Welcome to another issue of Net Interest, my newsletter on financial sector themes. This is zombie week … read on for an explanation! For paying subscribers there are also additional segments on Deutsche Bank, European bank earnings and UK challenger banks. If you’re not already signed up, and want to support my work, you can do so here:
“Slow and steady in their approach, weak, clumsy, often absurd, the zombie relentlessly closes in, unstoppable, intractable.” — Simon Pegg.
In October 1968, Night of the Living Dead opened on screens across America. Zombies had been depicted in movies before, as versions of mythological creatures described in Haitian folklore. But Night of the Living Dead redefined the genre. Director George Romero breathed life into his reanimated corpses, infusing them with the shambling, flesh-eating mores that characterise zombies today.
Made on a budget of just $114,000, Romero’s film was low key. Cast and crew were recruited from among the production team’s family and friends, and expenses were spared on props: Chocolate syrup was used for blood and consumed flesh consisted of roasted ham – donated by one of the actors who also owned a chain of butcher shops.
In spite of the critical response that railed against its graphic violence, the film was a major hit. In the ten years following its release, it returned over 250 times its budget at the box office. It also paved the way for a new line in horror movies. According to critics, slasher films of the 1970s and 80s such as Halloween and A Nightmare of Elm Street, owe much to the Night of the Living Dead. And while the zombie genre waned for some years after Michael Jackson’s Thriller music video, it re-emerged in film, books and video games to cement its place in popular culture.
As a metaphor, the zombie also proved deeply alluring.
At the time Romero was working on his second sequel, Day of the Dead, America’s savings-and-loan (S&L) companies were in a full-blown crisis. Higher inflation and interest rates through the late 1970s and early 1980s squeezed their profitability. Limited by what rates they were allowed to pay on deposits, savings-and-loan associations saw deposit balances shrink at the same time as fixed rate mortgages lost value against the tide of rising rates.
Federal regulators didn’t spot the problem emerging and by the time they did, they lacked sufficient resources to deal with the losses. In 1983, it was estimated that it would cost roughly $25 billion to pay off the insured depositors of failed institutions yet the savings-and-loan insurance fund, known as the FSLIC, only had reserves of $6 billion. Rather than address the issue head-on, policymakers first took steps to deregulate the industry in the hope it could grow out of its problems, and then adopted a policy of forbearance to give it time to recover.
The result was what Edward Kane, Professor of Banking and Monetary Economics at Ohio State University, identified as a plague of “zombie” S&Ls:
Zombie S&Ls are institutional corpses capable of financial locomotion and various forms of malefic behavior. Much as in George Romero movies, these undead entities wreak havoc in financial markets by feeding on the markets of the living.
Kane defined zombie S&Ls as those with negative economic capital. According to the US Government Accountability Office, there were 459 S&Ls holding more than $100 billion in assets with negative net worth in September 1985. Kane estimated that including losses not shown under generally accepted accounting principles, the number of zombies could be higher than 600, equivalent to around a fifth of FSLIC-insured institutions at the time. Although insolvent, these financial institutions were able to continue operating thanks to explicit or implicit support from the government.
Edward Kane died on March 2 this year, a week before the collapse of Silicon Valley Bank, so he didn’t get to see the rise of a new crop of zombies.
But a look at First Republic Bank shows what one looks like. The bank had $174 billion of deposits the day before Silicon Valley Bank failed and went on to lose over $100 billion of them in the weeks that followed. Not able to liquidate assets as quickly, it turned to other funding sources for help:
It borrowed $30 billion from a consortium of 11 banks as a term deposit with an initial term of 120 days at market rates.
It borrowed $63.5 billion from the Federal Reserve discount window, at a rate of 4.85%.
It borrowed $13.8 billion from a new Federal Reserve Bank Term Funding Program launched in the aftermath of Silicon Valley Bank’s collapse, allowing it to borrow against the face value of its securities holdings (regardless of their market value).
It upped its borrowings from “lender of second-to-last resort”, the Federal Home Loan Bank of San Francisco, from $14 billion to $28.1 billion. Bit of adverse selection here: As well as First Republic Bank, the Federal Home Loan Bank’s largest borrowers as at end 2022 included Silicon Valley Bank and Silvergate. But the Home Loan Bank makes funding freely available, as long as a bank has assets to secure against it, albeit at a high cost.
By the end of the quarter, 45% of First Republic’s balance sheet was funded via government-sponsored schemes, up from 7% at the end of 2022.
The problem is that these funding sources are more expensive than First Republic’s core deposit funding. On average over the course of the first quarter, the bank paid 1.4% on deposits; its government-sponsored funding sources cost between 4.8% and 5.0%. Given that First Republic Bank only earns 3.7% on its assets, the shift in funding mix turns its business model upside down. Paying more for funding that you are yielding on assets is not exactly a recipe for business success.
The low yield on First Republic’s assets is also reflected in their market value. Like Silicon Valley Bank, First Republic acquired a lot of held-to-maturity securities, which now constitute around an eighth of its balance sheet. As interest rates rose, these securities fell into an unrealised loss, amounting to $4.8 billion at the end of 2022.
