When the regulators put out the risk-retention proposal this week, one would have thought there was nothing to it but the exception under it for Fannie Mae and Freddie Mac. Legislation was promptly introduced to block this even as pundits – see, for example, the Floyd Norris column in today’s New York Times – pounced on it as proof that private capital won’t ever come back to mortgage finance. The controversy results in large part from the misguided way in which the proposal crafts the GSE provision – more on that below – but it’s compounded by a major missing fact. If the GSEs are forced to hold risk positions above and beyond their guarantees or private mortgage insurance isn’t allowed to back them, this high risk low down payment business will flow not to the private capital for which many hope, but straight into the arms of the FHA and, as a result, taxpayers.

For good measure, the cost of the GSEs to taxpayers will grow still higher and the challenges of privatizing them some day still more daunting.

First, to the specific GSE provisions in the risk-retention proposal. They aren’t in the definition of a “qualified residential mortgage” (QRM) allowed out from under risk retention, as mandated by Dodd-Frank. Instead, the issue is addressed in the definition of the type of capital at risk allowed to constitute risk retention. The issue is complicated by other provisions in the law, which expressly state that loans backed by the U.S. Government should be considered for an exemption, but that nothing backed by Fannie, Freddie or a Home Loan Bank qualifies as such.

In the NPR, the GSE exemption is listed in the course of the other backstops counted as permissible risk retention. The text states that GSE guarantees will count as risk retention while Fannie and Freddie are in conservatorship or receivership, with the preamble going on to state that this is being done because the U.S. stands behind the GSEs in so many ways as now to make them, for all intents and purposes, U.S. Government entities. This is true, but still problematic under the text of the law, let alone in the political context of GSE reform.

Instead of shoving the round peg of a GSE exemption into the square hole of the statutory prohibition against it, regulators had another way to address this point that would have worked better and roused far fewer howls. It should have permitted all guarantees and credit insurance provided by regulated entities that provide capitalized credit-risk enhancement to count as risk retention. We simply don’t understand why a five percent vertical tranche of a structured MBS counts as risk retention while a 100% principal-and-interest guarantee doesn’t. Is it’s because a guarantee is off-balance sheet? Not any more under recent FASB rules – see also the banking-agencies’ costly capital rules on consolidated ABS. Is it because federal regulators don’t trust each other – e.g., the FHFA for the GSEs – or state insurance regulators? If so, there’s a much bigger problem here that a misguided risk-retention definition can’t cure.

To get private capital back into mortgage finance, the risk-retention rule should have identified not just a reasonable QRM definition, but also solid forms of credit enhancement that provide what Congress demanded: capital at risk to ensure incentive alignment. The huge incentive-misalignment problems in residential MBS didn’t come in those backed by full guarantees from the GSEs or comparable private capital; they came from highly-structured, engineered and, sometimes phantasmagorical RMBS constructs that ensured no capital was ever at risk but that of the hapless investor. We’re working our way through this reg now to see if the complex structures allowed to count as risk retention are as meaningful as regulators said they were when touting the rule this week. But, even if the proposed risk retentions count for something, we still don’t see why others with proven records and real capital at risk were discounted in their favor.

But, ignore all this and take for a moment the proposition that the GSEs aren’t government agencies – correct – and, then, that guarantees shouldn’t count for anything from anyone – wrong, as noted. Then, as proposed by some, shove the GSEs back under the risk-retention rules. What then? The GSEs’ debt will rise, their portfolios will grow and the capital hole taxpayers must fill to restore them to a semblance of solvency will ascend to still more atmospheric heights. The only solution to this is to curtail the GSEs, which of course forces us right back into the arms of the FHA.