The Corker-Warner bill does a creditable job—certainly the best of any current proposal for a government-backed system—in insulating the taxpayers from risk. But Congress has to face the fact, made plain by what happened with Fannie Mae and Freddie Mac that any plan in which the government covers "catastrophic losses" is fatally flawed.

When the government backs any system—whether through deposit insurance, flood insurance, pension benefits or anything else—the beneficiaries have only limited interest in the risks they are taking. In a housing finance system, that means investors in mortgage-backed securities have no need to be concerned about the quality of the underlying mortgages. By putting the government's credit behind the securities their system would create, Senators Corker and Warner have fallen into this trap.

The Senators will of course say that they have built into their system plenty of private capital before the government has to step in, and that this private capital will take most of the likely losses before taxpayers might be called upon. What's more, there will be a mortgage insurance fund, administered by a Federal Mortgage Insurance Corp. This entity—like the FDIC's bank insurance fund—will absorb losses if they exceed those taken by private capital. The whole system will be overseen and regulated by the FMIC, which will have the authority to determine mortgage underwriting standards. 

What could go wrong? Quite a lot.

First, the insurance fund. We should recognize by now that insurance funds run by the government are not insurance in any meaningful sense. The government does not effectively price for risk; it prices to confer political benefits, just as Congress delights in spending. Even in the unprecedented event that the FMIC actually prices for its risks, there will come a time when the Government Mortgage Complex—the Realtors, homebuilders and community activists—will call upon Congress to stop accumulating money in the fund. It's a tax, they will say, on the people who are trying to buy homes, and the fund is certainly large enough for any future catastrophe. Congress, of course will relent, as they did when they capped the bank insurance fund at the behest of the banking industry many years ago. After all, why needlessly impose a tax on homeowners?

When the catastrophe actually occurs, the mortgage insurance fund—like the bank insurance fund—will be found to be inadequate and the taxpayers will once again be called upon to fill in the hole. Exactly this just occurred when Congress had to pony up $9.7 billion to support the federal flood insurance program after Hurricane Sandy.

Still, the bill requires private capital, which Corker-Warner decrees must bear the first loss on the securitized mortgages, ahead of the insurance fund. Here is where the Corker-Warner plan is most original. It requires that private investors take no less than 10% of the first loss on the mortgage-backed securities, which are issued by firms specially registered with the FMIC.  The issuers are basically utilities, not intended to take losses themselves but required to arrange for private investors to take the first losses, with the insurance fund picking up the rest.

Doesn't this structure create a class of private investors who have an interest in the quality of the mortgages underlying the MBS? Unfortunately, no.

They have an interest, to be sure, but it's an insurer's interest; they will require compensation based on their views about the likelihood of default by the mortgages in the MBS pool. This cost will have to be included in the cost of the mortgage. Thus, the private loss-takers have no particular interest in making sure that the mortgages are good quality—only that their compensation is commensurate with the risks in the pool they are presented with. A portfolio lender, on the other hand, operating in an environment in which the government is not setting low mortgage underwriting standards, has an interest in, and the ability, to set standards that will reduce the likelihood of default. That's why a fully private market will generally produce high quality prime mortgages.

So, in the end, we have investors in the securities that get paid no matter what, and we have loss-sharing participants who are paid only for the risks they are taking.

What's wrong with this picture? Clearly, there is one important group whose interests are ignored—the nation's homeowners. In a system where the government is establishing the underwriting standards—that is, the quality of the mortgages that are securitized—there is no question that those standards will be loose. The government's political interest is in handing out goodies, and easy mortgage credit has no costs today and only speculative costs down the road. As we saw with the affordable housing requirements imposed on Fannie and Freddie, there are many ways that mortgage-quality rules can be adjusted so that the maximum number of people—no matter what their credit standing—can get the financing they need to buy homes.

The Corker-Warner bill provides only one underwriting limitation; the down payment must be at least 5%. It also specifies that the Qualified Mortgage rule will apply, but that rule conspicuously ignores such basic mortgage underwriting standards as a good FICO score. A 5% down payment, coupled with indifference to the borrower's credit record, will assure that in the first serious recession, many families will lose their homes, many others will be underwater, and many more who are struggling to meet their obligations will live in neighborhoods where foreclosures abound. Sounds familiar.

So in the end the Corker-Warner proposal, by focusing on protecting the taxpayers and compensating investors, will leave homeowners with the losses.

Peter J. Wallison is the Arthur F. Burns Fellow in Financial Policy Studies at the American Enterprise Institute. His most recent book is Bad History, Worse Policy: How a False Narrative about the Financial Crisis Led to the Dodd-Frank Act.