In front of the children, the Federal Reserve is keeping its collective voice as soothing as possible, but the new CCAR shows that the grown-ups know how big the bumps in the night could prove if U.S. rates drop below the zero lower bound. The negative interest-rate test is the most alarming among a series of scary hypotheses in the new CCAR (see our Client Report) because it so flies in the face of everything the FRB said when it released its FOMC statement only days before. FedFin took negative rates to be a serious U.S. prospect last summer, issuing a paper laying this out at the time. Now that we aren’t the only ones talking out loud about what negative rates might do to banks, it’s worth also revisiting what they could well do to the rest of us. In short, get out the shoeboxes.

A lot of talk in the last few days about negative rates has been surprisingly sanguine, but that’s largely because those doing the talking are theoretical economists and central bankers thinking about what they know best, which for sure isn’t how financial systems respond to a structural inversion as tectonic as prolonged negative rates. To be sure, the forecasting challenge for financial stability isn’t simple. One must consider how low rates go below the zero lower bound (ZLB), which rates go where when across the yield curve, and how long which rate stays south of the border. The Fed’s scenario is far from certain, but it’s extremely informative as well as far from comforting.

Under it, three-month Treasuries start below the ZLB – at minus 0.2 percent in the second quarter of 2016, dip to minus 0.5 percent in the third quarter and then stay at minus 0.5 percent through the first quarter of 2019. In short, this isn’t an accidental or temporary drop, but rather the sustained nominal negative rates advocated earlier this week by former FRB-Minneapolis President Kocherlakota. In the CCAR, the FRB also gives its own yield-curve projections, with the longest-maturity Treasury (the benchmark ten-year) starting the test in the second quarter at a very skinny positive 0.2 percent with the rate bouncing gradually – very gradually — upward to close at 1.7 percent by the first quarter of 2019. And you thought two percent benchmark ten-year Treasuries were bargain basement.

If these rate scenarios are complex, the fundamental impact of negative rates isn’t: under them, depositors pay their bankers and bankers pay their borrowers. Theoretically, some of these disruptions are muffled by currency valuations that achieve central-bank objectives regardless of the paroxysms into which they turn the financial system – the Bank of Japan’s actions on Friday are an example of monetary policy with an eye towards currency valuations particularly critical in export-based economies. Adverse impacts can also be muffled by financial systems that like Sweden’s are increasingly cash-free. They can even be offset when economies depend only on banks so that, as in the Eurozone, depositors have no place to safeguard funds outside of banks so that banks charge depositors even as bankers still eke something out from borrowers who, without banks, are bust.

But what of the United States? It isn’t a system in which depends on exports nor one where banks are the be-all and end-all or where cash is an artifact of times gone by. Making negative rates still more challenging, the U.S. is uniquely an economy with a benchmark asset – those short-term or very low-return Treasuries. As a result, profound disruptions to finance as usual – indeed, to finance as we understand it – could ensue under the FRB’s sustained, significant scenario.

Take just one not-so-technical financial-stability problem: how to charge for trading, investment advice, or even custody services when the assets involved are at a negative rate? Positive fees for negative returns is a real no-win deal that would leave retail and wholesale market participants alike to reconsider what they do, how often they do it, and with whom. Many fees are already at break-even if not below it based on hoped-for rate normalization, but three years under the FRB’s scenario would break not just break-even, but also the backs of many of the firms offering these products.

Market disruptions could also result from incentives baked into the new rulebook. For example, financial institutions could try to meet liquidity, margin, and collateral requirements with above-ZLB sovereigns that still pay rates above the ZLB. Most of these positive-rate sovereigns are, of course, riskier sovereigns and, in the U.S., they might also be agency obligations issued by Fannie and Freddie. The financial system might thus stay operational but only because counterparties post higher-risk collateral. If nothing goes wrong, great; if something does, think doom-loop.

Sovereigns aren’t, though, the only asset class subject to significant distortions if low-risk assets go below the ZLB. The day the Bank of Japan went negative, higher-risk property stocks in Japan soared through the roof even as big-bank equities plummeted. Yield-chasing in the U.S. is starting to meet its maker in the energy sector due to geopolitical forces beyond the FRB’s sphere, but other sectors – commercial real estate, residential mortgages – are booming in ways the FRB’s stringent stress test can’t catch because so much of the lending is outside its jurisdiction. CCAR and negative rates could combine into a vortex in which banks pull back from sectors subjected to tough CCAR stresses, creating a vacuum of remaining, desperate demand to be filled by non-bank lenders and securitizers.

And then there’s the shoebox question. If depositors have to pay bankers to park their money, depositors may well hoard as much physical cash as they can. Central bankers comfort themselves by saying, rightly, that physical cash is bulky, but market participants have lots of ways to work around that, including by taking financial assets and commoditizing them into gold, housing, jewelry, or whatever they think will retain principal value or even appreciate a bit. The U.S. did this once before in a big way in the late 1970s when inflation undermined the value of financial holdings, and negative rates would have a similar effect, differentiated now from the past only by the way new technology will permit cash and deposit transformations with lightning speed, if dangerous consequences.

The thinking at central banks is also that macroprudential regulation could solve for these procyclical forces, but no one has contemplated macroprudential regulation under the twin strains of a financial system that doesn’t depend on banks in the grips of negative Treasury rates. Our negative-rate paper goes into depth here, as does a more recent speech on the limits of macroprudential policy.

All this is scary enough, but here’s another sobering thought: a Minsky moment. So far, we haven’t seen these shocks to the system and, then, their unintended and often awful consequences in other regimes with negative rates. None of them does, though, have a currency that’s the global benchmark nor an economy on which the rest of the world relies for growth. Further, at least until Friday’s BoJ decision, negative rates were instituted when markets were still pretty optimistic about recovery in general and most banks in particular. Not so much right now, making investor psychology particularly easy to spook.

To avoid not just the scenarios I’ve outlined above, but also the far worse ones that would befall in a Minsky moment, it’s critical for the FRB to think about negative rates not just as an academic construct that might solve for the monetary policy “liquidity trap” or temporarily boost growth. We learned in the financial crisis how closely macroeconomic stability depends on financial-market stability. To the extent negative rates undermine financial stability – and I fear they could do so in far-reaching ways – real economic growth and prosperity would suffer.   The FRB is engaged now in a vigorous debate over whether monetary policy should consider financial stability or if monetary policy and macroprudential policy should each proceed on a separate track. Negative rates would settle this, forcing them on the same track, but they’d be rumbling heading towards each other.