When FRB Governor Dan Tarullo testified before Senate Banking earlier this week, big banks and the market trembled as he announced a set of tough new rules for the biggest U.S. banks. They should have been even more afraid: his written testimony takes Fed policy one step farther by declaring that the biggest banks must be safer than a small one. The only way I know they can do so is to turn themselves into lots of small ones. If big banks don’t like this future, they’ll need to lay out why others should care – and fast!

Specifically, Mr. Tarullo said, “… [T]he Federal Reserve has been working to develop regulations that are designed to reduce the probability of failure of a GSIB to levels that are meaningfully below those for less systemically important firms and to materially reduce the potential adverse impact on the broader financial system and economy in the event of a failure of a GSIB.”

The no negative-externalities point is a frequent one in FRB statements – it means that big banks should fail without adverse impact on innocent bystanders (including the global economy). Based on feared negative externalities, U.S. policy sports both belts and suspenders – that is, it hopes that orderly resolution protects the world from harm, but it nevertheless demands an array of tough new rules so that the resilience of orderly resolution is never tested.

The new demand Mr. Tarullo lays out is not only that G-SIBs be both robustly regulated and resolvable, but now also that they be so secure as to prevent far less risk than far smaller banks. Can this be done?

Two big banks are the rock on which the repo market depends. Can this be shattered without also blowing up a market that, while risky, is also critical for fixed-income liquidity? What to do with the payment system, which goes through a few big banks at critical junctures? Simply splinter it? Do G-SIBs hog derivatives markets and securities financing because they can or because the infrastructure of these complex activities demands size and scope? Can the world live without derivatives and securities financing – or at least live with a lot less of it – if big banks are shattered into low-risk bits or will the business just migrate to new behemoths?

Can the FRB really level the failure field between very big banks and others without also restructuring the financial marketplace? Which smaller banks are sufficiently riskless as to satisfy the FRB and why? Actually, small banks fail a lot – mortgage, commercial-real-estate, and many other woes crater them on a remarkably regular basis. Is thus some risk for G-SIBs okay as long as it isn’t too much and, then, how much is enough?

In the past, the FRB has resisted asset-size thresholds because it recognized that risk – including systemic risk – can come from seemingly small corners in the financial market. Now, though, it seems to believe that anything really big is so risky that it needs to be smaller. Do we want bank rules graduated by size – Mr. Tarullo has also proposed flat-out exempting small ones from Volcker and otherwise tempering all the Section 165 standards – or do we want prudential and resolution rules premised on risk? With all the new rules coming at them and a new Senate sure to be even less hospitable than the current one to very big banks, this is a critical question big banks must answer persuasively if they want to stay in all the businesses that now rely on them.