As expected, the FDIC released an overview of the deposit insurance system yesterday, along with options for deposit insurance reform. I, like many of you I’m sure, stayed up late into the night reading through the 75 pages. It’s a nice report; it covers the history of deposit insurance in the U.S., objectives of insurance, tools to support those objectives, and proposed options for increased insurance coverage.
In short, nothing is going to change for the foreseeable future, and the FDIC’s preferred changes are not too dramatic.
At a high level, the FDIC proposes three options for deposit insurance:
Limited coverage: Maintain the current deposit insurance framework, which provides insurance to depositors up to a specified limit (possibly higher than the current $250,000 limit).
Unlimited coverage: Extend unlimited deposit insurance coverage to all depositors.
Targeted coverage: Offering different deposit insurance limits across account types, where business payment accounts receive significantly higher coverage than other accounts.
Of the three options, the FDIC is clear in the report that it favors number 3, targeted coverage. The change would require Congressional action, and there will likely be a lengthy period of discussion and commentary before we even get there.
In the following, we discuss the report, the proposed options, their pros / cons / general impacts, and our thoughts on the best way forward. Importantly for ModernFi and banks that use deposit networks, the FDIC’s report looks favorably on reciprocal products and sweep products (jump to section 6 if interested).
This is a longer post than normal, but any changes to deposit insurance will have a dramatic impact on the banking sector so it’s worth digging in to the implications (and it’s less than the 75 page report, I promise). Numbers in parentheses below refer to pages of the full report.
Table of Contents
Banking as a business model and the role of insurance
The importance of deposit diversification
FDIC’s focus on the “composition of deposits”
The role of technology
FDIC’s proposed options for increased deposit coverage
Limited coverage
Unlimited coverage
Targeted coverage
Our proposal for increased deposit insurance (and the FDIC’s view of deposit networks)
1. Banking as a business model and the role of insurance
Banking is an odd business model; you use short-term deposits to make long-term loans. This duration mismatch creates risk. As Martin Gruenberg simply states, “The business of banking, which accepts deposits that are available on demand while making long-term loans, remains susceptible to runs.” However, loans, or credit in general, is fundamental to advanced economics. Credit, and access to credit, drives productivity, efficiency, and growth. When banks do badly, such as in financial panics, they stop lending and credit dries up, which can cause deep recessions and depressions. As a result, the government has a special interest in supporting the banking sector.
Deposit insurance was introduced in 1933 to slow and stop bank runs (15). We’ve discussed depositor psychology before, but if a depositor thinks their bank might be in trouble, they are incentivized to pull their money out before anyone else does. If all depositors pull, the bank is in trouble. However, if the depositor’s funds are insured against loss, they have less reason to worry about their money and less incentive to withdraw.
For the most part, deposit insurance is quite effective (31). Most individuals hold less than the insurance limit of $250,000; in fact, 99% of deposit account are fully insured (1). However, the remaining 1% account for a huge amount of deposits. According to the report, ~49% of deposits are uninsured (12). Under the current setup, we have a relatively small amount of very large accounts contributing to large amounts of uninsured deposits in the system. This point will be important when we revisit the proposed options.
Deposit insurance is paid out by the FDIC’s Deposit Insurance Fund (DIF), and the DIF is funded by assessments on banks. Any of the FDIC’s proposed increases to deposit insurance will result in an increase in assessment fees. This is, and will be, extremely unpopular with banks.
2. The importance of deposit diversification
Only hinted at in the report, but deserving of more attention, is the importance of deposit diversification. Recall the banking business model: using short-term deposits to make long-term loans. This only works if all depositors don’t need all their money back at the same time. There needs to be diversification in the inflows and outflows, and there needs to be some diversification in depositors’ response to exogenous shocks.
SVB and FRC did not have diversified deposit bases. Their deposits were almost all uninsured, 94% and ~70% respectively, and concentrated in specific industries such as VC-backed companies and high net worth individuals living in a handful of places (7). This creates a large amount of risk.
Moving forward, deposit diversification will likely, and should, be highlighted by management teams and supervisors. I would not be surprised if there are explicit or implicit limits on the amount of uninsured deposits that institutions can hold; the report does hint at “simple limits on uninsured depositor funding for banks” (41).
