Wednesday, January 30, 2008

Question About Mortgage Loans

*** Edited to add a conversation with the GZA (at the end) ***

Maybe this is a really stupid question, but if anyone knows what the deal is, I would very much like to hear what you have to say. The question I have is about jumbo vs conforming mortgage loans.

Right now the conforming limit is 417K (and current talk of boosting this as high as 730K). So, any loan above this amount would be considered a jumbo, and therefore would be subject to significantly higher interest rates. This makes perfect sense, because larger loans would definitely carry more risk. Thus, lenders have the right to be compensated for the increase.

However, the notion of a hard cut-off seems really dumb to me. Why wouldn't there be many shades of risk which are then subject to a gamut of interest rates tied to each shade? Then, we would more or less have a weighted average of the risk and as a result a weighted average of the proper interest rate.

It seems stupid to me that a 415K loan can be had for 1% less than a 550K loan. What would make sense to me would be something more like a progressive schedule, with tiny incremental adjustments. Say we used 5 or 10K slices. First 10K is very low risk, and thus is given the very best interest rates... maybe 5%. The next 10K adds a little more risk, and maybe gets 5.02%. By the time you get to the 55th 10K slice, that part is much riskier given that in terms of an instantaneous snapshot, you've already borrowed 540K, and so it is subject to a much higher rate, say 6.89%.

The end result would be that you'd have a weighted rate for your loan.

Now, I'm just talking out of my ass right now. And, I'm thinking more like an engineer. Often times it's better to take a weighted average and apply a weighted rule set rather than applying completely new rules at some arbitrary cutoff point. You simply lose continuity.

I mean, qualifying criteria for a Roth IRA utilize a weaning off schedule (AGI < 99k =" full" 114k =" weighted"> 114K = no contribution). Why can't mortgage loans follow a similar model?

Anyone with more knowledge want to share with me why this general idea is not viable? I certainly would expect a lot of kinks to be worked out, but the overall concept of a risk-weighted loan makes a lot more sense than what we have today.

*** Edit #1 ***

Some spots are paraphrased/edited for clarity.

GZA: I saw your blog entry. It's not just that the jumbo loans have more risk from being larger.
GZA: Conforming loans can be bought and resold by Freddie Mac, where as the jumbo loans can't.
BruteForceX: right, so the thing is... why can't they figure out a way so that loans can somehow be partially conforming, partially not, etc. in terms of securitizing
BruteForceX: can't they break up loans into smaller chunks/slices? Or is that just too complex?
GZA: So you want to complicate the calculation of loan interest rates even further :)
BruteForceX: if it makes things less arbitrary... definitely! haha
GZA: Don't you think that shit is ridiculous enough to understand for the average person?
BruteForceX: well... that's for the professionals to figure out. The bottom line is that you will hopefully see your rate as a weighted avg. Who cares how your loan is bought and sold behind the scenes?
BruteForceX: You, as an avg person, should be oblivious to the behind the scenes shit. All you know is that a 450K vs 415K loan is not a HUGE interest rate diff, because that's plain stupid to have a discontinuity in the rates
GZA: Yeah. I hear you.


Okay, so I still don't know why there isn't some mechanism/model that prevents this sudden change in interest rates at arbitrary cutoff points. Still makes way more sense to me that there should be a smoother curve.

No comments: