The Obama Administration’s rumored plan to buy “troubled assets” to rescue banks presents the government with a “thorny valuation problem,” the New York Times reports.

The article gives an example of a bond held by an unidentified “financial institution” which is “backed by 9,000 second mortgages from borrowers who put down little or no money to buy homes. Nearly a quarter of the loans are delinquent, and losses on defaulted mortgages are averaging 40 percent.”

The financial institution values the bond at 97 cents on the dollar. Standard & Poor’s values the bond at 87 cents at the current default rate, but estimates the bond’s value could go down to 53 cents if the default rate doubles. But someone actually bought one of the bonds recently. The purchase price was 38 cents.

According to the article, financial industry critics “say that the banks’ accounting for those assets cannot be trusted because they have an incentive to use optimistic assumptions.”

“Optimistic”? Whoever paid 38 cents on the dollar for one of those bonds is a giddy optimist. The financial institution’s valuation of 97 cents on the dollar is pretty clearly fraudulent.

Almost half of home mortgages in 2006 had “piggyback” second mortgages. The first mortgage was typically for 80 percent of the value of the home, and the second was for the rest of the loan. So if a borrower had no equity, the first mortgage was for 80 percent of the value of the home and the second was for 20 percent. The lender then sold the first and second mortgages separately to investment banks like Bear Stearns and Lehman Brothers, which sold bonds on pools of mortgages, like the one in the Times’ article.

Here’s how it all works: the holder of a first mortgage gets paid everything before the holder of a second gets paid anything. If the holder of the first forecloses, the proceeds of the sale pay the foreclosure costs, then the first mortgage, then the second—if there is anything left. If the holder of the second forecloses, the holder of the second pays all foreclosure costs and the holder of the first gets paid in full from foreclosure before the holder of the second gets anything.

Home values nationwide have now fallen 25.1 percent from the peak, which is probably when the second mortgages that backed the bond in the Times’ article were made.

In short, most second mortgages effectively now have no collateral. Holders of second mortgages are unsecured creditors, just like credit card companies.

The Times article said that based on Standard and Poor’s computer models, “at the end of September financial firms had $600 billion in such hard-to-value assets.” That’s not the amount that assets are overvalued, that’s the current value of assets that are “hard-to-value,” so the assets might be overvalued.

Yeah, that’s probably optimistic too.

Goldman Sachs economists estimated a couple of weeks ago that losses from all debt—residential mortgages, credit cards, car loans, commercial real estate, business loans—was now $2.1 trillion, and only $1 trillion of that had been written down by the financial institutions holding the debt. “Loss recognition by US financial institutions still has a long way to go,” a Goldman economist said.

Not surprisingly, Nouriel Roubini is less optimistic still. Dr. Doom estimates that losses for U.S. financial institutions could reach $3.6 trillion, about half of that losses to banks and broker dealers. “If that’s true, it means the U.S. banking system is effectively insolvent because it starts with a capital of $1.4 trillion,” Roubini said. “This is a systemic banking crisis.”

So who can we believe? Financial institutions’ own valuations of their assets? Standard & Poor’s computers? The price actually paid for an identical asset in a single recent transaction? Goldman Sachs’ economists? Dr. Doom?

Some prominent economists appear to put more faith in Roubini’s analysis than in financial institutions’ valuations. Joseph Stiglitz, a Nobel laureate in economics, calls the rumored plan to buy such assets swapping “cash for trash.” Paul Krugman, also a Nobel laureate, calls the plan “Wall Street voodoo”- “the belief that by performing elaborate financial rituals we can keep dead banks walking.”

I don’t claim to know, but for some time now I have thought that financial institutions were acting like they knew something we didn’t. Some business practices have not made any sense for institutions that are safe and sound.

In the late 1980s, I represented an Iowa savings and loan in a small lawsuit against a North Carolina bank. The Iowa savings and loan held a mortgage on a house and lot in rural North Carolina. The borrower stopped paying, and the savings and loan foreclosed. When the savings and loan inspected the property, they were surprised to find an empty foundation. It turned out that the house was a manufactured home. The bank had loaned money to the borrower to buy the manufactured home when the home was still sitting on the dealer’s lot. The borrower had stopped paying the bank too, and the bank repossessed the manufactured home, jacking the home off the foundation and towing it away.

