Why is it that bankruptcy has less, or equal, an impact as "paid" collections?

[MattH]

---

@Anonymous wrote:
...

FICO should reprogram their algorithm to punish defaulters in proportion to the amount owed. A charge off should not just be a charged off. A $10,000 Charged off is not the same as a $150 charged off.

I am a programmer and If I were given the choice to modify FICO algorithm, I will classify defaulters as

...

---

You utterly miss the whole point of FICO scoring: such algorithms are not about moral judgements and punishments.  Their main customers -- lenders -- are not making moral judgements, they are making business decisions guided by empirical probabilities (ie., what is the risk if I lend money to this person?).  Certain factors such as race and gender may not be used because there are laws against using them.  Also after certain legally-mandated time intervals old negative information must be expunged from the records, again as required by law.  Within these legal constraints, Fair Isaac and their competitors use statistical algorithms to optimize their scoring models for predictive value.

[Anonymous]

---

@MattH wrote:

---

@Anonymous wrote:
...

FICO should reprogram their algorithm to punish defaulters in proportion to the amount owed. A charge off should not just be a charged off. A $10,000 Charged off is not the same as a $150 charged off.

I am a programmer and If I were given the choice to modify FICO algorithm, I will classify defaulters as

...

---

You utterly miss the whole point of FICO scoring: such algorithms are not about moral judgments and punishments.  Their main customers -- lenders -- are not making moral judgments, they are making business decisions guided by empirical probabilities (ie., what is the risk if I lend money to this person?).  Certain factors such as race and gender may not be used because there are laws against using them.  Also after certain legally-mandated time intervals old negative information must be expunged from the records, again as required by law.  Within these legal constraints, Fair Isaac and their competitors use statistical algorithms to optimize their scoring models for predictive value.

---

Hey Matt,

  Thanks for the clarification. I now get the point. My choice of word (punish) was rather poor and I already edited my post to reflect a more "civilized" usage (penalty).

I don't have anything against anyone who declare BK as we are all susceptible to such unfortunate condition leading to bankruptcy. While it is a sad thing to default on loans due to unforeseen circumstance like loss of job, loss of health, divorce or any problem whatsoever, we can not overlook the fact that some people are out there to game the system.

The premise of my argument is not to punish any particular group or person, but rather to ensure that penalty are justly awarded with respect to the type of default. However, like Matt pointed out, the business or financial world have no moral standard, but rather, profit is their main goal. If the current FICO system work for them, then let it be.

 I apologize if my opinion has offended anyone even though I have right to my own opinion as permitted by the 

1st amendment . All opinion expressed in my previous post are completely mine and does not represent the opinion of any other person, group or organization.

                                   Good luck to you all.

[Anonymous]

I concur with Matt - FICO is a tool for prediction and risk assesment to a creditor.  Its number, like beauty, is in the eye of the beholder.

In addition, there is not a linear relationship between a credit item and the score.  It is a multidimensional or relational data algorithm which attempts to find data points in common between files in order to give a semblance of intelligence to the otherwise random and vast data within the collective credit data base.

In other words, to find out what credit data points are in common with positive and or negative consequences and outcome.

FICO scores are not like a school grade, though they seem like one.  Because not all lenders treat them equally in evaluation.

FICO is not fair, except that it is applied universally and equally.  It is not judgemental nor designed to punish nor reward.  It is a numerical indicator of risk which moral or emotional judgments cannot play a part and still remain relevant to its intended purpose and use.

The problem was never FICO for current financial crisis.  It was that lenders used lower and lower standards - the everyone qualifies standard.  FICO did not create that, they continued to deliver the same consistent comparative analysis they always have.  Lenders chose to take on more risk because of the high profits.

FICO can be tuned and made more accurate in prediction, but if lenders do not utilize conservative or safe lending practices, all the FICO scores in the world are moot.

[MattH]

---

@Anonymous wrote:

I concur with Matt - FICO is a tool for prediction and risk assesment to a creditor.  Its number, like beauty, is in the eye of the beholder.

In addition, there is not a linear relationship between a credit item and the score.  It is a multidimensional or relational data algorithm which attempts to find data points in common between files in order to give a semblance of intelligence to the otherwise random and vast data within the collective credit data base.

In other words, to find out what credit data points are in common with positive and or negative consequences and outcome.

FICO scores are not like a school grade, though they seem like one.  Because not all lenders treat them equally in evaluation.

FICO is not fair, except that it is applied universally and equally.  It is not judgemental nor designed to punish nor reward.  It is a numerical indicator of risk which moral or emotional judgments cannot play a part and still remain relevant to its intended purpose and use.

The problem was never FICO for current financial crisis.  It was that lenders used lower and lower standards - the everyone qualifies standard.  FICO did not create that, they continued to deliver the same consistent comparative analysis they always have.  Lenders chose to take on more risk because of the high profits.

FICO can be tuned and made more accurate in prediction, but if lenders do not utilize conservative or safe lending practices, all the FICO scores in the world are moot.

---

Agreed.  And another all-too-common mistake made by mortgage lenders was to look only at the FICO score.  For a mortgage, good underwriting considers three main factors:

1. The amount borrowed as a proportion of a realistic market value for the property.  Lower is better for two main reasons (less risk of being underwater should the market drop, and also as a demonstration that the borrower has been able to save in the past).
2. The ability of the borrower to make payments, based on income and other financial obligations.  Amount of income is of course the main consideration, but so is its stability.
3. The borrower's past credit history.

FICO scores by definition consider only the third point above.  For the short term they are an excellent predictor, but a mortgage is longer term so its sustainability needs to be analyzed.  During the bubble everybody was thinking only of the short term.