Duration mismatch regularly causes some banks to fail. Matching the duration of short-term assets and liabilities will eliminate failures.
In my previous wire (The Risk of Higher Rates the RBA’s Overlooking, 21 March), I noted that the failures of the crypto-friendly Silvergate Bank and the tech-centric Silicon Valley Bank (America’s 16th-largest), as well as the forced takeover of Credit Suisse (founded in 1856, and not long ago one of the world’s top-20 banks) by UBS, are hugely symbolic: they occurred almost 15 years to the day after the collapse of Bear Stearns, its bailout by the New York Fed and rescue by JPMorgan Chase. Objectively, SVB’s failure is the second-largest (measured by its assets but unadjusted by CPI) in U.S. history; collectively, this is by far the biggest cluster of bank failures and bailouts since the Global Financial Crisis.
These events raise three vital questions. First, why did they occur? Second, do they portend another GFC? Lastly, what can be learnt and what should be done in order to minimise the recurrence of future ructions and crises?
In this essay, I show that
- The typical diagnosis of these banks’ failures isn’t wrong. Yet it’s unsatisfactory because it emphasises proximate and mostly ignores ultimate causes.
- Identifying the ultimate cause of the failures reveals that what many fear – another GFC – is unlikely. But what they’re overlooking – a banking crisis like the one in the U.S. from the mid-1980s to the early-1990s – is quite possible.
- Indeed, if just-released academic research (which I summarise) is accurate, then it’s not merely SVB and something much worse is brewing: the entire American banking sector is now effectively insolvent.
- It’s just plain silly to attribute banking crises to mismanagement and poor regulation, and in their wake to demand better administration and stricter rules. Illiquidity and insolvency are embedded in contemporary banks’ DNA; no amount of regulation can alter their nature.
- Yet there’s a straightforward way – which entails NO change of legislation or regulation – that will virtually eliminate the probability that a bank fails.
- Unfortunately, the likelihood is practically zero that any bank uses these means to make itself impregnable. Like today’s universities, contemporary banking is a “public bad” that enriches an anointed few but spreads destructive ideas throughout society.
Historical Context
Figure 1 plots the total number of banks (not including branches) in the U.S. since 1864. (In that year, Congress gave to the federal government powers to supervise commercial banks, and for the first time created coherent national banking statistics.) I’ve drawn data for 1864-1934 from Historical Statistics of the United States and since 1934 from the Federal Deposit Insurance Corporation. Congress established the FDIC, as well as the Federal Savings & Loan Insurance Corporation (FSLIC), in that year in order to insure all qualifying deposits. According to its advocates, it would also greatly reduce – and perhaps even end – bank failures.
Figure 1: Total Number of Banks U.S., 1864-2022
Figure 1 tells us four salient things:
- From 1864 to 1920, the number of banks rose from almost 2,000 to more than 31,000. Underpinning this increase was the rise of population (from 34 to 108 million) and of Gross National Product (GNP) from ca. $6 to $70 billion.
- Most of these banks were, in terms of size and geographical scope, small and fragile. Partly for this reason, in the 1920s their number began to fall; and largely for this reason, during the Great Depression it plunged by almost one-half.
- For roughly 50 years after the mid-1930s, the number remained roughly stable.
- Since the mid-1980s and largely as a consequence of mergers and acquisitions, the number has more than halved.
Figure 2a plots the number of bank failures per year in the U.S. since 1864; Figure 2b plots the failures during each year as a percentage of that year’s total number of banks.
Figure 2a: Bank Failures per Year, 1864-2022
Before the Depression, failures were a regular occurrence. Before 1921, every year an average of 69 banks (comprising an average of 0.7% of the total number in each year) failed. Between 1921 and 1929, the average swelled to 625 (2.25%). From 1930 to 1933, 9,106 banks – almost 36% of the number at the end of 1929 – failed. In 1933, more than 4,000 (27% of the total in that year) failed.
From 1934 to 1941, the average number of failures per year plunged to 43 (0.3% of the average number of banks during this interval). And from 1942 to 1979, the average number of failures fell to just 5 per year (0.04%).
