The bank that swallowed a fly

There once was a bank that swallowed a fly

There is an old nursery rhyme about an old woman who swallowed a fly, then a spider to catch the fly, then a bird to catch the spider, then a cat to catch the bird, then a dog, et cetera. I was reminded of that rhyme when writing about Basel III’s perversities last week. Both the structure of bank regulations and the path that led to them have much in common with the old woman’s strategy: increasing the scale of the underlying problem with each step. But this can’t go on forever. In the nursery rhyme the woman finally dies when she swallows a horse. The rise of alternative mechanisms for payments and credit, like stablecoins and “shadow banks,” suggest that Basel III may create a similar fate for banks.

You don’t know the half of it…

I’d like to start by expanding upon last week’s discussion of Basel III’s many inherent and well known flaws. Several readers noted that even with my exceptional skill in writing and exposition, Basel III appears incomprehensibly complex and self-contradictory. They’re right of course – both about my exceptional skill and Basel III’s inscrutability – but I only scratched the surface with my discussion of the regulatory framework’s four contradictory financial ratios. Here are some of my favorite examples of how utterly ridiculous Basel regulations are.

Stable funding is a financial oxymoron

I glossed over the Net Stable Funding Requirement (NSFR) last week because it wasn’t critical to the interaction of financial repression and fiscal dominance. But it is one of the sillier parts of Basel III when you consider the fundamental role of fractional reserve banking in society. Fractional reserve banking is fundamentally unstable, but most modern economies have adopted it because, through the magic of maturity transformation, i.e. turning zero-maturity deposits – the basis of non-cash payments – into illiquid, longer-dated loans, it fulfills societies’ needs for both a payments system and credit creation. The NSFR undermines both functions by requiring banks to increase the maturity of their funding as the average maturity of their assets increases. That eliminates maturity transformation and the deposits that facilitate payments.

Didn’t we solve that problem before?

That may seem sensible on its face, but only if you ignore the last four hundred years of financial and economic development. The Swedes invented central banking four centuries ago to provide troubled banks with emergency funding, lent against good assets at penalty rates to discourage abuse. Capital requirements followed to ensure that banks take the first loss on any bad loans, to cushion both depositors and the lender of last resort during a bank run. Americans created deposit insurance two centuries ago to reassure nervous depositors. Given these long-standing tools, regulators’ adoption of the NSFR amounts to an admission that they haven’t been doing their jobs.

Why make do with just one capital requirement?

Last week I noted that Basel III has not just one, but two different capital standards, and they contradict each other. But it’s worse than that: they encourage unhealthy risk-taking behavior. The Leverage Ratio (LR) constrains total leverage, regardless of what assets the bank owns. The Risk-Weighted-Asset (RWA) ratio rewards holding less risky assets (like government bonds) by weighting assets according to their risk (as defined by regulations). This creates a tension between the two ratios. When a bank runs into the LR it is incentivized to take more risk in its existing assets; conversely, when it runs into a limit on its RWAs, it is encouraged to instead take more leverage.

Technically, tar is a liquid

But nothing compares to the sticky mess of contradictions that is the Liquidity Coverage Ratio (LCR). Like the NSFR, the LCR, which requires banks to hold enough liquid assets to cover any stressed outflows, is fundamentally at odds with maturity transformation. Yet, ironically, the LCR undermines the longest-standing liquidity requirement for banks to protect against unusual deposit demand: cash in its vaults. Because vault cash is harder to monitor or transfer, it is treated as an inferior “high-quality, liquid asset” (HQLA) to bank reserves, the banks’ electronic deposits at the central bank. That is, of course, doubly ironic since the central bank’s original and primary function is to create reserves on demand for banks to borrow in times of trouble. And yet, the irony of the LCR continues: government bonds, like Treasuries, qualify as HQLA, but also are considered inferior to reserves. That not only creates “regulatory dominance” of the Fed, as discussed last week, but also impairs critical functioning of the Treasury market, particularly in periods of stress, as discussed below.

LCR can’t get along with anyone

The LCR also comes into conflict with all three of Basel III’s other required ratios. Its premium on liquidity conflicts with the RWA ratio’s prioritization of risk: safe assets (like secured loans) aren’t always liquid, while some liquid assets aren’t safe. The LCR also conflicts with the LR since HQLA count towards total leverage, pushing banks closer to the LRs boundary and heightening both ratios’ tension with the RWA ratio. Finally, despite likewise limiting banks’ ability to transform maturities, the LCR conflicts with the NSFR, too. While the former encourages banks to hold government securities of any maturity, the NSFR punishes holding longer-term bonds by requiring banks to hold term funding (rather than cheaper deposits) against them.

One minus one equals two

To illustrate how totally bonkers these rules are, consider the effects of the following transactions by a bank. In the morning, it lends cash to a client overnight through a repurchase agreement (repo), a loan backed by a Treasury security. Because the bank is taking a Treasury security on its balance sheet overnight – as collateral for a loan, mind you! – the NSFR requires it to increase its long-term funding. But when later that day it borrows the cash back through a repo with another bank, pledging the same Treasury security and removing it from its balance sheet, it gets no relief because the NSFR doesn’t consider overnight repo to be stable funding. Got that? The bank ends the day with the same cash and security position it began the day with, but now has an increased term funding requirement.

If you think 1 – 1 = 2, I can help. Subscribe to Seriously, Marvin?! today.

