cancel
Showing results for 
Search instead for 
Did you mean: 

FICO 2008: Better, But Not Good Enough

tag
Anonymous
Not applicable

Re: FICO 2008: Better, But Not Good Enough



cheddar wrote:


700Man wrote:

Do those GUARANTEE you'll be a good credit risk? Of course not. But they increase the odds tremendously.


What evidence is there of this besides the fact that it just "seems" like it should be the case?  Do you think the numbers crunchers at FI never considered this?  Wouldn't you imagine that all sorts of variables have been run through the wringer in order to generate the scoring algorithm?
 
I guarantee that FI has better evidence regarding the predictiveness of any particular variable than you or I do.

 


Well said Cheddar.
Message 11 of 92
Anonymous
Not applicable

Re: FICO 2008: Better, But Not Good Enough



700Man wrote:
Guys:

(1) If you work for Goldman Sachs instead of Mike's Financial Services, sorry, you're a better credit risk IMO working for Goldman Sachs.


Interesting example, but I'm curious why you chose Goldman Sachs to illustrate your point.  Why not some other large, publicly traded company like Bear Stearns?
 
Message 12 of 92
Anonymous
Not applicable

Re: FICO 2008: Better, But Not Good Enough



700Man wrote:
Sidewinder, income most absolutely SHOULD be included in a credit score. How can you say that income doesn't relate to willingness/ability to pay? That's absurd.
 
Income has NOTHING to do with willingness to pay. Yes it has to do with the ability to pay, but the lender judges that. Your credit history and/or score has nothing to do and is not there to reflect your ability to pay, it is there to reflect your willingness to pay. Credit reports and scoring is all about how you MANAGE your money not how MUCH money you have.

Yes, a guy making $500,000 can decide to stiff all his/her credit card companies on a lark. And someone making $35,000 can keep making minimum payments that consume 30% of his/her after-tax paycheck. BUT HOW LIKELY IS EACH ONE ?
 
How likely each is to happen would require a great deal of research. More research than I have time for. However, income has nothing to do with how likely each of those are to happen. A persons circumstances and morals have to do with that.
 


As I said above, this is about PROBABILITIES and forecasting LIKELIHOODS. Income most certainly should be a factor in FICO scoring since, in fact, the best scores are positively correlated with income (and also assets, home location, etc).
 
What do you have to back this theory? I know people who make WAY more than I do and have as bad or worse scores than myself. In fact one of them owns a substantial amount of land and other assets.I also know people who make less than I do and have way better scores than myself.

Do those GUARANTEE you'll be a good credit risk? Of course not. But they increase the odds tremendously.
If you don't think so, ask yourself this: if you were a lender, and were told to make a home mortage loan 100% TOTALLY BLIND, and the only thing you knew was that one applicant lived and wanted a mortgage for his/her house in the South Bronx, and the other wanted his/her mortgage for Great Neck, Long Island (a very posh place), which would YOU lend to ????
Unrealistic. Now you are basically saying where one lives has something to do with their credit risk/willingness to pay?
 

Message Edited by 700Man on 04-25-2008 09:01 PM

Why not include education level, parent's education level, your quality of life growing up, how many kids you have, your marital status, etc in with scoring?
 
By your reasoning, all of these factors would affect your credit risk.
 
Someone making $100,000 with 5 kids should be the same risk as someone making $25,000 with no children, right? It does cost a lot to raise children.
 
If you went to college or even graduate high school that means you are more financially responsible than someone who didn't, since you are more educated, correct?
 
Need to consider parent's education level as well, that way you can judge whether the parents were educated enough to teach their children financial responsibility.
 
Let's not forget your childhood quality of life. Everyone knows if you grew up with little valuables and didn't have many luxuries that you are gonna strive to have those things as adults and for your own children, therefore you will be more likely to run up giant amounts of debt buying those things and then spend your income on more things instead of bills.
 

 


Message Edited by sidewinder on 04-25-2008 09:45 PM
Message 13 of 92
Anonymous
Not applicable

Re: FICO 2008: Better, But Not Good Enough

KUDO'S TO SIDEWINDER!!!!!!!!!!!!!!!!!!!!!!!!!!
Message 14 of 92
Anonymous
Not applicable

Re: FICO 2008: Better, But Not Good Enough

Good God 700 man--are you nuts?Smiley Surprised


Message Edited by casinoannie97 on 04-26-2008 07:53 PM
Message 15 of 92
Anonymous
Not applicable

Re: FICO 2008: Better, But Not Good Enough

"What evidence is there of this besides the fact that it just "seems" like it should be the case? Do you think the numbers crunchers at FI never considered this? Wouldn't you imagine that all sorts of variables have been run through the wringer in order to generate the scoring algorithm? I guarantee that FI has better evidence regarding the predictiveness of any particular variable than you or I do."

Please...if FairIsaac's system were that good, they wouldn't be updating it.

