由买买提看人间百态

boards

本页内容为未名空间相应帖子的节选和存档,一周内的贴子最多显示50字,超过一周显示500字 访问原贴
Money版 - FICO信用分数进阶介绍 (转载)
相关主题
FICO信用分数进阶介绍 (转载)myfico vs creditkarma
哪里能查三大信用分数?请问到底有几个信用分数?
不同版本的FICO可以差近100分关于credit karma的分数
TransUnion的信用分数Discover的FICO挺准的啊?
震惊:我在于17%比例的人口里。。。信用分数急问
要FICO,不要FAKO,有兴趣的看看。。。为什么我在 Credit Karma 的分数和别家差别那么大?
discover 申卡给的 credit score 准么?请教怎么查自己的信用分数?
discover fico 正式上线了FICO和真的score差挺远啊
相关话题的讨论汇总
话题: fico话题: 分数话题: 信用话题: __话题: score
进入Money版参与讨论
1 (共1页)
t******g
发帖数: 937
1
【 以下文字转载自 Money2 俱乐部 】
发信人: tayshung (tayshung), 信区: Money2
标 题: FICO信用分数进阶介绍
发信站: BBS 未名空间站 (Tue May 27 16:40:59 2014, 美东)
在myFICO论坛上看到的一篇干货好贴,大致转贴贴一下。
1. 区分FICO 分数和FAKO 分数
2. FICO score的历史版本和应用
3. FICO 8模型中,各个指标的调整
4. 到底需要多高的分数才能批信用卡/车贷/房贷?(我自己总结的)
如果是信用分数方面的小白,请首先阅读Money版上的入门介绍贴:
http://www.mitbbs.com/article_t/Money/31596079.html
http://www.mitbbs.com/article/Money/31570407_0.html
另外各位有什么问题,欢迎提问,我会尽可能回答。这篇帖子若有不准确之处,也欢迎
指正!
1. 区分FICO 分数和FAKO 分数
FICO分数是90%lender借款时候参考的分数,由FICO公司做的信用分数模型。FAKO 分数
是其他非FICO的公司基于个人信用信息做的信用分数模型,比如“著名”的credit
karma。各大金融机构借款时从来不会看FAKO 分数,所以FAKO 分数除了给你个信用分
数的大致估算以外,毫无用处。
2. FICO score的历史版本和应用
FICO公司推出的FICO 分数模型并不是一尘不变的,历史上有很多大修订和小改进,好
比操作系统升级一样。每个版本的模型得出的信用分数会有差异。所以不同的商业公司
在借款给你的时候,所使用的FICO 分数也不一致,好比windows已经升级到了win8,但
是主流还是用的win7,而更大更老的公司还在用win XP。
历史上,FICO公司分别在95, 98, 04, 08年发布过四次大的模型修订。业界一般习惯称
之为FICO 98, FICO 04,和FICO 8(不叫FICO 08,就叫FICO 8)。目前消费者自己买
到的FICO分数一般都是FICO 8,而各大金融机构用的往往是FICO 04模型。
另外,由于每个信用局的提供信息存在差异,每个信用局又会有更细致的分数模型,比
如用TransUnion, FICO 98模型出来的信用分数,就叫做TU 98.
下面我会转帖一下各个模型的出处和应用情况(原帖还提供了每个模型分数的上限下限
,感兴趣的可以深挖)。
----------------------------------------------------------------------------
FICO 98 (introduced about 1998)
----------------------------------------------------------------------------
Transunion
__Official name: FICO Risk Score, Classic 98 (formerly called Empirica)
__Common name: TU-98
__Available directly to consumers: from MyFICO when buying Transunion FICO
score
Equifax
__Official name: Beacon 96
__This version appears to be seldom used, but a poster reported it used on a
mortgage application in 2014.
Experian
__FICO 98 edition no longer in common use

----------------------------------------------------------------------------
FICO 04 (introduced about 2004)
----------------------------------------------------------------------------
The FICO 04 edition is the standard used by almost all mortgage lenders.
All three scores will normally be pulled and the middle score (not the
average) will be used by the lender.
Transunion
__Official name: FICO Risk Score, Classic 04 (formerly called Empirica)
__Common name: TU-04
__Available directly to consumers: no
Equifax
__Official name: Beacon 5.0
__Common name: EQ-04
__Available directly to consumers: Some credit unions such as DCU provide
it free each month to members.

Experian
__Official name: Experian/FICO Risk Model v2 (also sometimes called FICO
II)
__Common name: EX-04
__Available directly to consumers: Some credit unions such as PSECU
provide it free each month to members.
----------------------------------------------------------------------------
FICO 8 (introduced about 2008)
----------------------------------------------------------------------------
Transunion
__Official name: FICO 8 Risk Score (also sometimes called FICO Risk Score,
Classic 08)
__Common name: TU-08
__Available directly to consumers: Some credit cards issuers such as
Discover, Barclays, and Walmart provides it free each month.