3. FDIC’s focus on the “composition of deposits”
In the report, the FDIC very clearly delineates between deposits used for “payment” and deposits used for “investment”; the differentiation then becomes central to their preference for targeted insurance for payment accounts. It is clear that they have a preference for protecting payment accounts to support banks’ “essential role in the payments system” (59). They state that deposits used for investment have other alternatives, such as money market funds (MMFs).
While I agree that banks play an essential role in the payments system, banks also play an essential role for savings. Bank accounts play a vital role for consumers and commercial entities that need a cash product providing convenience, security, and yield. Importantly, banks and the banking sector cannot survive on payment / transactional deposits alone. From our business model discussion, it is clear that banks need sticky deposits in order to make loans. If we create a system where bank accounts are only used for payments and all additional funds are swept into MMFs, we’re in trouble.
Bank accounts should be safe. Otherwise, in times of stress, we see a flight to safety OUT of banks, which is the exact opposite of what we want. Indeed, we’ve seen funds flow out of the banking system en masse in recent months because MMFs, especially Treasury funds, seem to offer better yield AND better protection (until politicians start playing chicken with the American economy in a couple of weeks at least).
In our opinion, the FDIC should care about flows out of the banking sector because they destabilize banks and create risk. Indeed, whenever deposit insurance has been increased, such as in 1969, 1974, and 1980, deposits have flown in to banks (17, 18). In an odd turn of events, the FDIC’s report doesn’t seem to look on favorably on that, even stating that unlimited insurance “may also generate large inflows of deposit funding to banks” (3). Well we wouldn’t want that now would we.
Simply put, we believe the FDIC should move to protect savings accounts as well. The FDIC’s proposed changes don’t cover savings accounts, but our proposed changes do.
4. The role of technology
The report, like many others, highlights the role that technology played in the recent bank runs (27). Mobile banking, digital account opening, and modern money movement rails made it easy for depositors to quickly move large amounts of deposits. Obviously, the solution shouldn’t be “don’t let people use phones.”
The truth is that technology has fundamentally changed banking. Regulators, institutions, and providers should internalize it and take action. We’ve hinted at it before, but we believe technology will also play a central role in the solution. The adoption of new technology can create risk, but it is becoming clear that inaction will result in more risk.
Specifically, technology can help with 1) deposit monitoring and grading, and 2) new deposit products. On the deposit monitoring and grading side, better technology can highlight issues in deposit diversification, correlated exposures, and intraday flows. In general, using technology to understand the quality of deposits and deposit stability is one of the perfect applications of data science and machine learning to banking, and something that we and others have been hard at work on post-SVB.
On point 2, reciprocal products and sweep products that offer extended insurance from a network of receiving banks are a beautiful, tech-driven solution to the bank run problem. We expand on these ideas in section 6 below.
5. FDIC’s proposed options for increased deposit coverage
The FDIC outlines three proposed options for increased deposit insurance coverage: 1) increasing the current limit, 2) providing unlimited coverage, and 3) providing increased coverage on specific account types.
a. Limited coverage
The first, which they call limited coverage, is the closest setup to what we have today. However, instead of $250,000 per depositor per bank, the limit could be increased to $500,000 or $1 million (they are do not provide a specific limit in the report).
While this model of insurance is the most common and the best understood, it does not solve the problem: that a small number of large accounts are contributing to bank runs. Increasing the limit across the board would provide most accounts with more insurance than they need and large accounts with not enough insurance.
b. Unlimited coverage
The second option is unlimited coverage, where all deposits are insured. Frankly, it is the cleanest solution, which simplifies insurance calculation and failure resolution, and would likely stop most, if not all, bank runs.
However, it has two serious consequences. One, it would create serious inefficiencies in the sector, specifically around moral hazard. You would now have institutions that take risk on the asset side on their balance sheet but have little to no risk on their liabilities side. Management teams might be incentivized to take more risk, and depositors would not be incentivized to care about bank risk taking. Second, it would be expensive for the Deposit Insurance Fund and therefore for banks. The DIF would have to roughly double in size for this to work.
c. Targeted coverage
The final proposal, and the FDIC’s preferred approach is targeted coverage to specific account types. In the report, they suggest providing large insurance limits, and possibly unlimited coverage, to business payment accounts because of their important role in the economy.
