The idea that the government can happily give “implicit guarantees” of Fannie Mae and Freddie Mac debt that will never cost it anything has come to an end. The two government-sponsored enterprises have used their government support to run their total obligations up to $5 trillion. Now they are experiencing financial stress, with their stock prices down more than 80% from their 52-week highs. The big $5 trillion turkey of GSE risk has come home to roost in the U.S. Capitol and won’t go away.

Treasury Secretary Paulson has asked Congress for unlimited authority for the Treasury to use the taxpayers’ money to both lend and make equity investments in Fannie and Freddie, so that they have continuing easy access to the global debt markets.

This is recognition that the “implicit guarantee” was always in fact a real guarantee, although occasionally the contrary was asserted. For example, “Nobody should be under any illusions that there is any guarantee, implicit, explicit, whatever-plicit,” the present chairman of the House Financial Services Committee, Barney Frank, said in 2003.

But bond investors across America and around the world bought Fannie and Freddie debt securities with the understanding that the guarantee was real — and government action will confirm that their understanding was correct.

A ranking member of the House Financial Services Committee, Rep. Spencer Bachus, recently pointed out the fundamental issue: with the proposed action, you “privatize the profits, socialize the losses.” He’s right, but it’s too late — the government is already on the hook for the risk. This contradiction in the very definition of a GSE was highlighted in a similar vein two decades ago by the late Rep. James Pickle of Texas. “The risk is 99% public,” he said, “but the profit is 100% private.”

Fannie and Freddie used the GSE structure to grow much bigger, make much bigger profits for many years, until their recent billions in losses, issue much more debt, pay their management handsomely, and take on more risk than any private company could have on the same capital base. They could do this because the real capital was the government guarantee, although it was provided for free and “implicit.”

Well, it is about to become explicit. This will parallel the action of Congress in the 1980s. When faced with the savings and loan collapse, it changed the Treasury guarantee of savings and loan deposits from implicit to explicit. It’s time to get rid of the fuzzy idea that “implicit guarantees” aren’t real and adjust our thinking and actions accordingly.

There are two pure organizational forms: a truly private company and a government agency. A GSE is a hybrid form, with an undesirable risk-reward structure and fuzzy recognition of risk to the government. The fiduciary duty of any GSE management is to figure out how to extract the maximum profit for their shareholders by running up the debt on the government’s credit card.

It wasn’t always this way. Fannie started out as a pure government agency, with its debt on budget, and was so for 30 years, between 1938 and 1968. Whatever we may think of the desirability of a government banking operation, at least the risk structure was clear. What happened?

Like some other bad ideas, Fannie as a GSE came from the Johnson administration. In 1968, it made Fannie a GSE in order to get it off the federal budget and keep down the reported deficit. In 1970, Freddie was created on this GSE model. In the early 1980s, Fannie was insolvent on a market value basis, but survived because of the government guarantee. It was at that point a much smaller risk turkey, in 1982 about 3% of its current size, with a full generation of battening ahead of it.

How should the Congress address the now huge GSE risk turkey roosting in the Capitol dome today? It needs to protect the creditors to make good on the government’s guarantee and to keep the mortgage market functioning, but let the equity investors suffer the consequences of losses — and do this while laying the groundwork for stronger capitalization in the future.

I believe the most efficient way to do this would be for the government to become a subordinated investor as an unambiguous signal to the world’s bond buyers. It should buy a significant amount, say $10 billion each, of senior subordinated debt in Fannie and Freddie. In other words, the investment must be senior not only to the common stock, but to all existing preferred stock and subordinated debt, being the last part of capital to suffer any losses and the first to get cash returns.

This action of the government becoming a direct investor would make entirely clear to the credit markets that the bond holders have nothing to worry about and can continue to finance the mortgage market in safety.

But there need to be serious conditions when it comes to the equity investors and the management. Something like this could work:

• A cash interest yield to the government’s senior subordinated investment of 4%, about the 10-year Treasury bond rate, to cover the government’s cost of financing its asset.
• No dividends to the common or any previous preferred stock as long as any portion of the government’s investment is outstanding, without the approval of the secretary of the Treasury.
• An additional pay-in-kind interest payment to the government of 11% per year, bringing the taxpayers’ potential annual return up to a fair return of 15%, in exchange for the risk they are taking.
• Reduction of management compensation and reconsideration of compensation structure.
• At least five members of the GSE board of directors to be selected by the Treasury Department to represent the government’s interests, with their fiduciary duty defined in statute to run to the taxpayers, not the common stockholders. At least two of these directors to be members of the Audit and Compensation Committees.
• The senior subordinated debt to be redeemable at par when the GSE achieves a capital ratio, not including the government’s investment, of 5% of assets plus 1% of off-balance sheet MBS — this is twice the current requirement, though still modest by financial institution standards.

This would be a sensible way for creditors to be fully protected while the existing equity holders pay the price for the risks they took. They would still preserve an interest, although an interest junior to the PIK accruals, in future upside and profits.

As for the fundamental structural contradiction of a GSE — well, we’re in the middle of a severe housing bust and we can’t fix that just yet.

Mr. Pollock is a resident fellow at the American Enterprise Institute in Washington, D.C.