Failures flared from 1980 to 1995 (average of 169 and 1.3%). At the peak in 1988, 470 banks (3.6%) failed. The GFC triggered a less severe and shorter burst (96 per year in 2008-2012, a loss of 1.5% per year of the average number of banks during that interval).
Figure 2b: Bank Failures as a Percentage of Total Banks, Annual, 1864-2022
Figures 2a-2b indicate that the key question isn’t merely “why do banks fail?” It’s equally important to ask “why, since the establishment of the FDIC – one of whose purposes is to minimise bank failures – have failures clustered during the years 1980-1994 and the GFC?” What key common traits do today’s banks – and their regulators – share with their predecessors?
The Cost of Bank Failures
Figure 3a plots losses to depositors since valid, reliable and comparable figures became available in 1921. Before 1934, as the Social Security Administration puts it, “when a bank failed, the depositors were simply left without a penny. The life savings of millions of Americans were wiped out by the bank failures (of the Great Depression).”
Figure 3a: Losses from Bank Failures, Billions of CPI-Adjusted $US, 1921-2022
In diametric contrast, since the advent of deposit insurance in 1934, the FDIC and FSLIC have made good any losses suffered by insured depositors (which is practically all individuals and small businesses). Importantly, the losses in Figure 3a exclude those incurred by owners of (a) uninsured amounts (that is, accounts whose balances exceeded the insured maximum) and (b) failed banks’ bonds and stocks; they also exclude (c) losses to the economy as a whole from reduced lending, output, etc.
In 1930-1933, depositors’ losses totalled approximately $14 billion; adjusted for CPI, that’s $310 billion today. In 1933, the largest loss ($7.6 billion or CPI-adjusted $171 billion) occurred. That’s more than the CPI-adjusted loss of $129 billion in 1988, yet the total loss in 1980-1995 ($448 billion) exceeds its counterparts in 1930-1939 ($311 billion) and 2008-2012 ($94 billion).
It’s essential to distinguish the completely-uninsured losses to depositors before 1934 and the mostly-insured ones since then. Uninsured deposits simply disappear; because they can’t be spent, they don’t generate further economic activity. Ignoring many complexities, $1 of uninsured deposits generally caused a decrease of at least $1 of GNP. Owners of insured deposits in failed banks, on the other hand, are made whole and to that extent their spending is generally unaffected.
The cost of each $1 of uninsured deposits is therefore far costlier to depositors than its insured counterpart. That’s why the losses of uninsured deposits in 1930-1933 contributed to an enormous shrinkage of GNP, whereas the larger (in CPI-adjusted dollars) losses in 1985-1995 affected GNP far less – and the losses during the GFC affected it hardly at all (Figure 3b).
Figure 3b: Losses to Depositors Arising from Bank Failures, Percentage of GNP, 1921-2022
Clearly, the key question isn’t “Do the recent bank failures in the U.S. portend another GFC?” The crucial but unasked question becomes: “is an extended crisis like the one from the mid-1980s to the mid-1990s (usually but inaccurately dubbed “the Savings & Loan (S&L) Crisis”) in the offing?
What Caused the S&L Crisis?
“In the 1980s,” writes Kenneth Robinson of the Federal Reserve Bank of Dallas on federalreservehistory.org, “the financial sector suffered through a period of distress that was focused on (but not restricted to) the nation’s savings and loan industry. Inflation rates and interest rates both rose dramatically in the late 1970s and early 1980s.” Does that sound familiar?
“This produced two problems … First, the interest rates that S&Ls could pay on deposits were set by the federal government and were substantially below what could be earned elsewhere, leading savers to withdraw their funds. Second, S&Ls primarily made long-term fixed-rate mortgages. When interest rates rose, these mortgages lost a considerable amount of value, which essentially wiped out the S&L industry’s net worth.”
From 1986 to 1989, the FSLIC closed or forced the absorption of 296 S&Ls. Despite repeated injections of billions of dollars of taxpayers’ funds into it as a result of the S&L Crisis, Congress declared it bankrupt and abolished it in 1989. From that year, the newly-created Resolution Trust Corporation paid billions of dollars to insured depositors. From 1989 to 1995, it closed or forced the takeover of 747 S&Ls whose estimated book value was $402-$407 billion (ca. $845 billion in today’s $US).
