Todd Baker is a senior fellow at the Richman Center for Business, Law & Public Policy at Columbia.
US banking has always been a risky business. The financial panics of 1819, 1837, 1873, 1907 and 1931-32 all sparked banking crises, recessions or full-scale depressions. Depositors lost everything every time a bank failed. But the Great Depression was the trigger for a seismic change.
Congress created the Federal Deposit Insurance Corporation (FDIC) during the Great Depression to stop the rolling banking panic by guaranteeing bank deposits. It was originally up to $2,500 in per eligible account — an amount quickly raised to $5,000 or about $112,000 in today’s dollars — and subsequently raised six times over the years to the current $250,000 limit.
At the same time, the US developed a comprehensive system of bank regulation to complement deposit insurance, including strict separation of banking from securities and other commercial activities. Banks effectively became private/public hybrids, with a federal deposit backstop and comprehensive regulation combined with private capital to enable banks to fulfil their critical role in maturity transformation with relative efficiency.
For a long time, it seemed like this private/public system worked perfectly. Bank runs stopped for almost 90 years, most banks were consistently profitable, failures were rare and, as the FDIC proudly states, “no insured depositor has lost a penny.” Losses were repaid through assessments on banks, and the prevailing view was that the theoretical problems of moral hazard and “private gains-public losses” were largely contained.
Here’s the FDIC ruminating on its own history in 1998, with Alphaville’s emphasis:
In its seventh decade, federal deposit insurance remains an integral part of the nation’s financial system, although some have argued at different points in time that there have been too few bank failures because of deposit insurance, that it undermines market discipline, that the current coverage limit of $100,000 is too high, and that it amounts to a federal subsidy for banking companies. Each of these concerns may be valid to some extent, yet the public appears to remain convinced that a deposit insurance program is worth the cost, which ultimately is borne by them. The severity of the 1930s banking crisis has not been repeated, but bank deposit insurance was harshly tested in the late 1980s and early 1990s. The system emerged battered but sound and, with some legislative tweaking, better suited to the more volatile, higher-risk financial environment that has evolved in the last quarter of the 20th century.
Oops. The 2008 financial crisis shattered that illusion, and this year’s expensive failures at Silicon Valley Bank and others show that the system has not returned to a state of stability. That was ably visualised by Toby Nangle here:
Where did things go wrong?
Many things have changed in the economy over the 90 years since the FDIC was created and the modern regulatory architecture was designed. “Shareholder value maximisation” theory gradually undermined traditional bank management conservatism, the 3-6-3 rule and the assumption that reasonable but stable returns were enough for banks.
Once shareholders grew more active and bank managers began to be compensated through equity awards tied to the bank’s stock price, they were strongly incentivised to take risks to generate personal wealth. The steady financialisation of the US economy and a wave of deregulation blurred the lines between boring banking and high-risk investment and capital markets activities.
Many new financial players and myriad new ways to create financial leverage, risk and return came into existence, like securitisation, private equity, derivative trading, venture capital, shadow banks, high speed trading, money market funds and private debt providers, to name only a few. Bank managers seeking to add risk had many new tools to choose from.
What had been a grand bargain with government — in which banks agreed to run their businesses conservatively, hold excess capital and liquid resources, only do certain useful and relatively low risk things and be subject to strict supervision in exchange for the business model stability that federal deposit insurance brought — turned into unbalanced and unbridled private risk-taking, subsidised by a public backstop.
The recent banking shenanigans show that post-crisis reforms have helped, but not sufficiently tamed these issues. As Stanford’s Anat Admati says:
When I say the banking system is too fragile, I am often told, “Things are just fine. It’s a good system. The rules were reformed.” Some of us have been arguing for years that these narratives are false and misleading. The failure to write proper rules in banking goes back decades.
The reform efforts after the financial crisis of 2007-09 — which was a direct result of terrible rules — have been pretty much a mess. Some were akin to reducing the speed limit for loaded trucks from 98 miles per hour to 95 miles per hour without even measuring the speed properly. Politicians and regulators are failing us. And when the flaw in the system becomes visible, as in 2008 or now, they cover up their failures and provide bailouts if that seems better than the alternatives. Then the public stops paying attention and bank lobbyists make sure the rules continue to tolerate recklessness.
Leading voices in finance, policy and academic circles are arguing that the source of the problem can be found in the deposit insurance system and that something about deposit insurance needs to change.
Mainstream pundits have proposed things like expanding insurance to cover all deposits; adding layers of regulation to control the risk du jour; creating specialised money-market funds with loss-absorbing capacity instead of deposits; privatising parts of the insurance system. Fringe commentators have even espoused solutions involving crypto and stablecoins (sigh), narrow banking and even a return to the good old days of wildcat banks.
It’s worth noting that there never has been a true consensus about the purpose and limits of deposit insurance. Even president Franklin Roosevelt (a surprising opponent) described deposit insurance as “dangerous,” for in time it “would lead to laxity in bank management and carelessness on the part of both banker and depositor.”
At his very first press conference Roosevelt’s scepticism was apparent in an off-the-record briefing for journalists:
I can tell you as to guaranteeing bank deposits my own views and I think those of the old administration. It is that the general underlying thought behind the use of the word “guaranty” with respect to bank deposits is that you guaranty bad banks as well as good banks and the minute the government starts to do that the government runs into a probable loss.
The whole objective of the plan we are working on can be stated this way: There are undoubtedly some banks that are not going to pay 100% on the dollar. We all know it is better to have that loss taken than jeopardize the credit of the United States Government or to put the United States Government further in debt, and therefore, the one objective is going to be to keep the loss in individual banks down to a minimum, endeavoring to get 100% on them, but not having the United States Government liable for the mistakes and errors of individual banks and not putting a premium on the future of unsound banking.
