatically expand upon this standard safety net. The Federal Reserve, the FDIC, and the Treasury Department announced a joint decision that all deposits at Silicon Valley Bank and Signature Bank would be covered by FDIC insurance—not just deposits under $250,000. So they took the subsidy for the banks that’s already implicit in standard FDIC insurance and scaled it way up.
Meanwhile, the Federal Reserve had already dramatically lowered the interest rate for loans through the standard discount window—in response to the 2020 crisis set off by the COVID pandemic—bringing it almost even with what private banks charge each other. Then, it responded to SVB’s collapse with a whole new lending program, offering U.S. dollars to the banking system on even more attractive terms. Most strikingly, the financial assets the banks put up as collateral for Fed loans would be valued “at par.” Meaning the Fed would value the asset according to what the bank originally paid for it, instead of what the private market would pay for the asset in the here and now.
Put in practical terms, when your bank offers you a credit card, it offers it with terms—interest payments, fees, lending limits, etc—designed to make the bank a profit and protect its bottom line. But imagine your bank offered you a credit card with terms that ignored the bank’s desire for profit completely. Instead, it specifically designed the terms for your sake, to keep you afloat through whatever your financial troubles were. That’s basically what the Fed offered the banks.
These were all big changes to the “rules” mid-game. And while they may not have saved the specific folks in charge of Silicon Valley Bank and Signature Bank, they were a massive financial boon to the ownership class of the banking industry as a whole. In the first four weeks after SVB’s collapse, bank borrowing through the standard discount window jumped from almost nothing to nearly $400 billion. Borrowing from the new lending program totaled around $200 billion.
Furthermore, given that SVB wasn’t even that large a bank in the grand scheme of things, it’s now much more likely that federal regulators will “break the glass in case of emergency” for any moderately-sized bank failure going forward. That would make the $250,000 limit to FDIC insurance, and the more stringent standards of the Fed’s normal lending program, just technical formalities. In practice, the terms of the banking system’s safety net will be much more generous from here on out.
Was all this actually necessary? I honestly have no idea. There’s no Earth 2 where we can run the same scenario, but this time without federal regulators’ massive intervention.
What I can say is that not knowing where to draw the regulatory line is itself a problem that’s intrinsic to mashing up the banks’ loan-making function with their money-system function. All the cut-offs we’re talking about here—the $250,000 limit on FDIC insurance, the terms offered in the Fed’s standard lending facilities—are arbitrary to begin with. Both the Federal Reserve and the FDIC are ultimately backstopped by the federal government’s power to create U.S. dollars, and neither can ever truly “run out of money” if policymakers don’t want them to. Therefore, the limits on FDIC insurance, and the terms on which the Fed offers to loan money, can be whatever the hell we want them to be. Once we’ve decided to enact policies that bring stability and reassurance into the banking system, there’s no “right” answer as to how far to take them.
The financial regulators at the FDIC, the Federal Reserve, and the Treasury Department are only human. And there are just a lot of unknowns when you’re in the teeth of that sort of crisis, trying to manage it on a moment-to-moment basis.2 So, when faced with the imminent collapse of several “mid-sized” banks that may or may not have been big enough to threaten the entire banking system, I can’t say I blame regulators for deciding to just nuke the entire site from orbit. As the lady said, it’s the only way to be sure.
Give the Finance Capitalists Everything And Ask Nothing in Return
At the time, Nathan Tankus argued that what should disturb us about the SVB episode isn’t so much that regulators made such creative use of their emergency powers. What’s disturbing is all the other times they didn’t do so. For instance, the Fed could’ve similarly used its leeway in emergencies to break norms and save Detroit and Puerto Rico from bankruptcy. Both of those crises resulted in massive cuts to public programs, and enormous damage to the lives of hundreds of thousands—if not millions—of everyday people. But the Fed decided that “the rules” and “market logic” militated against intervening.
Of course, a full-on banking system crisis, set off by SVB’s collapse, would’ve also done enormous damage to the lives of hundreds of thousands—if not millions—of everyday people. But it would also have done massive damage to the wealth and position of a good number of financial elites. And that’s why the Fed stepped in with the emergency-powers nuclear bomb for Silicon Valley Bank's collapse, but not for the people of Detroit and Puerto Rico.
Here’s yet another problem inherent to asking the banks to both provide a money system and make loans: the incoherence and contradictions provide abundant opportunities for exploitation. We will inevitably create policies to stabilize the money system for the sake and well-being of everyday Americans as a whole. But a knock-off consequence of that stabilization is the banks’ ownership class can privatize their gains and socialize their losses. And in an economy as unequal as ours, in which the rich and powerful wield such enormous influence and leverage, they can use that confusion to their advantage. They can tilt the design and implementation of those stabilization policies, seeing to it that norms get broken when it suits them and not broken when it doesn’t. They can make out like bandits when times are good, and leave the rest of us holding the bag whenever they fuck up.