Proximate versus Ultimate Causes
We’re now ready to answer the question “Why do banks fail?” Let’s use Silicon Valley Bank as an example. Summarising a recent contribution to Livewire, two factors collectively sealed its demise:
- Concentration of Deposits: SVB derived its deposits from effectively one industry and to some extent, geographic area, namely the Silicon Valley tech sector.
- No Hedging: SVB failed even to try to equalise (1) the very long-term interest rate risk on its ca. $212 billion asset (including loans) portfolio and (2) the very short-term interest rate risk on its ca. $173 billion deposit (or liability) book.
Neither of these points is incorrect; indeed, both are right. The problem is that they’re proximate rather than ultimate causes. A proximate cause is one which is temporally closest to, or immediately responsible for causing, some observed result.
The proximate cause is the final link in the chain of causality; the ultimate cause is the first one. The proximate cause is the straw that broke the camel’s back; the ultimate cause identifies who overloaded it – and explains why.
The Ultimate Cause of Bank Failures and Crises
This is a mammoth subject. Indeed, more than a decade ago I wrote a book about it called The Evil Princes of Martin Place. It explains, deductively from first principles, the inherent rottenness of banking (including central banking). It also applies these principles to the Panic of 1907, the Great Depression and the GFC. It’s available on Amazon.com.au, so if you want a comprehensive answer then read it.
For our purposes, a cursory and overly-simplified explanation will suffice. It’s sometimes mentioned in passing, but it’s rarely emphasised adequately: contemporary banks – including those which conform in all respects to legislation and regulation like Silicon Valley Bank did on 31 December – are EXTREMELY heavily leveraged. As a result, under certain conditions small transactions can have huge consequences upon their liquidity and solvency.
Satyajit Das (“Is a Full-Blown Global Banking Meltdown in the Offing?” 23 March) agrees: “banking is essentially a confidence trick because of the inherent mismatch between short-term deposits and longer-term assets … The essential structure of the banking is unstable, primarily because of its high leverage.”
To illustrate this crucial point, Table 1 summarises SVB’s balance sheet on 31 December. Two sets of points are noteworthy.
Table 1: Silicon Valley Bank’s Balance Sheet (Billions of $US), 31 December 2022
Death by Illiquidity
Like virtually all banks’ assets, SVB’s were extremely illiquid: of its total assets of $211.8 billion, only 19% were liquid (that is, in the form of cash or in a form rapidly convertible into cash). And that’s assuming that its securities held for sale could be sold immediately. If not, then SVB’s liquid (cash) assets comprised less than 7% of its total assets. For many businesses that mightn’t be a problem. But no business other than a bank holds such a large percentage of current liabilities relative to current assets.
But don’t banks have to maintain reserves? Not in the U.S.: on 15 March 2020, the Federal Reserve “reduced reserve requirement ratios to zero percent effective March 26, 2020. This action eliminated reserve requirements for all depository institutions.”
On 31 December SVB’s demand deposits of $173.1 billion vastly (by more than 12-fold) outstripped its cash assets ($13.8 billion) instantly available to redeem these liabilities. Normally that’s not a problem: rarely do more than a small percentage of depositors withdraw a significant percentage (never mind most or all) of their deposits. But when they do – that is, when a bank “run” occurs – the scene resembles the 1946 Christmas movie It’s a Wonderful Life.
At all times, banks (including banks fully compliant with legislation and regulations, which is virtually all banks) are highly illiquid. And on rare occasions, “runs” can render them unable to redeem their demand deposits. That’s death by illiquidity.
For today’s banks, including the allegedly impregnable big banks in Australia and Canada, this problem is inherent. “Maturity mismatch” (also known as “duration mismatch”) refers to the disparity between any entity’s (not necessarily a bank’s) short-term assets and short-term liabilities: the more of the latter relative to the former, the greater is the mismatch – and the potential threat to liquidity.
Banks add a specific twist to the general problem: they use short-term liabilities (deposits) to finance their long-term assets (investments and loans). Indeed, banks’ net interest margin – a key source of their profit – presupposes borrowing short-term from depositors (and paying them a relatively low rate of interest), and lending long-term to mortgagees, etc., at a higher rate.
Death by Insolvency
Banks aren’t merely at risk of quick death by illiquidity. They’re also prone to sudden demise by insolvency.