Undermining “exorbitant privilege”

Former French President Valéry Giscard d’Estaing complained that the US receives the “exorbitant privilege” of lower interest rates from issuing the reserve currency. But the US government’s true borrowing advantage comes from the deep liquidity of the Treasury market that makes US debt so desirable to hold as a security. That “privilege” is undercut by Basel III’s nonsensical contradictions. The combined punishment of the LR, the NSFR, the LCR’s preference for reserves over Treasuries, and recently introduced market risk rules so discourage banks from engaging in repo transactions that liquidity in that market has measurably fallen. It also has led to spikes in repo and Treasury rates on days with large tax payments or Treasury issuance that drain reserves from banks’ accounts at the Federal Reserve.

Complexity and non-transparency

Basel III’s complexity also impairs transparency and, arguably, effective supervision. As if four contradictory financial ratios – NSFR, LCR, LR, and RWA ratio – weren’t confusing enough, the last of those is not one ratio, but three! Because Basel III defines three different types of capital – CET1(common equity capital), Tier 1 (CET1 plus preferred shares and contingent convertibles, “CoCos”), and Tier 2 (Tier 1 plus long-dated, subordinated debt), there is a different RWA ratio for each. But that’s not all. Banks also must maintain a capital conservation buffer in case of losses; a countercyclical capital buffer to avoid a systemic reduction in lending during recessions; and additional capital for global, systemically important banks (G-SIB surcharge).

Never fear! I have a model

Yet, no one should feel sorry for big banks’ extra capital charge, since they’ve rigged the system to give them an advantage over smaller banks and regulators. While there are standard regulatory risk weights for different types of assets, sophisticated banks can create complex models that allow them to reduce those weights. For instance, suppose that a corporate loan has a standard risk weight of 100%, meaning that the full loan value counts against a bank’s RWAs. If the bank can show through its modeling that its corporate loans have a lower default rate and that it has further mitigated credit risk with off-balance sheet hedges, it might reduce the effective risk weight on the loan to 60%.

If we swallow a dog…

Unsurprisingly, regulators belatedly figured out that G-SIBs were gaming these models and – swallowing a dog to catch the cat – have tightened the rules on modeling to constrain banks’ leeway in both what assets can be modeled and how. Further, the new rules institute a floor of 72.5% of standardized risk weights, adding yet another non-linear constraint for banks to optimize that reduces everyone’s transparency: banks, regulators and markets. Like a bad movie sequel, the new rules – currently being introduced in Europe but not yet approved by US authorities – are called “Basel III Endgame.”

Bad medicine

All this complexity has a number of negative effects. First, it is anticompetitive: only big, incumbent banks can afford and have the historic data to develop complex risk models that reduce their capital. Even with standardized risk weights, the cost of regulatory compliance is onerous and scale impervious, meaning that its costs fall more heavily on smaller banks. Further, complexity reduces transparency and thus market discipline that might be imposed on riskier banks if the scope of their risks was more easily observable. Extremely complex regulations also likely reduce the efficacy of regulatory oversight. With so many different, conflicting variables with non-linear effects to track, regulators are much more likely to miss the basics as we saw with Silicon Valley Bank. Finally, it is highly likely that overwrought regulations are allowing the build-up of large, unobserved risks, hiding in the shadows.

Swallow a goat to catch the dog…

The complexity isn’t an accident, however. It is a product of the political economy of factional reserve banking, which is prone to crises. Societies face a trade off between crisis frequency and severity, and economic growth promoted by credit. Capital ratios are the dial that tune that trade off: more capital, less crises and less growth; less capital, more of both. When the inevitable crisis occurs, everyone’s incentives align towards complexity: governments want to “do something” without damaging growth; regulators and central banks want to shift blame to inadequate prior regulations and expand their own budgets in the process; and surviving banks lobby for complex, anticompetitive regulations instead of increased capital. This is exactly what happened following the Global Financial Crisis (GFC). JPMorgan, the strongest bank to emerge from the crisis, has its fingerprints all over Basel III, regulators more than doubled their staffs, and politicians turned out legislation better weighed than word counted. Like swallowing a goat to catch the dog, all of this layered onto previous regulatory complexity.

There was an old lady who swallowed a horse…She’s dead, of course!

Yet, as I described in Monetary revolution in the making, the political economy appears to have shifted, arguably in response to the ineffective, confusion of regulations and policies to emerge from the GFC. Most of Basel III’s internal contradictions result from trying to force an institution that evolved maturity transformation to deliver both credit and payments to separate those functions and eliminate maturity mismatches. It’s like swallowing a bird to catch the spider you swallowed to catch a fly; it doesn’t work. Eventually, the old lady went too far swallowing a horse and died.

A final irony?

Another of Basel III’s ironies is that, through regulatory arbitrage, it helped accelerate the growth of so-called “shadow lenders.” Despite the sinister name applied to them, most are financial firms that make loans funded with term-debt or locked-in capital, thus avoiding the maturity mismatch that plagues fractional reserve banking. Coincidentally, crypto payments platforms arose in the background and evolved stablecoins as fiat currency payment systems backed solely by short-maturity, liquid securities. Thus, the very features that Basel III attempts to cram into one legacy institution have arisen naturally without the complex regulation, or perhaps one could argue as an antidote to it. The independent development of stablecoin “narrow banks” with shadow lender “merchant banks” fulfills a century-old dream of a group of University of Chicago economists to create a safer financial system.