Excerpts from Peter McCorkell's Testimony on 07/22/99 at FTC hearing in D.C

We have reviewed the testimony of Peter McCorkell the chief legal counsel and senior vice president of Fair Isaac. We have noted direct quotes from his testimony at the FTC hearing in Washington D.C on 07/22/99. This is in no way meant to harm anyone, but is designed to show Congress and the general public that the credit scoring system is flawed and that one of the most important individuals who makes the decisions at Fair Isaac needs to do some more homework and treat all Americans as "equals" no matter what their ethnic origin.

  FAIR ISAAC CAN'T PREDICT THE FUTURE OR CAN THEY? 

MR. McCORKELL: "Casey Stengel actually had it about right. Making predictions about the future can be kind of dangerous. But in fact, that's what credit grantors have to do. They have to make predictions about how borrowers will behave, how borrowers will repay that obligation. And when you come right down to it, there is "NO WAY" to predict with certainty how any individual borrower is going to behave, how they're going to repay. But what you can do, is that you can predict how groups of borrowers will repay with reasonable certainty".(page8 line 8-17)

OUR COMMENT: This seems to be a lot like placing people into groups with no way for a consumer to fight if they are unjustly put into the wrong group based on no fault of their own. 

 MR. McCORKELL: "You can predict that if I have a thousand customers with this characteristics, 900 of them, or 950 of them, or 990 of them will repay in a satisfactory manner and the other's won't. But nobody can predict that this particular borrower will or will not repay in a satisfactory manner".(page8 line18-23)

OUR COMMENT: How can you admit that you can't predict the future on one hand, but than state that you can predict the future on a group of people. I wonder how he would feel if he where grouped in a group that had to pay higher rates on loans, insurance, cars and so on. And the reason you where in this group was because of your common name, x-spouse, wrong data in your credit file or ethnic origin.

MR. McCORKELL: "Well, scoring models don't assume anything. All of the numbers that are in those scoring models are the result of very painstaking, detailed statistical analysis of real data".(page20 line15-18)

OUR COMMENT: One thing we found out at this meeting was that "NO" governmental agency has checked the model out for equality and fairness to all. Has the government ever taken any person or companies word on anything that could possibly hurt or effect Americans in an unfair way before checking it out and placing protective restrictions?

MR. McCORKELL: "So we get through all of this, and somebody is probably still out there saying, oh, gee, but why do you think scoring can do a better job of predicting credit behavior of estimating risks than judgmental decisions. It's real simple. It's better to count than to guess".(page35 line18-22)

OUR COMMENT: Well that would work if this was a world run with people who never made mistakes, but the reality is that we are all human and there are mistakes on peoples credit files for many reasons from human errors, creditor errors, common names and so. So in reality we are guessing and in the old system consumers had rights to protest under the Fair Credit Reporting Act. The consumer does not have the opportunity to know their Fair Isaac credit score even though it's used to determine their destiny in life. The consumer can not fight something they can't see. 

     FAIR ISAAC'S PREJUDICE/DISCRIMINATION  

MR. McCORKELL: "It means that legally age is not supposed to be used in judgmental  credit decisions because of ECOA in rate B. But they do allow it to be used in scored decisions". (page 15-16 line21-23,1)

OUR COMMENT: I thought that our Constitution protected all Americans against people or companies that discriminate because of Race, Color, Creed, Gender, Age and so on? So what I understand is that the model discriminates against younger people or at the least gives them a lower score based on their group.

MR. McCORKELL: "Well, in fact, what we found out was that "NOT SURPRISINGLY" the lower income borrowers as a group scored lower. That's why that yellow line is lower than  the other lines in the graph".(page52 line 1-5)

OUR COMMENT: Well I really don't even have to say anything to anybody about this comment I believe that it stands on it's own. Watch out low income people your group is a higher risk and subject to higher fee's and interest.

MR. McCORKELL: "If you go to the next slide, what we also found out is that at any given  score, the risk of those low to moderate income borrowers was in fact a least as great, and usually in most cases just a little bit greater, than the general population". (page 52 line 6-10)

MR. McCORKELL: "In other words, they were just a little more risky than the general population in any score. So, in fact, the scores were doing a very good job of predicting the risk  for "THAT POPULATION". (page 52 line 13-16)

OUR COMMENT: I believe what the last three comments mean is that if your in a group of low to moderate income family's your scores will be lower because everyone in your group is a higher risk based on facts that no one has ever seen. Which means everyone in that low to moderate classification is being harmed, and is based on there group and your score may not qualify you for your home or at least they will be forced  to pay higher fee's (i.e.: interest, points, term). If you are low income how can you afford to pay higher fee's than the rest of  the general population? The consumer can't so you don't get the God given right of home ownership.