One way to tackle this problem would be to impose excess profits taxes on the banks. Yet another strategy is regulation. The policies we’ve been talking about thus far—the Federal Reserve system, the FDIC, etc—aim to stabilize the banking system as a whole. But the financial regulations created by laws like 2010’s Dodd-Frank Act aim to make individual banks safer and more stable. Dodd-Frank does this by limiting how much banks can borrow relative to their assets, and by requiring banks to keep a certain amount of low-risk assets on hand relative to liabilities. And those individual-level regulations also make banks less profitable, by forcing them to devote more money to financial buffers and less to making more of the loans that make them money.
So while financial elites and the banks’ ownership class are generally fine with things like FDIC insurance and generous lending from the Fed, they really don’t like Dodd-Frank and similar regulations. The latter rules function as a de facto excess profits tax on the banking industry, as Matt Klein put it. The point is partially to ensure stability and coherence in the system. But another point is to prevent the banks and their owners from acquiring too much power and wealth. That strong and overbearing regulations prevent enormous profits is a feature, not a bug.
In the wake of the SVB fiasco, much was also made of the change that Republicans—with the aid of centrist Democrats—made to Dodd-Frank’s regulatory framework in 2018: Originally, the law subjected all banks with assets above $50 billion to much stricter regulatory demands and scrutiny. But the 2018 change lifted the threshold to $250 billion. Silicon Valley Bank was one of many institutions directly lobbying for that change, and of course it had roughly $200 billion in assets when it collapsed.
Admittedly, there’s no guarantee that stricter regulatory oversight would’ve prevented SVB’s implosion. But the episode is still galling, because of the cravenness of SVB’s owners specifically. And because the whole justification for the 2018 change—that banks under $250 billion in assets aren’t really “systemic risks”—was belied by the massive effort regulators took to stabilize the money system in the wake of the bank’s collapse.
Yet even the implicit excess profits tax imposed by Dodd-Frank’s regulations before 2018 was pretty easy-going compared to the New Deal-era regulations banks faced after World War II. Back then, banks were flat-out forbidden from opening branches in more than one state, which put a pretty hard cap on how big they could get. There was a hard separation between retail banking, investment banking, and insurance—an institution could only do business in one of those three areas—which again limited size and revenue. Regulations even controlled the design of loans, dictating who the banks could loan to and what interest rates they could charge.
Policymakers dismantled the New Deal-era rules once American politics took its post-Reagan turn towards small government and free market ideology. And we only began inching our way back after the 2008 financial crisis.
Our political system is happy to stabilize the banks—and the money system they create—when that stabilization comes with implicit subsidies for the banks and their owners. But when those stabilizing policies also limit the banks’ power and profits, our leaders approach them gingerly, embracing them only after the worst sorts of debacles—and then retreating from them as fast as they can once memories start to fade. As Peter Conti-Brown observed, government regulations for banks have come and gone throughout the country’s history. But thus far, FDIC insurance has been forever.
So what do we do about all this?
I don’t think it’s strictly impossible to have the banks perform both functions—making loans and providing a money system—while also imposing a regulatory framework on them that’s both coherent and sufficiently strong. But I do think it’s a very difficult thing to achieve. It not only requires exceptionally smart policy design, it requires putting some pretty draconian limits on the wealth and power that banks’ owners can amass, because they’re always going to be pushing back at the political system for more freedom and leeway.
Frankly, most of the talk about what to do in response to Silicon Valley Bank’s collapse struck me as pretty unambitious: doing away with the 2018 change to Dodd-Frank, re-inscribing the limits on FDIC insurance and Fed lending, and taking away some for those institutions’ emergency powers. None of which strikes me as nearly enough. If we want to truly make banking boring again, we need a return to something like the old New Deal-era rules. And it will be a very long walk from here to there.
To my mind, what’s more promising is just splitting the two functions off into different institutions. More specifically, separate the money system—checking and savings accounts and payments—from banks taking deposits and making loans. Have the government provide that money system as a public institution, and let people put their money in it for the mundane tasks of saving and making transactions. That way, depositing your money at a bank really would be like playing the stock market. And the only people who would do it would be the ones who wanted to gamble.
I’ll try to dig into all of that in more detail in a future post.
If you know your money is insured by the FDIC, then there’s no reason to panic when there’s a crisis, and no reason to pull your deposit from the bank. That’s important, because if everyone panics and pulls their deposits at the same time, then you’ve got a bank run—whether that panic was justified or not.
When a bank collapses, the threat isn’t just that the depositors at that one bank will see their deposits go up in smoke. It’s that the banking system is inescapably interconnected, and the collapse of one bank destroys streams of money going to other banks, which can then drag them down as well. There’s also the mass-psychology aspect of bank runs. If you see other people lose their deposits when their bank folds, and you think—rightly or wrongly—that there’s any chance the same could happen to your bank, you may well pull your deposits too. No bank on earth can survive all its depositors pulling their money at the same time, no matter how well-designed its finances and portfolio are. Bank runs are a kind of irrational crowd panic, which can spread from a failed bank to bring down healthy banks. And that’s even more true in an era when you can pull your deposits electronically and instantaneously. Silicon Valley Bank’s depositors yanked $42 billion out of their accounts in a single day, making it the biggest—and by far the fastest—bank run in U.S. history.