Not to mention that their scoring system was totally given the end-around by savvy scammers who manipulated the scoring system thereby qualifying for mortgages they had no business even sniffing. Someone with FICOs in the 500's getting a $700,000 mortgage? Please....

I watched the entire housing bubble come to fruition because companies like New Century, Novastar, Freemont, etc, enabled and contributed to the lax FICO scoring with dubious lending standards.

At the end of the day, FairIsaac is a company run by fallible individuals, the same ones who stuck tens of billions in CDO's on the books of Bear Stearns, Merrill Lynch, Citibank, and AIG. Those were people whom you would no doubt say had much better evidence regarding the safety of those instruments than you or I.

WRONG !!

If I were a lender, would I look at a FICO score? Sure.

Would I trust it? No.
Message 16 of 92
Anonymous
Not applicable

Re: FICO 2008: Better, But Not Good Enough

No, CasinoAnnie, I'm not nuts -- I'm someone who's worked in banking and has never had to report credit losses because I do focus on the stuff that I mentioned, that's all.
Message 17 of 92
Anonymous
Not applicable

Re: FICO 2008: Better, But Not Good Enough



700Man wrote:

Not to mention that their scoring system was totally given the end-around by savvy scammers who manipulated the scoring system thereby qualifying for mortgages they had no business even sniffing. Someone with FICOs in the 500's getting a $700,000 mortgage? Please....


Think about what you just said, and notice the contradiction.
 
If those people's FICOs were in the 500s, how exactly was scoring system manipulated?
 
Perhaps lenders were approving loans for people who didn't have the FICO scores to qualify for them.  I'm not sure how that is the fault of the scoring system, though.  If people in the 500s were being approved for $700K mortgages, then it was the lenders who completely ignored the fact that they had FICOs in the 500s!
 
If the scoring algorithm assigns someone a score in the 500s, but lenders ignore the warnings that should go up with that kind of score, it does not mean the scoring algorithm didn't do its job.
 


Message Edited by cheddar on 04-26-2008 08:22 PM
Message 18 of 92
Anonymous
Not applicable

Re: FICO 2008: Better, But Not Good Enough

Cheddar, I mis-typed....I meant they were really people whose scores SHOULD have been in the 500's or 600's and instead were graded closer to 720-750. People games the scoring system, lied, committed fraud, and also used that 'authorized signer/borrow someone's credit score' rig quite frequently since 2005.

I spoke to the president of a large local bank. He told me that FICO used to be EXTREMELY reliable for years -- decades, really -- when credit was much less ubiquitous, people felt an obligation to pay off debts, and they didn't go overboard. But the whole thing began to unravel bigtime about 2002-03 when lending and scoring systems were gamed.

I don't blame Fair Isaac -- but I do think that they would be doing the right thing for both good and bad credits by getting the scores more accurate. I think including more variables is a means toward that end. If the variables I cited don't work, eliminate them or underweight them. But they DO work -- and maybe that's why they aren't being included.
Message 19 of 92
Anonymous
Not applicable

Re: FICO 2008: Better, But Not Good Enough



700Man wrote:
Cheddar, I mis-typed....I meant they were really people whose scores SHOULD have been in the 500's or 600's and instead were graded closer to 720-750. People games the scoring system, lied, committed fraud, and also used that 'authorized signer/borrow someone's credit score' rig quite frequently since 2005.
 
This is part of the reason that FICO is being updated. To eliminate people using someone else's good credit in order to make theirs look more desirable.

I spoke to the president of a large local bank. He told me that FICO used to be EXTREMELY reliable for years -- decades, really -- when credit was much less ubiquitous, people felt an obligation to pay off debts, and they didn't go overboard. But the whole thing began to unravel bigtime about 2002-03 when lending and scoring systems were gamed.

I don't blame Fair Isaac -- but I do think that they would be doing the right thing for both good and bad credits by getting the scores more accurate. I think including more variables is a means toward that end. If the variables I cited don't work, eliminate them or underweight them. But they DO work -- and maybe that's why they aren't being included.

If lenders are looking at ONLY FICO scores and not actually looking at the CRs themselves........well then it is the lenders fault.
 
It doesn't matter what your FICO score is, a manual review should ALWAYS be done.
 
FICO has nothing at stake, they aren't lending the money based on the scores. So why would they not include variables that you say do work to better predict a person's probability of not defaulting.
 
Do you think lenders should be able to just say ok give me your name, dob, SS# and let me pull your credit score and I'll base my decision?
 
Absolutely not, they ask for income, they ask for family size, they ask for employment history, they SHOULD investigate these things and make sure they are accurate.
 
FICO scores are not the sole factor in deciding who to lend to and who not to lend to and shouldn't be used as such.
 

 
Message 20 of 92
Advertiser Disclosure: The offers that appear on this site are from third party advertisers from whom FICO receives compensation.