I just came back from the ASF 2012 conference in Las Vegas. I have to admit, the first day (Sunday) was really all about football, but it was all business afterwards. Generally there seemed to be lot of activity and optimism this year (perhaps helped by the fact that the Giants are making another Super Bowl appearance). All the sessions appeared to be well attended, including the Wednesday morning sessions.
One session that was particularly of interest to me was on Securitization Data and Analytics Innovations. It was lead by Ned Meyers, SR. VP from Lewtan. On the panel were reps. from Bloomberg, Equifax, 1010Data, Morningstar, and LPS. Following are some highlights from the session:
- Equifax stated generally the biggest lift in the default models came from updated credit scores. I think this is little bit contrary to what Laurie Goodman has been touting which is that CLTV is the biggest determinant of default.
- Lewtan stated accurate property level valuations are critical for estimating losses. This may be obvious to those outside of the RMBS world, but identifying property level data has been unavailable until companies like us were able to figure out a way to match the data and is a fairly new best practice that is being adopted by more and more investors in this space.
- LPS stated that they definitely saw discounts on REO resales that helped estimate losses. Their method for estimating REO discounts is to use non-distressed HPI and compare against distress only sales.
- LPS stated according to their analysis the REO discount rate was generally the same across all geography. Our analysis, however, shows that there could be differences based on geography and is a function of market saturation.
- LPS displayed zip code level HPI heat maps of LA and NY with a simple message that MSA/CSA level HPI should not be trusted. It’s funny that these were the maps we were demonstrating two years ago. It’s good to see that our work is being validated by others as well.
- It appears that rating agencies are now taking a deeper interest in analyzing loan level and even property level data supporting the residential bonds that they are rating.
I would love to hear your impressions of the conference if you were able to attend.
Go Giants!