• Posts Tagged ‘credit_score’

    Credit Card Conundrum

    by  • June 17, 2010 • Tagged: ,  • Comments

    ccards.jpg

    photo: Andres Rueda

    Over the past few months I’ve chosen to close a few credit card accounts. I think that in each case, the card carriers were going to start charging an annual fee unless we started using the card. The most recent example was my Citibank card, as they wanted us to spend at least $2400 per year to have them waive a $60 fee.

    The problem with all of the recent cancellations are that 1) the cards were among the oldest that I had and 2) they had pretty high credit limits. Cancelling those cards has reduced the average age of my credit history and has decreased my overall credit limit. I do still have a relatively high overall credit limit around $35,000 and my oldest card was opened in 2000, with others opened from 2005 to 2007. But still, it sucks that I had to cancel those cards.

    Of the remaining credit cards that I have, I use two of them regularly. Both are rewards cards, but one of them has a $45 yearly fee. Oh, and Her also has that card as well. So yes, we’re paying the $45 annual fee twice.

    So naturally, I’m contemplating closing that card as well so that we don’t have to pay the annual fee more times than we need to. The problem is that card is also the second oldest card that I have and has the second highest credit limit; closing the card would effectively halve my overall credit limit. I don’t want to take another ding on my credit score. I’ve had this card for almost 5 years, so we’ve paid $180 in fees so far – soon to be $225. We have definitely received much more in rewards (5-star hotels in Europe are a very nice perk), and so far have accumulated enough points to go on more awesome vacations.

    Do you think it’s worth it for me to still have this card? What alternatives could you think of instead?

     

    FICO as a Mathematical and Sociological Challenge

    by  • March 17, 2006 • Tagged: ,  • Comments

    Mathematically speaking, a FICO score is nothing more than the result of an equation that considers multiple variables. The exact equation is a closely guarded industry secret. After all, the credit bureau couldn’t sell you your FICO score if you could calculate it yourself. But I’ve been pondering this equation for about a week now and I just can’t come up with a good reason why nobody has cracked the equation.

    First, consider it as a mathematical challenge. I assume the equation could probably be determined by using linear regression. This is a math trick that allows you to isolate a single variable among many and solve for its value. It is used most commonly to determine real estate values. For example, imagine that you and your neighbor have identical houses on identical properties. The only difference is that your house has a fire place and your neighbor’s house does not. Now imagine you both sell your respective houses on the same day. He sells his for $100,000 and you sell yours for $105,000. It’s pretty easy to determine that, all else being equal, your fireplace was worth $5,000. It gets more complicated in the real world; because of course no two houses are completely identical. Even in a development with rows of identical houses, one might have a nicer view of the lake while the other may be closer to a busy street. So real estate agents use linear regression to isolate the base home price and the value of each add-on (like a fireplace or nice view) or detractor (such as noise). Then they can predict a fair price for your home.

    It makes sense to me that the FICO formula would work the same way. You would get a base number to start with, and then each good thing (on-time payments, low debt-to-credit ratio, etc) would add points and each bad thing (missed payments, late payments, etc) would detract from your score. This is the perfect situation for linear regression.

    While linear regression is time consuming, it isn’t hard. And it isn’t as though we don’t have some basic FICO equation information already. Here’s what we do know, because the FICO company has made this information public:

    1. The solution (your FICO score) can range from 250 to 900.
    2. There are not more than 33 variables and we know how much each one deducts from your score.
    3. If your credit report is accurate, you have all the same raw data they use to calculate your FICO score.

    So why hasn’t anyone cracked the formula? With all the brilliant mathematicians on the planet today, can’t one of them whip out a FICO equation? Imagine the fame and fortune that would follow such a revelation! That person could sell “FICO” scores at a discount and undercut the big guys, or he could work to improve the formula and sell a better score.

    The final aspect of this mystery is human nature. The FICO score equation is a secret only so long as nobody spills the beans. With all the employees who have worked at the Fair Isaac Corporation, surely one or two of them have seen the equation. Are their employees really that loyal and trustworthy?

    How does such a simple secret stay a secret for so long?