| sebastian_dangerfield |
12-18-2014 11:52 PM |
Re: Good White People
Quote:
If you're not using the individual's creditworthiness to perform your credit analysis, and instead increase the interest rate based on the neighborhood's characteristics, then what are you doing?
|
Reverse redlining. I get it. I'm not avoiding that acknowledgment.
Quote:
If you're saying that houses in certain neighborhoods tend to lose their value more than in other neighborhoods, then all you're telling me is that your model for determining how much the underlying property is worth is incorrect.
|
If I'm telling you that one area sees more f/cs, I'm telling you that area sees more chance of lower priced comps fucking up valuations. This increases two forms of risk: 1. Poor f/c recovery; and, 2. Higher likelihood of buyer being underwater or holding poor equity, which equals higher default rate.
No lender wants residential r/e. Good neighborhood or bad.
Quote:
If you're telling me that the average mortgagor in one neighborhood with the same credit score as the average mortgagor in another neighborhood has a higher default rate because of cultural reasons, I'm going to say that you need to prove that to me because it sounds like you're throwing bullshit at me to cover up a practice that is based on stereotypes.
|
I'm not, and I never would. I don't think an algorithm can tease "cultural reasons" out of the data. Nor does it have to do so. All it has to do is look at geography. Like a r/e agent pricing a home based on comps, it sets rates in part based on likelihood of performance, or non-performance, of similar surrounding assets.
|