The Treasury unveiled (PDF) new data last Friday showing the progress of its major foreclosure prevention program. Treasury officials touted the new data as evidence that pressure on mortgage servicers to improve performance was working, but the numbers continue to show real problems. And though there is a wide disparity in performance among the participating servicers, the four largest all continue to lag.

First, a look at the overall numbers: 66,465 homeowners had received permanent modifications as of Dec. 31. That’s up from 31,382 reported last month. On a conference call with reporters Friday, Treasury official Phyllis Caldwell called that a “significant acceleration of the rate at which borrowers are being approved.” That may be, but it’s still well below where it should be if the program were working properly.

Under the program, homeowners are first given three-month trial loan mods. About 900,000 homeowners have begun such trials, which are supposed to give borrowers enough time to turn in all their documentation and demonstrate an ability to make the lower monthly payments. But as we’ve reported, trials have been stretching much longer. The servicers have largely blamed homeowners for this, saying that they’ve been slow in turning their documents in. Homeowners, meanwhile, routinely complain that servicers lose documents and make mistakes.

In response to all those problems, the Treasury has allowed servicers to lengthen some trial periods to as long as five months. But trials have stretched still longer. As we’ve reported, services have kept some homeowners in “trial” purgatory for as many as nine months.

The numbers reflect that logjam. In October, the Treasury reported (PDF) that servicers had begun a total of 487,081 trial modifications. But three months later, less than 24 percent of those loans have had any sort of final resolution. In other words, 76 percent of the trial mods started three or more months ago are still in limbo.

As our chart shows, just 66,465 trials have been converted into permanent modifications, and 48,924 trials have failed – a total of 115,389. The Treasury officials made a point Friday of saying that some homeowners, about 46,000, have been approved for a permanent modification, but haven’t yet signed the paperwork to make it final. Even including that number, only one-third of the homeowners who began trials three or more months ago have had final answers.

Our chart also shows the wide divergence in performance among servicers. To show how the borrowers in trials are faring at each servicer, we’ve shown each servicer’s number of permanent modifications and failed trials as a portion of the trial modifications they had begun as of three months ago. Overall, only about 14 percent of the trials begun three or more months ago have been made permanent. Ten percent have failed.

The big names are among the worst-performing servicers. Bank of America, JPMorgan Chase, CitiMortgage and Wells Fargo together account for more than 60 percent of the 3.4 million mortgages eligible for the program. All four have converted a small percentage of the trials begun three or more months ago into permanent modifications. The highest is Wells Fargo, with only 13 percent.

The worst among them is by far the largest: Bank of America. The data show the bank is moving a relatively low number of homeowners into trials (206,775 homeowners have received trials, while over 1  million are eligible, according to Treasury) and then converting very few of those to permanent modifications. Of the trials begun by Sept. 30, it has only converted 3,305, or about 3 percent, to permanent modifications. In a press release Friday, Bank of America said that it had offered permanent modifications to many more (about 12,000). If so, those numbers will be reflected in the numbers released next month.

There doesn’t seem to be a good example of a model servicer in the program. The closest would seem to be GMAC Mortgage. The numbers show that it has moved about half of its eligible loans into trials, and then converted about half of those into permanent modifications. No other servicer has performed quite so well in both categories.

Some servicers have moved very few homeowners into the trial stage, but have converted a high percentage of them into permanent modifications. Carrington Mortgage Services, for instance, had begun 584 trial modifications as of the end of September. That’s a very small portion of its eligible loans, suggesting a high barrier of entry to the program. But Friday’s report shows that 608 homeowners have received permanent modifications, meaning the company is, on average, offering trial participants a permanent modification in under three months.

One other servicer, Ocwen, a company that specializes in troubled loans, has performed similarly. Its executive vice president Paul Koches told ProPublica that Ocwen’s high conversion rate is a result of requiring documents from homeowners before they enter the trial stage – as opposed to some other servicers that have accepted homeowners based on what they say their income is. The Treasury program allows for both approaches. Koches disputed the idea that Ocwen was putting few (20 percent) of its eligible homeowners into trials, as the data produced by Treasury suggests; the numbers do not reflect Ocwen customers who are on in-house modifications, he said, and reflect a large volume of loans that it began servicing only at the very end of the year.

Some companies, meanwhile, have a remarkably high failure rate on trials. Two servicers, Litton Loan Servicing and Aurora Loan Services, have denied more than twice as many homeowners in trials as they have approved for permanent modifications. We’ve asked both companies for explanations, but trials can fail for a number of reasons: a homeowner might not make the trial payments (Treasury official Michael Barr said Friday that overall, 25 percent of homeowners in trials were defaulting), the homeowner might never have turned in the necessary documents, or the homeowner might not meet the program’s requirements.

Asked Friday about the disparities among servicers, Barr said that some are “just doing a better job.”

Have you applied for a loan modification under the Obama administration’s Making Home Affordable program? Are you thinking about it? If so, we at ProPublica want to hear your story.