MR. McCORKELL: "The next study that we did was using credit bureau information. And there, rather than looking at the low to moderate income population, we looked at consumers  in zip codes with high concentration of "MINORITIES, BLACKS and HISPANICS". And I  could get into a lot detail about why we chose the various CATEGORIES we did. But basically we looked at zip codes with those high MINORITY concentrations. One of the things we found was that those zip codes represented a little bit less of the credit bureau database than they did of the entire population, although certainly they weren't totally unrepresented, as I think a lot of people assume. And indeed there are folks in that population that simply don't use mainstream credit. They are NEVER going to have a credit report. But they're NEVER going to go to a lender that would pull a credit report in the FIRST PLACE, and secondly the degree of under representation isn't so great that it would really affect the ultimate scorecard design". (page 54 and 55 line 17-23 and 1-12)

OUR COMMENT: I believe this proves our point that the scoring model is unfair and must be discarded because it singles out certain minorities groups as high risk and certain groups who don't use credit as high risks. The method in which these humans are grouped seems to me to be unfair. Why did Fair Isaac single out Low Income, minorities, Blacks, and Hispanics? I thought that all men are "CREATED  EQUAL".

MR. McCORKELL: "That in fact that zip code definition winds up looking a lot like the low  to moderate income definition. So not surprisingly in economic terms, those high minority zip code populations look a lot like the low to moderate income populations, and so not  surprisingly we came up with very much the same results. Again you'll see the score  distribution is a little bit  lower, which again means that at any given cutoff score, you're  going to "ACCEPT FEWER OF THESE FOLKS". But if you go to the next slide, you'll see again exactly the same pattern. In this case the red line is the MINORITY population. The  dotted yellow line is the rest of the population. At any given score those folks represent a slightly higher degree of risk. So again, the scorecard  is doing an effective job of rank ordering risk".(page55-56 line 14-15, 20-23,2-12)

OUR COMMENT: Again, I state that you can not group people into groups or zip codes and be fair. You can not group people by there gender, race or age and be fair to all. The Constitution did away with this type of discrimination years ago.

  FIVE POINTS THAT FIGURE INTO THE FAIR ISAAC SCORE

MR. McCORKELL: (1) "First is previous credit performance. Has this person performed well, or have they gone delinquent on other obligations. Not surprisingly, that's the MOST PREDICTIVE piece of information we've got". (page 43 line 14-18)

OUR COMMENT: Well here is the first comment I can agree with. Previous credit performance should have some sort of predictive power with regard to how a consumer will pay his debt obligations, however. If this was true than why do we have thousands of credit reports that defy this position. We have reports that show consumers with only one trade line (which they are not even responsible for paying) have Fair Isaac scores  in  the high 700's. We even have reports with consumers who filed bankruptcy within the last 4 years with scores again in the 700's. Then we have consumers with "NO DELINQUENT" accounts of any kind on their credit file and they have scores in the 500's.

 MR. McCORKELL: (2) "And the next most predictive piece of information is what's labeled, current level of indebtedness. In one sense, you can sort of think of that as how close to the edge are these folks. Is it somebody -- you know, somebody who has got five credit cards, all of them they have borrowed right up to the max and they're making minimum payments, are a lot riskier than somebody that their balance -- the limit ratio is down in the 10 to 20 percent range". (page 44 line 10-18)

OUR COMMENT: Again this might work if all creditors reported the same way and I tell you that they "DON'T". If the scoring model is looking at the "high credit to balance ratio" most if not all consumers are being "GREATLY" harmed. Some creditors report HIGH CREDIT as the following (i.e.)  the balance, zero, unknown, the most ever charged, these are just a few examples. Also, the bureau's don't report on consumers credit file if they pay off there credit cards each month, so how can the model know that you are only making minimum payments? The National Credit Bureau's can't be  held accountable for this, only the creditor must be held accountable. This is the reason why you can't make a "COMPUTER" do the job of a human. 

 MR. McCORKELL: (3) "The amount of time that credit has been in use. And actually that line is a bit mislabeled. It probably ought to say sort of the amount of credit history available, which is really a combination of the time that credit has been in use and also the number of trade lines that are on the file". (page 44-45 line 19-23 and line 1) 

OUR COMMENT: Well again this may work on fantasy Island, but in the real world credit only stays on a consumers credit file 7 years from date of last activity. The biggest problem with this type of statement is that their are a lot of creditors that don't report good credit on consumers credit reports they only just report  the "BAD" so consumers are again getting harmed. Again, there are a lot of creditors that don't report because of the following reasons: (1) fear of losing a great account to their competition (2) they just don't have to or want to; nobody forces creditors to report (3) they only want to report negative data, these are just a few reasons why again you CAN'T use scoring, because it is "NOT" a true picture on how a consumer pays their bills or how a consumer may pay in the future.