Central to their argument is that this sort of targeted insurance existed before. From 2008 to 2012, the Transaction Account Guarantee (TAG) program provided unlimited insurance on non-interest bearing transaction accounts to stabilize the banking sector during the Global Financial Crisis (10). The program was well used, peaking at $1.5 trillion deposits in 2012, and likely contributed to stability (22).
The FDIC admits that it will be complex to specifically identify business payment accounts (57). If this approach is implemented, the FDIC will likely have to extend coverage to all non-interest bearing transaction accounts, like under TAG. However, for this to work, they’ll likely have to roll back rule changes to make transaction accounts more distinct from savings accounts. These high-coverage accounts will have to pay zero to little interest (reinstating Regulation Q), and savings accounts might once again need restrictions on monthly withdrawals (reinstating the six withdrawal rule to Regulation D). Otherwise, depositors could just keep all their funds in a fully-insured transaction account paying decent yield.
Our main hesitation with the approach is that it doesn’t seem to solve the bank run problem. If a depositor has their funds in an uninsured savings account and becomes worried about their bank, they will move their funds to the insured transaction account. However, now the depositor is 1) still worried about the bank, and 2) earning no interest. It feels like that depositor will still leave. Importantly, yields were zero from 2008 to 2012, so depositors could leave everything in unlimited insurance transaction accounts with little penalty.
In addition, sweep accounts, which are already popular, will become the norm where deposits are fully insured during the day at the bank and are swept into MMFs to earn yield at night.
So while the FDIC’s proposal makes sense and has precedence, we don’t believe it fully solves the problem.
6. Our proposal for increased deposit insurance (and the FDIC’s view of deposit networks)
We’ve learned time and time again over the past few years that it is easy to criticize and hard to create. So after poking holes for the past 1800 words, we propose our solution here.
Ideally, you want a system where depositors who care about extended insurance can get extended insurance. Because they care about it, they should pay for it. This is how almost every other insurance market works.
Reciprocal products and sweep products, which sweep deposits to other banks, more or less work this way.
Depositors who care about additional insurance can get extended insurance through these products. They opt in to the product, and they pay for it through small fees and yield impact. Importantly, these products tend to have an inverse relationship between the amount of insurance offered and the amount of yield, which is exactly how it should work. By allocating deposits across a network of banks, the products diversify exposure and reduce risk.
In a huge win for deposit networks, the FDIC seems to agree. They state, “Although larger, institutional depositors are better equipped than smaller depositors to perform due diligence, they may also use their resources to expand their deposit insurance coverage beyond the $250,000 limit by using deposit services such as brokered deposits, reciprocal deposits, and sweep accounts. Use of these products shows that there is a demand for deposit insurance protection at higher levels. Further, the presence of brokered deposits, sweeps, and reciprocal deposits demonstrates that the current system already provides deposit insurance coverage for large depositors” (32). In fact, one of the reasons the FDIC focuses on business payment accounts in their report is that, “Relative to investment accounts, business payment accounts … are more difficult to diversify across banks in the current system to obtain full deposit insurance” (49). Again, they mention, “It is likely that deposit accounts used for operational purposes are more difficult to maintain across multiple banks to obtain greater deposit insurance coverage” (57). Stopping short of a full endorsement, the above essentially says that folks should “diversify across banks in the current system to obtain full deposit insurance.”
So reciprocal products and sweep products can solve the deposit run problem without negative implications and are looked upon favorably by the FDIC. However, this raises the question of why didn’t they help SVB and FRC? Both had relationships with deposit networks, and FRC especially had been vocal about sweep products in recent weeks. (The management teams of those networks are subscribed to this newsletter). The simple fact is that the current products are clunky for depositors and banks. As a result, while popular and important, most folks had never heard of them until a few weeks ago, and most folks still don’t use them.
Specifically, we need better reciprocal and sweep products that integrate seamlessly into banks’ product suites, so that they feel just like regular deposit accounts to depositors. For large value depositors, these account types should become the default, not the exception. A strong focus on user experience, for both the depositor and the bank’s management team, is needed. If only there was a tech company working on that.