When Is A Score Not A Score?
It appears to me that one fundamental factor has not been discussed in the thread. Let’s assume UltraFICO is truly an option as stated in the various articles. And let’s further assume that this new model is actually “…a secondary scoring model that is only used when you would have originally already been denied.” Then how does a lender compare scores? How does a lender set a true cutoff point? This 700 is okay, but that 701 has been “Ultraed” and not good (or vis-a-versa).
Let me explain. We all know that FICO currently states that “Payment History [is] 35% of a FICO Score. Amount of Debt [is] 30% of a FICO Score. Length of Credit History [is] 15% of a FICO Score. Amount of New Credit [is] 10% of a FICO Score [and] Credit Mix [is] 10% of a FICO Score.” This makeup is 100% of an individual’s FICO score; however, according to the article, the new model will also include “Current checking balance, Length of checking history, Transaction frequency, [and] Overdraw history.” If these items have any scoring value at all, then some input variable’s value above must be reduced, or the scales must be different – it’s simple math and logic.
The first case means that one individual’s score may have different (additional) variables when compared to another person’s score. These two scores cannot be compared to each other nor can they be compared to a common baseline – there is none. In the second case it is simply comparing apples to elephants.
I believe this is a very important point for lenders. If you and I have a 760 score, any lender can be assured that we present the exact same risk to them. Sure the way FICO arrived at the score may (and most likely will) differ; but our credit risk (based on credit score) is the same to the lender. Now let’s say the trigger for Ultra is 670/671. 670 and below trigger the Ultra factors while 671 and above does not. If you have a 671 you will have the old risk value, but my 670 will provide something else (let’s say 1 point). Now to a lender, is your 671 the same as my 671? Do we both represent the same risk factor to the lender? I suggest we don’t and this is where the problem lies.
For example, what if my trigger gives me a 10-point boost to 680; am I a better risk than you at 671 (especially if the lender’s cutoff is say 675)? What if your low because you have a young, thin file; while I’m low because of a late some time ago with a thick file. Since we don’t know the scale, divisions or weight of adjusted variables, no one knows the answer and I don’t think lenders will like (or use) this model. Ultra seems to have a fatal flaw that could potentially increase the risk to lenders or lower the effectiveness of computerized qualification technology. Nonetheless, as Denise Miller often says, “But then again I could be wrong.”
Y