MR. McCORKELL: (4) "The next category we look at is the pursuit of new credit. This is where the inquiries come in and also new account opening. Somebody that has a lot of   inquiries "GENERATED BY THE CONSUMER" -- and I'll get into that in a little bit -- or has opened a lot of new accounts recently, is more risky than somebody that hasn't been out there actively pursuing a lot of new credit. I mentioned inquiries. This has been a topic of some controversy. Fair Isaac about a year ago redesigned our bureau systems to do what we all  call inquiry de-duping. We know that in a lot of cases somebody shopping for a mortgage or for an auto loan will wind up with multiple inquiries from the same transaction. And of course not many people buy more than one house in any given month, or more than one car". (page 45-46 line 20-23, 1-4 and 10-17)

OUR COMMENT: This only takes into account mortgages and auto's what about other types of accounts? What about insurance? When you are shopping for the best rate on Homeowners, auto, and business insurance these companies are also using your credit score to quote your rate and are making inquiries into your credit file. 

MR. McCORKELL: (5) "The thickness of the file, in a sort of philosophical sense how many trade lines, etc., there are, the age of the file, how long has this person had established credit". (page 50 line 1-4)

OUR COMMENT: The thickness of your file! Trade lines only stay on your credit file for 7 years from date of last activity. Because of the economy that we are in now consumers are  refinancing all of their loans from mortgages, cars, credit cards and so on. Again, consumers are harmed because creditors don't report their good credit to the three National Credit Repositories.

WRONG INFORMATION STATED WITH REGARDS TO  C/R'S

MR. McCORKELL: "Secondly, now for those people where the answer is yes, all of their  prior credit they've performed in a satisfactory way in the sense that they've met all of the contractual obligations an they've never been late, etc., that accounts for about 85 percent of the population. And so there are only, again, rough numbers, 15 percent of the population  that's got serious delinquencies noted on their credit report at any given time". (page 43-44  line 19-23 and 1-4)    

OUR COMMENT: I believe that these numbers do not represent the true numbers of  delinquencies in consumer credit files. If these numbers where true this issue over credit scoring would not be here if 85 percent of consumers had no delinquencies and credit reports were 85% accurate.

MR. McCORKELL: "In fact, again, those of us who have been around for a while, sort of after  we got into the credit world, we probably have a lot of accounts that we've "PAID OFF AND CLOSED". That Good information, in Most Cases, will stay on the file "KIND OF INDEFINITELY". (page 45 line 3-7)

MR. McCORKELL: "And so somebody that's got 20 or 30 years of credit experience and ANY kind of reasonable number of trade lines, maybe a lot of which have been PAID OFF -- you know, student loans that have been paid off. Mortgages that have PAID OFF. You know, your first credit card may have been a gas card or a store card, because that was the only place you could get a credit card when you, you know, first got out of School. So those have been PAID OFF. All  of those historical pieces are "LIKELY" to still be on the file, and se we LOOK at that. It's really the amount of credit history available". (page 45 line 8-19)

MR. McCORKELL: "Again, somebody who is 50 years old and has 30 trade lines, most of which are "NOW CLOSED", that may be a very typical pattern. Somebody who is 22 years old and has 50 trade lines, and they're only been in the file for 24 months and they've got 50 trade lines already, that's probably kind of a strange bird". (page 50 line 5-10)

OUR COMMENT: I can not believe that this is the head legal counsel and senior vice president for Fair Isaac . This comment is why the credit scoring system must be heavily scrutinized. He has made the largest  statement of mistruth or the largest mistake of anyone in the credit business. This statement goes back to what he should have learned in "BASIC CREDIT 101". Just for the record accounts paid off and "CLOSED" will stay on a consumers credit file for "ONLY 7 YEARS" without exception. Note the inaccurate statement, will stay on the file "KIND OF INDEFINITELY" or will "LIKELY" to still be on the file. Again just for the record those trades lines "WILL NOT" be on the credit file. Can we believe that these statements are wrong? or that he made a mistake? or he was just giving examples? I believe that the above statements show that we have a company that needs to be "REGULATED" and brought before the Congress and FTC. This statement scares me to "DEATH" and to think these guys are the ones controlling our FINANCIAL FUTURE. 

