Treasury REALLY wants to promote this idea of TARP as a success, don’t they? They released yet another report yesterday on the cost of the bailout, with the number whittled down to $29 billion.

The Treasury Department expects to lose $29 billion on the federal bailouts stemming from the financial crisis, with most of the losses in its housing finance program and the auto rescue.

In a report released on Tuesday, the administration said it expected a $17 billion loss from its investments in General Motors, Chrysler and the auto finance companies, as well as a $46 billion loss from housing programs like the mortgage modification program known as the Home Affordable Modification Program.

OK, it’s time to go to the internals of this report, because the last time anyone looked at the spending on HAMP, it was the Inspector General for TARP, and he put that spending at about $500 million. And actually, the report acknowledges this. So they expect to spend $45.5 billion more in the next few years on HAMP, when the pickup for new modifications is dropping? I know that Treasury expects to pay out incentive payments to lenders over time if modifications become permanent, but that level of spending would require something close to the 4 million mods they estimated at the beginning, when the program for now looks stalled out at 400,000 permanent mods, and advocates have such poor feelings about it that they have begun to encourage their clients not to use it.

And people are quoting from this?

(Note too how they shaved off $4 billion from HAMP without anybody noticing; the expected commitment was $50 billion.)

But here’s the nut of the issue: none of this TARP outlook includes the losses that will incur on Fannie Mae and Freddie Mac:

Recently, the Congressional Budget Office put the cost at $66 billion. And, after the bailout, the government will still be left with its investments in Fannie Mae and Freddie Mac, the government-backed housing finance companies. The report said Fannie and Freddie were expected to cause “substantial losses,” but it noted that they were financed using other funds, not the troubled asset funds.

I’m trying to decipher this, or rather why it matters. Fannie and Freddie losses come as a direct result of asset purchases on toxic loans – in other words, TARP as it was originally envisioned. The banks off-loaded their crap onto the government, and we paid the price. Basically, creative accounting is making TARP look better. Incidentally, the bailouts of Fannie and Freddie are unlimited.

But don’t worry, a “substantial part” of those “substantial losses” will be offset! How substantial? This is where the report goes vague.

The Treasury Department today said it expects “to incur substantial losses” from the two government-sponsored mortgage companies, which have operated under federal control since 2008. But the administration believes a “substantial part” of these losses will be offset by income from two other areas, according to an evaluation of the Troubled Asset Relief Program it released today. (Somewhat confusingly, the government assistance to Fannie Mae and Freddie Mac had nothing to do with TARP).

First, Treasury said a 2008 law that permitted it to buy $200 billion in mortgage-backed securities is bringing in “notable returns.” It wasn’t more specific.

Also, “as a result of its emergency financial programs,” money from Federal Reserve operations that is shifted to the federal budget has “increased sharply in 2009 and 2010, and they are projected to remain elevated for some time.”

This reveals the true bankruptcy of this report. The TARP cost $29 billion, and the hundreds of billions of losses in Fannie and Freddie don’t count, because the money was separate, even though it was for the exact same purpose. But don’t worry about Fannie and Freddie, because we have these unnamed “other programs” that will offset the cost in unnamed ways.

Either this entire set of emergency lending to the banks is part of a whole program, or it isn’t. You can’t slice and dice the numbers when convenient to come out with a preferable outcome.

The main take-away from these numbers is that the banks got their money very quickly, while a year and a half later, the spending to help homeowners has reached a whopping 1% of what was alloted.