On 31 December, SVB’s $211.8 billion of assets were poised precariously upon a sliver – just $16.3 billion – of equity. If the value of its total assets declined by just 7.7% (i.e., 16.3 ÷ 211.8= 0.077) then its liabilities would exceed its assets and it’d be insolvent.
The problem was that on 31 December the market value of its $91.2 billion of securities it was holding to maturity was just $76.2 billion. The value of fixed-interest securities varies inversely with interest rates; since the middle of last year rates have been rising; hence these securities’ market value has been falling.
Remove $91.3 billion from its asset column, replace it with $76.2 billion, and (assuming no other changes, e.g., assume very optimistically that it loses not a single penny on its portfolio of loans and depositors don’t withdraw a penny of their deposits) total assets sum to just $196.7 billion – and equity collapses to just $1.2 billion.
That’s a ratio of assets to equity of 196.7 ÷ 1.2 = 164! If depositors demand more than a mere 0.7% of their deposits (1.2 ÷ 173.1 = 0.007), which they did in early March, then SVB’s liabilities would exceed its assets and it’d be bankrupt.
Why and When Do Major Clusters of Bank Failures Occur?
Banks fail because they’re extremely illiquid and heavily leveraged. That explains why a handful of relatively small and unlucky banks fail, on average, in the U.S. each year. But it doesn’t explain spasms of failures such as the ones that occurred during the Great Depression, S&L Crisis and GFC. This, too, is a mammoth subject, and The Evil Princes of Martin Place analyses it comprehensively. For our purposes, a brief summary of Charles Kindleberger’s book, Manias Panics Crashes: A History of Financial Crises (John Wiley & Sons, 5th rev. ed, 2005), suffices to answer this crucial question.
“When you consider (the market’s) cycles of mania, panic and crashes,” a review of the book notes, “you will notice that there are some typical factors that usually lead to the formulation of a mania and, sometime later, an inevitable crash. In such a scenario, the foremost factors have (included) … an increase in (bank) credit.” In response to technological or other developments, people become exuberant and bankers participate in the exuberance by writing dud loans – or what they initially regard as good risks but subsequently learn are bad ones.
Bankers’ extension of credit generates “an increase in stock and real estate prices, leading to further (lending) … Most asset bubbles are a monetary phenomenon and result from the rapid growth in the supply of credit.”
“Though market manias lead to a dizzying rise in prices across asset classes,” the review continues, “the party stops when creditors start getting worried about the ability of debtors to pay back the loans that they have taken to invest in the rising market.” Lenders, in other words, come to their senses as significant numbers of their loans fall into arrears. “Such concerns indicate the beginning of panic mode, which prompts creditors to stop issuing new loans. In such a situation, the debtors who were relying on additional loans to cover their interest payments, face bankruptcy …”
And as a rising tide of debtors become insolvent, so do their lenders; hence the generational clusters and spasms of bank runs and failures. Occasionally a bank-specific event causes a bank to collapse. But often the same factor causes multiple banks to fail. In particular, notes Kindleberger, ”most of the bank failures in the 1980s and 1990s were systemic events that involved a large number of banks in a country, and in many episodes virtually all of a country’s banks.”
Using data collated by the Federal Reserve Bank of St Louis, Figure 4 plots the face value of American banks’ non-performing loans (which are 90 or more days in arrears) as a percentage of Gross Domestic Product (GDP). This is the longest series of valid and reliable data I’ve been able to locate; unfortunately, the Fed discontinued it in 2020.
It doesn’t merely substantiate Kindleberger’s thesis: it also indicates that in this key respect the S&L Crisis was more severe – and lasted much longer – than the GFC.
Figure 4: American Banks’ Non-Performing Loans as a Percentage of GDP, Quarterly, 1984Q1-2020Q1
How to Prevent Bank Failures and Banking Crises
If duration mismatch regularly causes a few banks to fail (and much less frequently brings a given country’s banking system to its knees), and if we wish realistically to minimise (and ideally to eliminate) failures, then the solution is obvious: match the duration of short-term assets and liabilities. Table 2 provides an example.
