Insure or Invest? Or does it make a difference?

As the talks on Capitol Hill went from apparent progress to partisan disorder on Wednesday and Thursday, a one-page set of principles from House Republicans circulated, titled

As the talks on Capitol Hill went from apparent progress to partisan disorder on Wednesday and Thursday, a one-page set of principles from House Republicans circulated, titled “Economic Rescue Principles.” It was, in some ways, the model of a rescue proposal from people whose fundamental philosophy abhors government intervention in the marketplace.

Among the key provisions: Instead of using taxpayer money to help recapitalize struggling financial companies, the government should offer tax breaks and ease regulation to encourage private capital, now sidelined by anxiety, to return to the market. And, replacing Paulson’s proposal to invest $700 billion in troubled mortgage-related assets, the GOP plan would have the government insure those assets instead.

We were intrigued by the insurance concept. The proposal said the government already insures about half of mortgage-backed securities — that would be through Fannie Mae and Freddie Mac, primarily — and “can insure the rest,” though it shouldn’t do so at taxpayers’ expense.

It sounded intriguing. But at the end of the day, the plan has much the same flaws as Paulson’s — and leaves essentially unanswered how the capital markets would get the capital and liquidity they need. “To me it seems like just another bad trade,” says Barry Ritholtz, chief executive and director of equity research at FusionIQ, who also blogs at The Big Picture.

One problem with the insurance model is that many of the securities in question are known to be toxic. It's almost like proposing, Ritholtz says, "go down to Galveston, Texas, and write insurance policies on all those houses that blew away."

And pricing that kind of protection accurately -- the Economic Rescue Principles say Treasury should "charge premiums to the holders of [mortgage-backed securities] to fully finance this insurance" -- means it could well prove too costly to succeed.

"You’re buying insurance on toxic assets, so what would your premium be? It’s pretty huge," says Frank Lawatsch Jr., a law partner with Day Pitney LLP in New York. "They couldn’t afford these premiums; it would kill them."

In addition, the assets backed by this insurance would effectively become government-backed securities, Lawatsch said. That has its own drawbacks, as the saga of Fannie Mae and Freddie Mac has taught us.

Now, Lawatsch represents banks -- including some that hold said toxic assets -- and also "potential vulture investors," so he has a distinct point of view. (He's not much of a fan of the Democratic-led compromise plan, either, and grudgingly says Paulson's original plan could work; what he really wants is government purchase of preferred shares from the financial companies, which he says would be simple, inject capital and give taxpayers a share in any upside.)

But there's more on the insurance angle: It turns out that selling financial companies insurance on their assets works out to be about the same, economically, as buying and reselling the assets. In both cases, the nagging question remains: How much does the government risk of the taxpayer's money in the process?

Here's the reasoning:

When the government buys and then re-sells the troubled assets, its gain or loss is the difference between what it paid and gets in the sale. If the government "overpays" -- ie, it pays more than the assets are ultimately worth -- then taxpayers lose, and the companies get a windfall. If the government "underpays" -- ie, gives the companies less the assets are ultimately worth -- then the taxpayer wins, profiting at the company's expense. If the government manages to price the assets exactly right, then it's a wash.

However -- and this point has been raised elsewhere, including in BusinessWeek -- if the Treasury underpays or manages to pay just what the assets are ultimately worth, the move isn't likely to do the financial markets much good. After all, if the problem is too little capital, the companies will essentially be swapping one asset one asset (mortgage-related securities) for another (cash) or equal or lesser value. Only if the government overpays does the scheme pump capital into the system -- to the taxpayers' detriment.

Now consider insurance: Companies would pay the government a premium up-front to insure assets against default, or losing value. Then, if or when some of them do lose value, the government pays the company. Again, the government can over-charge the company -- charging more than the companies eventually recoup from insurance payments, in which case taxpayers win -- or it can under-charge, in which case the companies win when they collect more insurance proceeds than they paid out in premiums. If the government gets it just right by managing to charge a perfect premium, it's a wash.

And, once again, setting the premiums too high (company loses) or just right (break-even) would in all likelihood leave the companies without any additional capital. Only if the government charges too small a premium (taxpayer loses) does the financial system as a whole see any benefit in the form of additional capital.

"I think that in the end, the fundamental choice is really between whether or not you price risks at accurate levels or whether you try to inject capital," at taxpayer expense, says Lucian Bebchuk, a professor of law, economics and finance at Harvard. (Bebchuk has his own proposal, based on Treasury paying fair market value for troubled assets, buying new securities from financial institutions to provide additional capital, and requiring them to raise capital through rights offerings to existing shareholders.)

In the end, the two concepts -- government insurance vs. government investment -- aren't quite identical, of course. With the Treasury proposal, the government would put up the cash, and recoup at least some of it later by selling the assets it buys. With insurance, the companies hand cash over to the government up-front, and get at least some of it back later, when they collect on their insurance policies. Then there's the problem of setting that price or premium. One problem with the more complex securities at the heart of this mess: no one is really sure what they're worth.

"Insurance isn’t going to do it, tax incentives aren't going to do this," Ritholtz says. "If you want to do this, you've got to get cash to the banks."

Before it's here, it's on the Bloomberg Terminal.