The savings and loan didn’t know that the borrower owed money to the bank, and the bank didn’t know that the borrower owed money to the savings and loan. And both were counting on the same home as collateral.

The lot was worth about $30,000 less with an empty foundation than it had been with the manufactured home on it, which meant that the value of the lot was not nearly enough to cover the debt to the savings and loan.

There was no North Carolina case on whether the savings and loan had the first right to the manufactured home as collateral or the bank did. I found a handful of cases from other states, and they split maybe sixty percent for the bank, forty percent for the savings and loan.

So I filed a lawsuit for the savings and loan asking the court to make the bank pay the savings and loan the $30,000 difference. I included a claim for unfair and deceptive trade practices, which had a pretty vague definition in North Carolina at the time. A judge could award triple damages for an unfair and deceptive trade practices claim. The claim wasn’t frivolous…exactly…but it was aggressive.

The bank’s lawyer did the same research and found the same cases. He and I agreed that it made sense for the bank and the savings and loan just to split the difference, rather than fight it out in court. So I called my contact at the savings and loan and told him that the bank had offered to settle for $15,000, and I thought they should take the offer. He said that the savings and loan was carrying the lawsuit on their books as a $90,000 asset, so a $15,000 settlement would show on their books as a $75,000 loss. That valuation assumed complete victory, including triple damages, which was pretty unlikely.

I argued that $90,000 was wildly unrealistic, that the settlement offer was a bird in the hand, and that settlement would save litigation cost. He wouldn’t budge. I was puzzled, but I had no choice but to go forward with the lawsuit.

I litigated valiantly, and lost.

A few months later I got a letter from the Resolution Trust Corporation. They had taken over the savings and loan, and wanted to know more about the lawsuit.

The savings and loan’s decision not to settle the lawsuit made no economic sense for a solvent institution, but it made perfect sense if their principle objective was to maintain the false appearance of solvency for as long as possible. The savings and loan was undoubtedly inflating all of their assets, including my homely little lawsuit, to postpone the inevitable.

What reminded me of that incident from my late, unlamented law practice was the persistent failure of financial institutions to modify mortgages voluntarily. It makes perfect economic sense for a safe and sound institution to avoid the ruinous costs of foreclosure by agreeing to reduce the principal and monthly payment for homeowners who can pay a mortgage, but not the one they’ve got. But according to the National Association of Consumer Bankruptcy Attorneys, fewer than ten percent of mortgage modifications in November reduced the principal. About half added late payments and penalties to the principal, and either increased monthly payments or added payments at the back end of the mortgage. If a borrower was in default already, what’s the chance the borrower can make a higher monthly payment?

Lenders are quick to say that they have every right to full payment, that they don’t have to agree to modify mortgages. The savings and loan had every right not to settle their lawsuit, to let a court decide their claim against the bank. But refusing to do what makes obvious economic sense is suspicious. If lenders agreed to modify a mortgage to reduce the principal, they would have to change how they value the mortgage as an asset. How do lenders value a modified mortgage that does not reduce the principal, but that is destined for default?

John Kenneth Galbraith wrote that embezzlement is “the most interesting of crimes” for an economist. Embezzlement is almost always eventually discovered, but for a time results in “a net increase in psychic wealth,” when the embezzler “has his gain” and the victim doesn’t miss it. Galbraith called the undiscovered and therefore unfelt loss “the bezzle.”

According to Krugman, the stock in banks that are solvent only by virtue of an “optimistic” valuation of their assets “isn’t totally worthless,” but the stock’s value is “entirely based on the hope that shareholders will be rescued by a government bailout.” The “huge gift to banks shareholders at taxpayer expense,” Krugman said, was likely to be “disguised as ‘fair value’ purchases of toxic assets.”

So maybe insolvent banks are stalling for time, hoping that the economy turns around, that home prices will go back up, or that sick borrowers will get well and unemployed borrowers will find jobs. Maybe they want to enjoy the “psychic wealth” of paper solvency for as long as possible.

And maybe they’re hoping we’ll buy their bezzle.