Table 2: an Impregnable Deposit-Taker’s Balance Sheet
In this example, the deposit-taking institution holds $15 billion of cash (an amount equal to its shareholders’ equity), invests all of its remaining assets in one-month Treasury bills (or its nearest Australian equivalent, etc.) makes no loans and doesn’t borrow. Short-term assets (cash and Treasury bills of $100 billion) exceed short-term liabilities (demand deposits of $85 billion). If depositors seek to withdraw any amount up to $15 billion, cash finances their withdrawals; if they suddenly demand more than $15 billion, then the institution sells the required quantity of T-bills.
Regardless of the amount demanded for redemption, depositors receive $1 cash per $1 of deposits. The risk of a “run” – and, more generally, of insolvency, assuming that the U.S. (or Commonwealth Government, etc., as appropriate) doesn’t default – is effectively zero.
In order to finance its expenses, this bank’s rate of interest on deposits would be less than the short-term Treasury rate. And although Table 2 doesn’t show it, this bank could also issue term deposits at discounts to corresponding Treasury yields.
This simple example – which any existing American, Australian, Canadian, etc., startup could emulate – resembles “full-reserve” banking (also known as “100% reserve banking” and “narrow banking”), but is even more conservative: a full-reserve bank doesn’t lend demand deposits; instead, it finances loans only from time deposits. The bank in Table 2 is even more conservative because it doesn’t lend. For this reason, it’s much less profitable (in the short-term) than a typical bank.
The conservative depositor’s bank doesn’t lend because lending is inherently risky. Contemporary banks obscure this risk. Decades typically pass from one banking crisis to the next. These long stretches don’t mean that lending is usually riskless: it means that, like detritus on the forest floor, risks build steadily over time – until some spark suddenly ignites them.
Although it’s far safer than the typical bank, the non-lending bank nonetheless faces three risks: (1) that short-term government securities default); (2) that the volume of one-month government securities bought and sold falls materially); and (3) that their yield rises and hence their price falls). Technically, these risks aren’t zero; realistically (and particularly if the bank hedges its interest rate risk), they’re as close to zero as the real world gets.
It’s important to emphasise: the non-lending bank writes neither commercial nor consumer or mortgage loans. Where would people go if they needed a credit card, car loan or mortgage? To an existing – that is, lending – bank. Who would buy the equity of the institution described in Table 2? People, businesses and other entities that don’t need a high dividend yield or total return but seek to place large deposits and thus place high value upon the maximum possible safety of their cash. Why would anybody deposit their cash in this bank? Doesn’t the government guarantee deposits? Don’t the events of the past month underscore the unarguable success of deposit insurance?
Deposit Insurance Amplifies Rather than Abates Systemic Risk
Congress established the FDIC and FSLIC in 1934. They initially insured 100% of the first $2,500 of each depositor’s account (increased to $5,000 in mid-1934 and, in a number of steps, to $100,000 in 1980 and $250,000 today). A flat annual levy upon deposit-taking institutions, equivalent to 0.5% of each’s deposits, finances the insurance.
Deposit insurance’s advocates contend that it provides safety for small depositors and protects the smooth operation of the payments (cheque-clearing) system. Most importantly, it mitigates or even eliminates bank runs and failures. Indeed, for 50 years after their establishment, runs virtually disappeared, rarely did more than ten banks fail per year (recall Figure 1) and costs to the FDIC’s and FSLIC’s insurance funds were negligible (recall Figure 2). The logic and evidence thus seemed unassailable: regardless of a bank’s solvency, the government promises to make depositors whole; so why should they bother to remove their deposits (“run”) at the first sign of possible trouble?
For years, opponents contended that this alleged advantage is actually an Achilles’ heel: precisely because insurance weakens the incentive for depositors to care about the risks their banks and S&Ls might take, they gradually take ever more – and eventually take excessive – risks. Insured depositors may sleep more soundly in the short-term, but the edifice becomes ever more unstable.
The S&L crisis vindicated the critics. George Kaufman’s article (“Deposit Insurance”) in the Library of Economics and Liberty chronicles its shortcomings. Among its more bizarre – indeed, perverse – attributes: during the 1970s and 1980s, as the crisis grew, the FDIC and FSLIC were slow to close insolvent institutions. As a result, and in an effort to attract depositors and restore their solvency, those that traded whilst insolvent often increased the rates of interest they paid to depositors – and thereby helped to create “runs” from solvent to insolvent banks that worsened the crisis!
