The Federal National Mortgage Association — known as Fannie Mae — and the Federal Home Loan Mortgage Corporation — Freddie Mac — were poorly structured from the time, 40 years ago, when they were set up as so-called government-sponsored enterprises. Both of these technically private companies, designed to foster the issuance of home mortgages, enjoyed implicit federal backing in the event they got into financial trouble but only weak regulation to prevent such trouble. Essentially, the federal government insured the companies’ liabilities but never charged a premium.

Fannie and Freddie had a license to print money. They could borrow at an interest rate only a bit over the Treasury rate and then accumulate large portfolios of mortgages and mortgage-backed securities earning the market rate. What a deal — borrow at the low rate, invest at a higher one, hold little capital and let the federal government bear the risk! Investors enjoyed high returns, and management enjoyed high salaries. Incidentally, politicians also got a steady flow of campaign contributions from the companies’ executives.

Fannie and Freddie’s risky policies led to their near collapse; in September 2008, the federal government brought them under federal conservatorship. Fannie and Freddie have cost taxpayers about $150 billion so far.

On Tuesday, the Obama administration plans to hold a conference to address the question of what to do with the two companies. Clearly, it would be an inexcusable mistake to reconstitute them as private companies in anything close to their prior form. Some people have suggested recasting them as a single new “Fan-Fred agency” that would continue to securitize and guarantee home mortgages. It’s true that Fannie and Freddie played an important role in developing the market for mortgage-backed securities. But they have completed that work, and they should not be preserved in any form. They should be thanked for their successes and gracefully retired.

Can the home mortgage market stand on its own, without support from federally sponsored mortgage companies? Experience tells us that the answer is an unambiguous yes. When Fannie and Freddie curtailed their operations after the disclosure of accounting irregularities in 2003, there was no effect on mortgage rates. We have seen how the jumbo mortgage market, for loans too large to be eligible for Fannie and Freddie purchases, has long operated efficiently, with rates only slightly above the rates on smaller mortgages. And many other asset markets, like the one for securitized auto loans, have functioned well without federal intermediaries.

If Fannie and Freddie were to continue to operate with the government absorbing all their risk, they would keep a large share of the market. But that system has been terribly expensive for taxpayers. The evidence shows that the private market can originate, securitize and distribute home mortgages efficiently on its own. While it’s true that the private market brought on the financial crisis by creating so many subprime mortgages, Fannie and Freddie did not block that parade; they joined it — indeed, in some respects led it.

In principle, it ought to be possible for government financial agencies to be self-supporting. But decades of observation have convinced me that there is no practical way to prevent the government from inserting hidden subsidies and special interest mandates into the agencies’ operations. If there are to be more federal housing subsidies — and I hope there are not — they should be legislated transparently.

The danger in having any new mortgage agency is that its guarantees would subsidize mortgage risk, eventually leading to further taxpayer losses. The only sure way to prevent that outcome is to phase out Fannie and Freddie. If the home finance market were fully private, then it would bear the losses from its own mistakes in pricing and insurance. The proper government role is regulatory oversight and not direct operation of financial firms.

Fannie and Freddie could not be shuttered immediately; they are too large. A sensible transition plan would have them stop buying new mortgages, and their portfolios would decline as the mortgages they own are paid down. Within 10 years, the portfolios would shrink to insignificance.

Their securitization business, whereby they purchase mortgages and issue securities against them, should likewise be wound down. A practical approach would be to set a gradually rising schedule of fees, motivating private companies to enter the securitization business.

In 10 or 15 years, the companies would be gone, closing a chapter in American financial history that enjoyed considerable success but ended very badly and at great taxpayer cost.