Mortgage modification datapoint of the day, Ocwen edition

By Felix Salmon
October 13, 2009
Shahien Nasiripour does some more digging into the HAMP mortgage-modification figures today, following on from last week's analysis. This time he's looking at the percentage of trial loan mods which have been converted to permanent status, and the numbers are startling, to say the least:

" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">

Shahien Nasiripour does some more digging into the HAMP mortgage-modification figures today, following on from last week‘s analysis. This time he’s looking at the percentage of trial loan mods which have been converted to permanent status, and the numbers are startling, to say the least:

Total number of trial modifications at the end of May: 50,130

Number of those being serviced by Ocwen: 1,058

Total number of permanent modifications as of September 1: 1,711

Number of those being serviced by Ocwen: 763

To put it another way, of the Ocwen has managed to convert more than 72% of its end-May trial loan mods to permanent status. The equivalent number for everybody else? 1.9%.

Ocwen’s looking good on other fronts, too, such as its redefault rate, which is much lower than the rest of the industry.

Is there any way to import Ocwen’s best practices to other servicers? Shahien quotes Valparaiso University’s Alan White as saying that Treasury “should start firing the under-performing servicers and bidding their work out to the successful companies”; what’s more, White is a law professor, which means that might even be legal.

More helpfully, Shahien explains one big difference between Ocwen and everybody else: Ocwen insists on having proof of income before it starts any trial modification. (And this doesn’t seem to have slowed it down at all; quite the contrary.) Other servicers might do well to follow suit.

Comments
One comment so far

http://www.google.co.in/

Posted by tomharry9090 | Report as abusive
Post Your Comment

We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see http://blogs.reuters.com/fulldisclosure/2010/09/27/toward-a-more-thoughtful-conversation-on-stories/