Anat Admati, a professor at Stanford and one of Time magazine’s 100 Most Influential People in the World (2014), agrees. In The Weekend Australian (25-26 March) she stated that the events of the past month evoke the S&L Crisis. At that time, regulators allowed “many insolvent institutions to persist. Everyone pretended they were OK … but meanwhile they were incredibly reckless and ultimately collapsed and caused much harm … Unlimited deposit insurance (which the Biden Administration is seemingly offering, albeit without the necessary approval of Congress) is a very dangerous situation; for bank managers, it can pay to be negligent under the current regime.”
Implications
Contemporary banking inevitably and repeatedly presents to governments two bad choices: (1) let banks such as SVB fail and require big depositors to accept haircuts on uninsured deposits, and risk a nationwide bank run and economic calamity, or (2) rescue illiquid and/or insolvent banks, extend the deposit guarantee – and ensure an even bigger crisis down the road. A decade ago, The Evil Princes of Martin Place and other analyses that inspired it were well outside the mainstream. Today, acceptable opinion – as exemplified by The Wall Street Journal – has moved towards them. Each banking crisis, it seems, narrows the gap. But in one crucial respect a chasm remains.
It’d Be Comical If the Consequences Weren’t So Serious …
On 10 October 2022, The Royal Swedish Academy of Sciences announced that it had awarded the Sveriges Riksbank Prize in Economic Sciences in Memory of Alfred Nobel (commonly but erroneously called “the Nobel Prize in Economics”) to three economists, including Ben Bernanke, “for research on banks and financial crises.”
Banks, the Academy’s press release correctly noted, are “vulnerable to rumours about their imminent collapse. If a large number of savers simultaneously run to the bank to withdraw their money, the rumour may become a self-fulfilling prophecy – a bank run occurs and the bank collapses.” The press release then lapsed into obvious (based upon historical evidence) falsehood: “These dangerous dynamics can be prevented through the government providing deposit insurance and acting as a lender of last resort to banks.” From this false premise, the Academy reckoned that these three economists’ “discoveries” have “improved how society deals with financial crises.” In particular, “the laureates’ insights have improved our ability to avoid both serious (banking) crises and expensive bailouts.”
What a hoot! Since the beginning of this month, the Federal Reserve which Bernanke once headed has backstopped more than $600 billion of American banks’ mark-to-market losses; it’s also guaranteed all of SVB’s ($175 billion) and Signature Bank’s ($89 billion) deposits; moreover, Switzerland’s central bank is providing $100 billion of liquidity to smooth UBS’s forced takeover of Credit Suisse.
“The banking system is safe,” Joe Biden assured Americans on 13 March. Indeed, things are so liquid and the system apparently so strong that it’s taken almost $1 trillion of emergency liquidity to keep the system liquid. Maybe Bernanke will cringe in shame and return his gong? Perhaps the Royal Swedish Academy of Sciences will change its mind and rescind its award? Pigs will fly!
… And Today’s Crisis Is Very Serious
America’s banking system is facing not just a crisis of liquidity but also one of solvency. Economists at the University of Southern California, Northwestern, Columbia and Stanford, in a research paper entitled “Monetary Tightening and U.S. Bank Fragility in 2023: Mark-to-Market Losses and Uninsured Depositor Runs?” and dated 13 March, quantify American banks’ exposure to recent increases of bonds’ yields.
They find that “the U.S. banking system’s market value of assets is $2 trillion lower than suggested by their book value of assets … Even if only half of uninsured depositors decide to withdraw (their cash), almost 190 banks are at a potential risk of impairment to insured depositors, with potentially $300 billion of insured deposits at risk.” The researchers also note that “the recently failed SVB does not stand out (much) … About 11% of banks suffered worse mark-to-market losses … In other words, if SVB failed …, more than 500 other banks should also have failed.” If so, that’s the most since the S&L Crisis.
But in one crucial sense, today’s banks’ condition is the worst since the Great Depression. “These ($2 trillion of MTM) losses amount to a stunning 96% of pre-tightening aggregate bank capitalisation.” In other words, and as Adam Creighton noted in The Weekend Australian (25-6 March), “almost all of the capital in the U.S. bank system, based on mark-to-market valuations, has been wiped out.” If so, then it’s hardly just SVB: at present the entire American banking sector is effectively insolvent.
