Have you ever put any thought into what actually counts in your FICO scores? Furthermore, have you ever wondered why some things count and why some don't? Everyone who knows anything about credit scoring knows that your payment history, your debt and your credit shopping habits "count." That's not rocket science and you certainly didn't need me to tell you this. But there's more to it than meets the eye. There's a darn good reason why some things actually count toward your credit scores and why some things don't count. Join me…
- Your payment history – Okay, this is a no-brainer. Everyone knows this counts and counts a lot. So tell me why it counts? I'll save that little nugget of gold for later.
- Your amount of debt – Another no-brainer. This counts almost as much as your payment history. Why?
- Your credit applications, the age of your credit file and the mix of accounts - Again, you can find all of this at any reputable credit scoring website. We all know it counts. Why?
All of these things are what I like to call "the company line." It comes straight from the Public Relations folks at the credit bureaus and Fair Isaac. Let's talk about something else that doesn't get so much attention…
- Your age - You can certainly figure this out from a credit report. Your date of birth is on your reports and if you can subtract you can figure out someone's age. Does it count? Yes or no? Take a guess. I'll say that it does NOT count and I'll say that it doesn't count because age is illegal to look at when determining credit risk. What's your answer? I'll give you a hint. The answer will surprise you and it doesn't have anything to do with being legal or illegal.
Okay, I'll quit teasing you with questions and start giving some answers. The word that you need to become familiar with is "predictive." All of the things I listed at the beginning of this post (payment history, debt, etc) are all predictive of something. And since we're talking about credit scores that "something" is credit risk. They are all predictive of future credit risk.
What that means is that people who don't do well in these categories are more likely to miss payments in the foreseeable future. And, their scores will suffer because of it.
How do the credit scoring companies determine what's predictive and what's not? It's not really that hard. They use a process called regression where they simply look in the past to determine what people who miss payments all have in common. Guess what they have in common…payment history problems, a lot of debt and excessive credit shopping. Voila!! That's one mystery solved. Let's tackle another one.
What about your age? Does it count? Nope, it doesn't. Why not? It's certainly not illegal to count your age when determining risk. Did I say that? Sure did. Take a deep breath and after you get done dog cussing me I'll explain why it's not illegal. Think about when you apply for auto insurance. Better yet…guys…think about when you were 22 and you were single and you paid for your own car insurance. It was pretty expensive wasn't it? Ever wonder why? It's because you were in a high risk group.
What happened when you turned 25, got married, moved to the burbs and sold that GT Mustang and started driving a Buick? Your insurance premiums dropped off the table because you went from a high risk group to a lower risk group. You didn't seem too worried about your age being counted against you as a way of assessing your insurance risk.
It's certainly not illegal to use age to determine credit risk but it's a public relations nightmare. Think about it…what if an 87 year old applies for a 30 mortgage? Do you honestly think he is going to be able to pay it off over 30 years? Of course he isn't. But, you can't hold his age against him despite it being very predictive of him defaulting on that loan someday. That's because he is a part of a protected class…the elderly.
You can certainly count his age against him unless he's in a protected class. In that case you have to build your model so that it gives him more points than the statistics say you should. So basically you have to do the opposite of what your model tells you to do. Not a good thing.
So you add those two things together and you get a characteristic that's predictive and legal to look at but is a loser from a PR perspective.
Plus, there's more than one way to skin a cat, isn't there? How do I wink on this keyboard…oh there it is…semi-colon, closed bracket… ;)
What if I looked at your credit report and I took the oldest account on it and then added 18-25 years to the date it was opened? I bet I can get pretty darn close to your age without having to actually ask for it.
I have a Citibank account that I opened in 1988. It was my first student credit card. That was 18 years ago. Add 18-25 years to 18 and you get 36-43 years old. Guess what…I'm 37. I figured out how old I was without actually telling anyone. Why can't a credit-scoring model do the same thing? It can. Do they? I can't go there. Not allowed.
There's more than one way to skin a cat. ;)