CONTRADICTIONS IN TESTIMONY

MR. McCORKELL: "And so we're sort of able to say, if we count all of the auto loan  inquiries and all of the mortgage loan inquiries within any given 30 day period as a single inquiry, because it's probably only one auto or one house that they're shopping for, that is really a better representation of their credit behavior. And the also, again in mortgage lending especially, people may wind up shopping around for some period of time looking for the best rates, going to different lenders or go through a mortgage broker who shops at different lenders. And so in fact what we're said is , we don't count "ANY INQUIRIES" in the 30 days before this particular score is computed. So you can have a "HUNDRED  INQUIRIES" in the MONTH before the score is computed and the "WON'T" count at all". (page 46-47 line 18-23 and 1-9)

MR. McCORKELL: "And you can have any number of "MORTGAGE" or any number of "AUTO LOANS" in any -- I'm sorry. It's a 14 day period at any time before the score is computed, and they're only count as one inquiry each". (page 47 line 10-13)

OUR COMMENT: Well what is it? 30 days or 14 days? We have been told for about a year that  it  was 30 days and we have other public announcements that where made by FAIR ISAAC in which they stated  publicly it was 30 days. We have also been told by the Three National Credit Repositories that it was 30 days and we have told all our Mortgage Companies that information. We will have to now come out publicly and correct the information we where given a first. With these kinds of contradictions coming out of a company that is not checked or regulated how can we "TRUST THE SCORE" that comes out from this company.  Maybe today it's right and maybe tomorrow it's wrong. This is the most important purchase a consumer will make in their life time; "HOME OWNERSHIP", we can't leave it up to a "GUESS" or a "HOPE" that Fair Isaac is right.

OVERALL COMMENTS FROM TESTIMONY

MR. McCORKELL: "Well, we may have considered that implicitly because of all of these other factors that are in the scoring system, and so for a number of reasons, they're characteristics that people try to keep out of scoring systems. One of them -- something that, you know, sort of might be a very obvious characteristic is "INCOME". It sounds like a nice numerical characteristic. It's probably got something to do with how people are going to be able to pay their bills. And it actually turns out to be a very messy characteristic. And so somebody says, well, you don't ask for my income. How could you possibly make an  intelligent credit decision". (page 32 line 1-9 and 17-21)

OUR COMMENT: This author believes that without income you will discriminate against the self-employed and other people making good incomes. That is why the credit scoring system must be done away with in some instances. Income should not be used on one hand, but if you look at a person's debt, that they pay on time, and lower the score based on to much debt because of some study of "NORMAL PEOPLE", then you must give the consumer the right to add income to off set the debt. Remember that it's the score that "MOST" underwriters are using. They are no longer looking at the credit report as the predictive power. As a final note on this topic of income, Fannie Mae and Freddie Mac systems use this scoring system. The consumer has "NO RIGHT" or ability to fight what they "CAN'T" see and or understand. 

MR. McCORKELL: "Now, these are scores that Fair, Isaac and others -- there are other scoring developers out there. I don't know why anybody would use them, but some people do". (page 42-43 line 21-23 and 1)

OUR COMMENT: Well, I think this comment speaks for its self. I believe and this is only my opinion, that this company is too "BIG" and there is nothing worse than a company that "THINKS" they have no competition and have no need for improvement.    

MR. McCORKELL: "Because of the multiple scorecard design, it's not possible to tell somebody,  if you do this, your score will go up by X, because in fact the credit bureau information is a "MOVING TARGET". There's always information coming into the credit bureau. Essentially on a daily basis credit bureaus are getting information from credit  grantors. And so I could sit there with somebody's credit report and say well, if you close this account, or if you paid off that account, here's what would happen to your score if nothing else changed. But something else will always change. And then one of the things that may change, is that by doing something you may wind up on a different scorecard, because the  way that the system is split". (page 47-49 line 22,23 and 1-5 and 21-23 and 1-5)

OUR COMMENT: Wow! you can create a system that you can never be held accountable for because the credit bureau data is a "MOVING TARGET". Also, it's not my data so you can't hold me accountable. Because of the way the system is designed I don't have to explain anything to anybody about how it works. Consumers and loan officers could end up hurting themselves or their  client by trying to instruct them to "do this or that" in hopes of raising their score. I've seen, on a daily basis, how scores can change by 10's or 100's during the real estate process so it really depends on "WHEN" you pull the credit report. You could have loan approval "TODAY" and lose it "TOMORROW" it all depends on how you hit that "MOVING TARGET" on a good day or a bad day.