The Post-2008 Regulatory Regime Has Failed
Most people who occupy seats of economic, financial and political power typically attribute banking crises to mismanagement and insufficient regulation, and in the wake of each crisis robotically demand “better management” and “stricter rules.” Such comments demonstrate how little they know – or, at least, how little they choose to disclose.
They fail to comprehend (or admit) that crisis and failure is embedded in the DNA of today’s banks. No amount of regulatory meddling can alter their nature: only a fundamental change of what constitutes deposit banking can.
“One of the great conceits of the past 15 years,” acknowledged WSJ on 16 March, “has been that … regulations would keep banks safe.” Whether on the eve of the GFC (Washington Mutual) or today (Credit Suisse), major banks’ capital and liquidity buffers have met regulatory minima – but some still fail or require bailouts.
“This weekend’s rescue (of Credit Suisse),” WSJ added on 20 March, “is a warning that two weeks into the current banking panic the post-2008 rulebook has already failed. Taxpayers are on notice that the solution to any crisis will be to amplify too-big-to-fail rather than (reduce) it – as it was the last time around (in 2008-2009). Hang onto your wallets.”
Michael Faulkender (an economist at the University of Maryland and a former assistant Secretary of the Treasury) and Tyler Goodspeed (of the Hoover Institution at Stanford University and a former chair of the White House Council of Economic Advisers) agree. “Efforts to make the U.S. banking system less risky have had the opposite effect,” they write in “Want to Prevent SVCB-Style Collapses? Scrap Dodd Frank” on 21 March. “The stress testing mandated by the Dodd-Frank Act led banks to diversify in the same way, which elevated systemic risk even as individual banks became less risky …”
“This means that if something brings down one major bank, others are more likely to fall, snowballing into a major financial system collapse … Even a small error in government rules … will be multiplied across the entire financial architecture … SVB’s fall shows that this is a real possibility.”
Deposit Insurance Doesn’t Abate – It Boosts – Systemic Risk
Before the GFC, this contention was controversial; today it’s becoming commonplace. Nor is it contentious – indeed, it’s unarguable – that the Fed’s recent chairs, including Janet Yellen, kept rates far too low way too long. As a result, as rates have risen SVB is hardly the only one that’s taken hefty MTM losses on its portfolio of bonds. Consequently, said Yellen on 16 March, the Treasury (which she heads), the Fed and the FDIC will “act in a manner that fully protects all depositors” – including those whose balances of more than $250,000 hadn’t previously been insured.
“This DC trio of heroes,” WSJ warned on 19 March, is “effectively guaranteeing all bank deposits … This creates a classic moral hazard, encouraging bankers to let rip – no need for a chief risk officer – and then maximise profits because the federales will step in later. The only thing this guarantees is future bailouts.”
On 22 March it added: “Financial regulators have ignored the post-2008 rule book to contain the latest banking panic. And Treasury Secretary Janet Yellen tore it up by announcing a de facto guarantee of all $17.6 trillion in U.S. bank deposits … It’s important to understand what this moment means: the end of market discipline in U.S. banking.”
What Causes Clusters of Bank Failures?
“This is becoming a global reckoning for years of policy illusions and financial excess,” WSJ editorialised on 16 March. “The biggest illusion is that there is a monetary and fiscal free lunch … The fiscal-monetary gusher distorted investment decision around the world – from Silicon Valley app developers to Italian government bonds to Chinese real estate. Banks, like fish, have absorbed the pollutants in the water around them. Little surprise that crises develop on bank balance sheets …”
“SBV was the lender of choice for tech dreamers,” Kimberley Strassel added (“Did ESG Help Sink SVB?” 16 March). “It claims to have banked nearly half of all venture-backed tech and healthcare start-ups. Yet in recent years those clients have skewed even more in one direction. ‘We serve those creating positive environmental change,’ SVB’s website brags, noting that the bank worked with some 1,550 companies in the ‘climate technology and sustainability sector.’”
