What Big-Bank Breakup Hawks <i>Should</i> Be Talking About

Seven years since the financial crisis, talk of "too big to fail" banks continues, even as major legal changes have established new resolution regimes, and a pending Federal Reserve Board rule will require the eight U.S. banks designated as systemically significant to hold $1.6 trillion of equity and long-term, bail-in debt to absorb losses.

Meanwhile, potential sources of systemic instability from bank failure — the potential causes of bailout — have been significantly reduced for all large banks through central clearing, a stay on derivatives close-outs in bankruptcy and numerous other risk reduction requirements. As a result, bank debt spreads show no evidence of a TBTF subsidy, and rating agencies either have removed or are scheduled to remove any uplift in ratings.

But as bank critics assemble Monday at a meeting hosted by Federal Reserve Bank of Minneapolis President Neel Kashkari, who recently called for breaking up the biggest banks, it is not clear how much participants at the symposium will analyze the impact of these changes.

Participants seemed poised to take TBTF as a given and then pivot to two options: Capital requirements so high as to make failure impossible, or — always a fun discussion — how to break up the banks into harmless little pieces.

But it turns out that breaking up is indeed hard to do. TBTF talk is cheap, and actually transforming large, diversified banks into monoline Lehmans and Washington Mutuals — the first to go in the financial crisis — seems to many a very odd approach. Thus, the safest and most common ground has been simply to say that large banks need more capital — with a bidding war on just how much and how soon to commence shortly.

If panelists at the Minneapolis Fed meeting choose to look at the more realistic scenario of recapitalization, they should understand that this prescription, too, suffers from two very significant problems.

First, large U.S. banks already hold levels of capital unprecedented in the modern era. The ratio of common equity to risk-weighted assets at large banks has more than doubled since the financial crisis. (At the same time, liquidity has increased even more dramatically, with the largest banks now holding about 25% of their assets in cash and cash equivalents.)

But of course, one can always say that double is not enough; it should be triple or quintuple. How does one know that simply adding a lot more capital is enough?

As a tangible benchmark, consider this year's Federal Reserve Comprehensive Capital Analysis and Review stress test. It requires large U.S. banks to hold sufficient capital to weather a series of simultaneous shocks that exceed the worst of the financial crisis. These include a precipitous 6.25 percentage point drop in real GDP (over 50% more than during the worst of the financial crisis), and an increase in the unemployment rate from 5% to 10% in less than two years (far more sudden than we saw in the financial crisis). Simultaneously, there are calamities in asset prices: the stock market loses 50% of its value in 12 months; corporate bond spreads blow out 300 basis points in two quarters (versus 50 basis points in the financial crisis). Meanwhile, banks suffer huge litigation losses — suddenly, not years later, as has been historical experience. And interest rates turn negative, crushing bank net interest margins. Finally, for banks engaged in the securities business, there is a further market shock and the failure of their largest counterparty. In sum, this scenario is not only worse than the recent financial crisis, but dramatically worse — not a stress test, but a worst-case test.

Of even greater significance, though, is what it takes to pass this stress test. The standard for passing is not a bank holding sufficient capital (and liquidity) to fail in an orderly way, without recourse to taxpayer support — though that could be a quite reasonable standard. The standard is not even surviving the shock without failing, as most probably assume. Rather, it means (1) emerging from these simultaneous shocks with over 4.5% Tier 1 common equity to risk-weighted assets and over 8.0% total capital; (2) while keeping the bank's balance sheet constant, and assuming no effort to shrink money-losing business lines, as any rational bank (or other company) would do; and (3) while continuing to make all planned dividend payments and share repurchases (in other words, while depleting rather than husbanding its capital).

In essence, the stress test does not require large banks to survive a depression and market crash of unprecedented severity, but rather to be so well-capitalized that their businesses are largely unaffected by such a depression and market crash. That is a lot of capital.

Advocates of impossibly high capital sometimes assume there is no cost to higher capital requirements. Alas, there are costs, which brings us to the ultimate problem with the "big equity" prescription. Reducing a bank's leverage, of course, reduces its ability to lend, to make markets and more generally to support economic growth. (So, too, does requiring it to hold a quarter of its assets in cash and cash equivalents.)

Note further that high and granular capital requirements do not have just general effects but also particular ones: regulators are effectively choosing winners and losers among asset classes. Force banks to assume a rapid rise in unemployment, and they will be disincentivized from making loans whose repayment is most sensitive to unemployment rates — most notably, loans to households without significant savings. Assume an unprecedented drop in GDP, and loans to small businesses look less attractive on both an absolute and relative scale. Assume the failure of a bank's largest derivatives counterparty, and hedging becomes more expensive. And so on.

Regulators have not been eager to investigate these economic-growth and income-inequality costs of higher capital requirements, and I suspect this topic won't be on the agenda in Minneapolis. To critics continuing a bidding war for still-higher capital levels for large U.S. banks, though, it seems fair to ask a few questions. On what historical or analytical basis do you conclude that existing levels are not more than sufficient? Are you confident that you understand how higher capital levels affect lending, market making and ultimately economic and job growth? Are you confident you understand how banks incorporate higher capital standards into their businesses, and how those decisions are changing the market for financial services? Because while unduly lax capital standards can place the economy at risk, unduly punitive ones can do so as well.

Greg Baer is the president of The Clearing House Association.