In addition, the bank made a lot of fixed rate mortgage loans, which react the same way to higher rates. At the end of 2022, two-thirds of its loan portfolio wasn’t due to reprice for at least another five years – a very high share compared with other banks. At the end of last year, First Republic estimated the mark-to-market loss on its loan portfolio to be $22.2 billion. Those losses may now be smaller as interest rates have come back in since the end of the year, but their sum is still likely more than the tangible common equity that the bank has available ($14.2 billion at at end March) rendering First Republic technically insolvent even if that is not reflected in accounting numbers. Under normal circumstances, the value of the deposit base would have risen to compensate, as higher rates make low-cost deposits more valuable, but…they’re not there anymore.
Edward Kane identified the problem with zombie banks as two-fold. First, as long as they roam the earth, zombies risk accumulating more losses. Second, by bidding up funding costs, they infect other institutions. The impact they have on funding costs can either be direct, as they force up the price all banks have to pay for deposits, or indirect, through higher deposit insurance assessments on the entire industry.
Many savings and loans companies took excessive risks as they “gambled for resurrection” but this is not the only response available to zombies. An alternative is to hoard cash, do little new business and wait for profits slowly to pay down the losses. That was the approach taken by Japanese banks in the years after the S&L crisis. In a paper in 1993, Kane warned the Japanese of the risks incurred in creating zombies. Unfortunately, they didn’t pay heed.
Before its crisis in the 1990s, Japan had a tried-and-tested formula for resolving bank failures: Authorities would orchestrate their acquisition by a healthy bank. When the Japanese bubble burst in the early 1990s, the scale of banking problems overwhelmed the formula but nor could the deposit insurance scheme step in: It was legally constrained in how much support it could provide in a single case of financial assistance. Instead, a hougachou approach was developed – a reference to a traditional festival for raising money from the community – that combined private and public sector support. Private sector institutions were invited to make capital contributions alongside liquidity and capital assistance from the Bank of Japan.
But the hougachou approach didn’t address underlying non-performing loan problems at the banks, and relying on other often weakened private banks led to consistently under-powered responses. In 1998, Long Term Credit Bank of Japan failed after having participated in the bailout of Nippon Credit Bank the previous year. By then, much of the sector comprised zombie banks, muddling along with weak balance sheets and insufficient capital to restructure and undertake new business.
Over the next five years, the authorities moved to resolve the crisis, initially by providing capital to the zombie banks, then by nationalising and shutting down banks. They also formed a resolution trust to take over banks’ bad assets. When it was done, banks had written off around $1 trillion in bad assets at a cost to the government of over $200 billion. The cost to the country was greater: A “lost decade” of subpar economic growth.
That lost decade cast a long shadow elsewhere in the world. In 2017, Andreas Dombret, a board member of Germany’s central bank, highlighted the lesson:
“Because zombie banks were propped up for years and were offered continued protection, they were actually able to support insolvent enterprises. As a result, the economy fared considerably worse than if authorities had intervened more courageously.”
Dombret warned about the risk of creating zombie banks in Europe, even though many had shuffled through its lands in recent years. The oldest bank in the world, Italy’s Banca Monte dei Paschi di Siena (BMPS), spent many years in a zombie-like state, under the burden of high non-performing loans and insufficient capital. BMPS suffered losses in nine out of the past twelve years. After delaying a capital raise for two years, it finally got one away at the end of last year. “Let me allow to say that Monte dei Paschi is not anymore a systemic problem, but it should be considered for what is a truly country’s asset,” its CEO announced in February this year.
Since the S&L crisis, American authorities have been acutely aware of the risk of creating zombies. They were quickest to recapitalise their banking system after the global financial crisis, and they are the most supportive of their banks making money.
Which brings us back to First Republic Bank.
The bank could shamble along in a zombie-like state. At 9.3%, its regulatory capital ratio is above the threshold of a “well-capitalised bank” (unrealised losses are not included). Meanwhile, government-sponsored schemes can continue to feed it liquidity. But its mounting losses will eventually deplete capital and as that becomes more obvious, authorities have a duty to shut off the liquidity taps.
So that leaves three solutions: a private-sector solution, a public-sector solution and … a hougachou solution.
A pure private-sector solution is difficult because accounting principles require any buyer to mark First Republic’s balance sheet to market which means even for the price of a dollar, the deal wouldn’t be economic.
A pure public-sector solution is also difficult. When authorities took Silicon Valley Bank and Signature Bank into receivership in March, they invoked a “systemic risk exception” which allowed them to cover uninsured deposits as well as insured out of the deposit insurance scheme (at an incremental cost of $19.2 billion). It’s not clear that a failure of First Republic could be deemed “systemic” now that everyone knows it’s happening, which could mean uninsured depositors are relegated to unsecured creditors. Uninsured deposit balances have fallen, from $119 billion before the crisis to around $18 billion at the end of last week, but forcing losses on the last few standing without forcing losses on similar depositors at the other failed banks or even those depositors quicker to withdraw could attract controversy. And that doesn’t include the $30 billion of deposits placed by the 11 large banks. Reprising the “systemic risk exception” to cover them would definitely be seen as controversial.
Which leaves a mixed public/private solution. Various proposals seem to be on the table, although it’s interesting that after years of crafting detailed frameworks to deal with bank failure, a new, unique model may be exercised. As in the case of the S&Ls and Japanese banks, the key question is how to allocate losses between the various parties. Putting off the day of reckoning only serves to feed the zombie – and nobody wants that.