Strassel is charitable: SVB offered “banking services to start-ups that often weren’t profitable, in some cases didn’t have a product, and would otherwise have a hard time getting a line of credit or a loan from a larger bank.” She also cites an entrepreneur’s scathing description: SVB’s products are “basically sub-prime business loans.” She asks: “what inspires a bank to disregard risk and shower money on products and services that nobody is clamouring to buy? One answer is easy money and misguided regulation … The other (is) Washington handouts, via President Biden’s effort to engineer a climate industry that otherwise wouldn’t exist.”
What’s been true historically continues to apply today: in response to technological developments, crazes, etc., people become irrationally exuberant – and bankers participate by writing dud loans.
Banks Won’t Become Impregnable and Depositors Secure
“The inference from … the number and severity of banking crises since the mid-1960s,” Charles Kindleberger concluded, “is that the lessons of history have been forgotten or slighted. These decades have been the most tumultuous in international monetary history in terms of the number, scope, and severity of financial crises.”
Yet as I’ve shown, there’s a straightforward way – without ANY change of legislation or regulation – virtually to eliminate the probability of a crisis like the one that shuttered SVB. James Mackintosh (“There Is a Cost to Moral Hazard,” 17 March) contends that “the second (way) to remake the financial system (the first is to let insolvent and “non-systemic” banks like SVB go to the wall) is “to eliminate the risk of bank runs.” Mackintosh’s key idea resembles mine: the creation of an exclusively deposit-taking (and not lending) institution.
Deposits, proposes Mackintosh, “would be held with ‘narrow banks’ that look very (much) like a … money-market fund, holding nothing riskier than short-term T Bills and making their money from transaction fees, not borrowing short and lending long. The rest of the (deposit) banking system would be replaced by funds that take only term deposits, matching the maturity of deposits with loans.”
This arrangement, he notes, “would have significantly less flexibility but would need far fewer regulators and no deposit insurance. This might have been possible in 2009 when the entire system failed, but concern about banks isn’t (sufficient) for such an overhaul now. Instead, we will get compromise, dither and cost. More deposit insurance … means even more regulation … and higher cost. It will probably prevent more bank runs for a while … But in the long run, we can expect complacency (and therefore recklessness) to take hold again because it always (has).”
The likelihood that any financial institution renders itself impregnable is, for all practical purposes, zero. Why on earth should it? The existing regime allows it – indeed, encourages it – to privatise its profits and socialise its losses. It’s a case of “heads, banks win; tails, taxpayers lose.”
Start-up banks won’t organise themselves as 100% reserve banks, existing banks won’t reorganize themselves along these lines, and politicians and regulators will emphatically reject any proposal to do so. They won’t merely succumb to heavy pressure from banks. As Satyajit Das puts it,
“Since the 1980s, the economic system has become addicted to borrowing-funded consumption and investment. Bank credit is central to this process. Some recommendations propose a drastic reduction in bank leverage from the current 10-to-1 (actually, it’s can be 20-1 or even more) to a mere 3-to1. The resulting contraction would have serious implications for economic activity and asset values.”
Like today’s universities, contemporary banking is a seemingly ineradicable “public bad” that spreads immensely destructive ideas throughout society. Indeed, like lithium-ion batteries, today’s banks are highly combustible, their conflagrations are extremely difficult to extinguish and they spew toxins far and wide. And nothing consequential will be done about it. It’s possible that a banking crisis has commenced. Perhaps it’ll be bigger than the GFC and as big as the S&L Crisis; it certainly won’t be the last.
max says
Honest Money by Gary North:
“Fractional reserve banking violates the Biblical principle against multiple indebtedness. When bankers violate this law (with the consent of the State), it leads to inflation and economic booms, followed by deflation and economic depressions. Fractional reserve banking is a form of fraud, as surely as a borrower who uses one piece of collateral to get a dozen loans is fraudulent.”
Roderick Russell says
Satya — a Sanskrit word meaning true; rather like the Mandarin word Fang.
Fault here lies with the short – termism of the political class.
‘The Long View ‘ by Fisher quotes Jean- Claude Junker ,” We all know what to do. But just don’t know how to get re-elected when we do it.”
Get into Romanian cigarettes, produced by the China International Tobacco Company. They have a 1000% margin in Denmark. Illegal but honest.