In a speech last week, Paul Tucker – the man of the moment for all matters systemic in the U.K. – focused on an issue often overlooked in the furious debate on financial reform: what to do about the increase in systemic risk sure to result when derivatives trading moves to central counterparties (CCPs). Regulators demanded CCPs because they believe that organized clearing prevents disorderly panics when uncentralized counterparties are forced to fend for themselves. Right they are, except for the fact that the word “central” in CCP says it all: the risk one distributes around the financial system comes home to roost in a CCP. This question has been on the back-burner, where all the pots up front are bubbling on what to do with SIFIs. But, we expect CCPs to crowd up front with all the other systemic pots early next week, when House Republicans argue that too many too-big-to-fail firms are left unscathed by Dodd-Frank.

As Mr. Tucker points out, the challenge for CCPs is embedded in their mission: clearing for members. CCPs are membership organizations, not single institutions with third-party creditors and shareholders. Thus, for CCPs, the first line of defense is demands on their members, what used to be called the “last drop of blood” at commodity clearing houses. However, as Mr. Tucker rightly notes, that last drop of blood may be drained from big banks at, shall one say, an inconvenient moment. This of course exacerbates systemic risk instead of abating it.

What to do instead of calling on stressed members to pony up when they likely can’t? CCPs generally now have default funds or similar kitties intended to insulate members from awkward calls. But, how much is in them varies dramatically from CCP to CCP. This is not just the result of the natural preference for CCP members to keep their money for themselves. It can also reflect significant differences in the types of trades handled on a CCP, the legal obligation of the CCP (widely different in different jurisdictions and for different financial instruments) and the degree to which funding is based on net or gross settlement risk. As a result, these funds are necessary buffers, but hardly sufficient bulwarks against systemic risk.

The Dodd-Frank Act seeks to solve this by creating a class of CCPs (including those in the payment and settlement arena) dubbed financial market utilities (FMUs). The word “utility” is no accident – the structure of FMU regulation in the new law is premised on the view that these CCPs are the equivalent of the electric grid. The lights must stay on, so FMUs must be regulated to ensure they are always powered up.

Dodd-Frank tries to ensure this by giving FMUs access to the Federal Reserve’s discount window so that, under stress, they have an alternative to demanding the last drop of blood from anemic members. This is what so upsets the law’s critics, who fear that CCPs will become a new class of TBTF entities with still bigger claims on the taxpayer than other systemic financial institutions.

But, we see no way around this – either CCPs have a backstop or they won’t work. Without CCPs, the market goes back as before, with a few very large dealer banks essentially acting as CCPs backed not only by their own access to the liquidity window but also by many other SIFI safeguards. Thus, CCPs don’t add more risk to the discount window and, by inference, the taxpayer – they just concentrate it from a few big banks to still fewer big CCPs.

This isn’t exactly a comforting thought. CCPs must have a backstop to ensure they can meet market needs under stress, but that backstop could of course make them slipshod – moral hazard rides again. One solution to this is to be found in Mr. Tucker’s speech: require CCPs to serve as systemic risk managers. If they sit between counterparties, they can and should ensure that those with whom they do business can meet their commitments, including under acute stress.

To defend FMUs against their foes, the Fed will need to craft a balanced regulatory regime for CCPs and other infrastructure entities. It can and should include discount-window access, but only at the cost of demonstrated risk-management capabilities at any eligible CCP. Before the Fed opens the window, it needs to put the screens up