(While this view still prevails among those dedicated to the idea that deposit insurance creates moral hazard risk, there’s no good evidence that this happens in practice.)
Some think deposit insurance is a consumer protection tool, like limits on credit card fraud losses. Others view it as one of the pillars necessary to control funding volatility and shore up the eternally fragile lend-long/fund-short model of banking. New ways to spread money around banks have made the cap on insurance is essentially meaningless to savvy consumers, so it should surprise no one that the current $250,000 limit was dreamt up by Congressional staff as a sweetener to get a larger bill passed, without any analysis.
Isn’t it possible that we’re asking the wrong question about how to fix problems with banking system stability?
Maybe we shouldn’t do anything to change the deposit insurance system, beyond the FDIC’s eminently reasonable suggestion of greater protection for business operating and payroll accounts. And maybe more regulation isn’t the answer either, as it’s hard to argue that regulation and regulators have proven effective in addressing the novel and fast-moving circumstances that lead to bank failures.
As for returning to the low-risk world of Glass-Steagall? Even a supermajority in Congress couldn’t unbake that cake.
Maybe, just maybe, the best way to reduce risk in the banking system and restore the balance in the Depression-era bargain is through people?
After all, it is human error that causes bank failures. Citibank, Bank of America, Wachovia and Washington Mutual all failed in 2008 (three were bailed out and one was seized and sold) because managers fatally misunderstood mortgage risk. Silicon Valley Bank failed this year because its executives made a ruinous bet on interest rates and deposit duration that its board failed to adequately understand or control.
When it comes to people — whether it’s Chuck Prince saying “you’ve got to keep dancing while the music is playing” or Greg Becker deciding to spin the roulette wheel on interest rates and durations — you get what you reward. Change the incentives and you change the outcomes.
There is some precedent for thinking incentives are important in banking. In the days when bankers were compensated with salaries, cars, retirement plans, country club memberships and prestige instead of equity-based bonuses, they had little incentive to ramp up risks for maximum return at the cost of safety.
They did well as long as the bank stayed healthy, even if financial performance was only “above average”. They also spent as much time thinking about customers, employees and communities as they did thinking about shareholder returns.
While this system was by no means perfect (this was the source of economists’ dreaded “agency cost” problem which in part led to the equitisation of executive pay), a revised approach to incentives could help restore the Depression-era banking bargain, and tip the balance back towards systemic stability.
Let’s start by banning equity-based compensation for bank management and boards. A total ban goes far beyond what was contemplated by the Dodd-Frank Act Section 956, which was meant to curb compensation plans that encourage bankers from taking inappropriate risks but has had only a small impact.
Congress actually required regulatory agencies to finalise a rule implementing Section 956 by May 2011, but it has never happened. ¯_(ツ)_/¯ As senator Gary Peters — one of the authors of the Dodd-Frank Act — wrote to regulators including Gary Gensler and Jay Powell in March:
Section 956 of Dodd-Frank was intended to require financial regulators to quickly and collaboratively issue rules requiring financial institutions to disclose any incentive-based executive compensation arrangements that encourage excessive risk-taking at financial institutions or that may lead to financial loss. The recent bank failure of Silicon Valley Bank and reported bonuses issued to its leadership further underscore the urgency and importance of this rule’s implementation. Given how incentive-based compensation can continue to lead to certain financial institutions and professionals taking excessive and reckless risks, implementation of this long-delayed rule is an important reform to ensure reckless financial risks and financial mismanagement do not put our banking system at risk.
So why a total ban? Well, equity-based compensation just isn’t appropriate in the private/public hybrid banking model.
Bank managers in effect serve two masters — the board/shareholders and the FDIC insurance fund and its backer, the US Treasury — so bank executive bonuses and director compensation should be based on balanced targets serving the needs of both masters, with a mix of profitability, risk management and other metrics, and large holdbacks for subsequent negative or positive developments.
The same concepts should apply for line manager and “risk-taker” compensation — balanced scorecards with risk management as well as profitability targets and extended holdbacks. The total amount of compensation isn’t the problem. It’s the behaviours incentivised by the compensation structure.
More fundamentally, we probably need to rethink which parts of our corporate law should be applicable to private/public hybrid institutions. As Martin Wolf pointed out recently, they should be seen as quasi-utilities:
We get pretty upset if all the electricity got cut off because our electric utility failed to manage perfectly obvious risks. And in the same way, we expect banks to continue to function. And the way we’ve ensured that is to give essentially an open-ended guarantee from the state that we will allow them to do so . . .
. . . I think that to me, the essential point is that banks are, really are utilities. They’re not and shouldn’t be the prime risk-takers in the economy. That’s what we have equity for. And if they are to be seen as utilities, they don’t need to be vastly profitable. They need to be run as utilities and be capitalised in ways that ensure that they will survive in tough times. Because surviving in tough times is the most important thing banks can do.
Even if you don’t see them as de facto utilities, it’s clear that banks really aren’t like other companies, and corporate law should recognise that difference.
State and federal law should be changed to explicitly require bank boards to consider the interests of the FDIC insurance fund and the larger economy in the exercise of their fiduciary duty as corporate directors. And the FDIC should be able to sue them if they don’t.
These solutions aren’t very complex, nor do they require new deposit insurance structures based on economic theories that have regularly failed to conform to our human-centred reality in practice. Will they work? It’s certainly worth a try?
The structural changes adopted with great fanfare after 2008 didn’t stop SVB from collapsing, taking a couple of other banks with it, and requiring a $16bn refill of the insurance deposit scheme.
Maybe incentive changes would have caused management at Silicon Valley to think twice, and consider the impact on their personal pocketbooks before they created a balance sheet that breached all their internal risk limits and turned into a loaded gun pointed at the FDIC and their shareholders.