Equifax
__Official name: Beacon 09 (name apparently derives from year of adoption
= 2009)
__Common name: EQ-08
__Available directly to consumers: from MyFICO when buying Equifax FICO
score
Experian
__Official name: Experian/FICO Risk Model v8
__Common name: EX-08
__Available directly to consumers: from MyFICO when buying Experian FICO
score
3. FICO 8模型中,各个指标的调整
借款额占信用额度比例 (Utilization): 高比例的欠款会进一步降低分数
High utilization of credit lines (both overall and for individual cards)
will have a more negative effect on score. FICO has not said anything more
than this generic statement, and I have not seen anyone attempt to quantify
the effect in terms of score points.
是否及时还款 (Late payments): 一次迟付账单的影响会减小,但是多次累计迟付会进
一步降低分数
A single, isolated late payment will have less negative effect, but a
pattern of numerous late payments will have more negative effect.
是否有追讨债务记录 (Collections): $100以下的的债务追讨将不会降低分数
For debts with an original balance less that $100, public records and debts
that have been assigned to a 3rd party collection agency will no longer
have a negative effect on score.
信用种类记录,如信用卡/车贷/房贷/学生贷款 (Types of credit used): 更多种类的
信用记录,会提高分数
About 10% of your FICO score is based on what types of credit accounts you
have, including credit cards, retail accounts, installment loans, mortgage
loans, and finance company accounts. Having a variety of different credit
types improves your score, except for finance company accounts which have a
negative effect. With FICO 8, the "types of credit used" category has
greater weight than before. So someone like me who has only credit cards on
my report will have a lower score under FICO 8 than under previous versions.
副卡的信用历史计算 (Authorized users): 只有同一个家庭里的人申请副卡,信用卡
的记录才会算到副卡人的信用分数里
FICO 8 changes the way it treats authorized users who are not family members
of the card holder. This is intended to reduce the benefit of "piggybacking
" whereby people pay credit repair companies a fee to become an authorized
user of someone with a good credit score.
信用卡数量 (Number of accounts): 多开卡分数比较高
Ethan Dornhelm of FICO said that under FICO 8, having "a relatively large
number of open accounts in good standing will be viewed more favorably than
having only a few accounts in good standing." Having only a small number of
accounts that are both open and active will now have a more negative effect
on your score.
开通但是不用的卡 (Accounts that are open but inactive): 平常不用的卡,算进信
用分数的比例会很少
FICO 8 is more aware of which accounts have been inactive with a 0 balance
for a long time, and those accounts will factor less into your score.
信用查询 (Inquiries)
When FICO 8 was originally introduced, hard inquiries had less negative
effect on your score. But a later revision changed this so that inquiries
continue to have about the same effect as before.
4. 到底需要多高的分数才能批信用卡/车贷/房贷?(以下为我自己总结)
先说明下,信用分数高并不一定意味着你就一定能申到信用卡/批车贷/批房贷。但是如
果你的信用分数不高,这些一定批不下来,或者是批下来的利率很高,你要付很多利息。
这是因为信用分数是给金融机构进行参考的一个分数。一般来讲,大多数信用卡公司只
要你到达了一定分数就会批准你的申请(个人经验,FICO 700分以上问题就不大)。但
是也有些信用卡公司比较严格,它们的系统里会设置其它条件,比如它通过pull你的信
用报告,发现你最近的hard pull很多,同时又新开了很多账户,它们的系统会认定你
的风险太高,因此系统会自动拒批。另外有些信用卡公司内部会设置一些条件,比如60
天内最多批准不超过它家的3张卡。下面就拿最常见的几家银行来讲:
Citi: 一个星期内最多批1张,60天内最多批2张。所以有人总结出了8天-65天的法则。
Chase & Amex: 一般来讲,personal cards(不包括charge card)不超过4张。第5张
以后申请的新卡被系统拒绝的概率比较高。
BOA: 没啥特别的限制。
Barclay: 非常谨慎的银行。会以最近hard pull太多拒绝批卡。
US Bank: 同样非常谨慎的银行。不过据说它家经常pull三大信用局以外的信用报告,
所以把那两家信用局的信用报告freeze能大大提高申请通过概率。
Wells Fargo: 申它家的信用卡之前,最好在它家先有checking account或者saving
account,它们会要求有所谓previous relationships。
然后讲车贷。以我个人的经验来讲,信用分数达到720以上,基本就能拿到最好的利率
了。不同车厂经常有各种promotions,什么0.9%,0.0%利率在最近的利率市场环境下很
常见。
再讲房贷。房贷我目前还没申请过。不过据我收集的信息是这样的:740以上的分数没
有差别,能拿到最好的利率。一般你去找金融机构申请房贷的时候,它们运行的往往是
FICO 04的模型,和你自己买的FICO 8模型不完全一致,所以分数会有差别。金融机构
会同时pull三家的信用分数,然后取中间值(median, not average)。另外房贷agent
也会看你和配偶的信用报告上的其他信息,比如最近有没有hard pull,有没有新开的
账户。有些房贷agent会对这些"seek credit behaviors"比较敏感,因而不给你最好的
利率。所以一般建议在做房贷前的6-18个月,最好不要开新卡
关于FICO 04和FICO 8模型的分数差,我记得版主saintknight以前好像发帖讲过,有时
候可能相差到20分。所以最最保守的算法,如果你能查到的FICO分数在760以上,就不
用担心信用分数不够而拿不到最好的利率了。如有纰漏,欢迎版主斧正。
-------------------------------------------
myFICO论坛原帖地址:
http://ficoforums.myfico.com/t5/Understanding-FICO-Scoring/The-
g********3
发帖数: 396
2
关于FICO 04和FICO 8模型的分数差,我记得版主saintknight以前好像发帖讲过,有时
候可能相差到20分
-----------
这个也许不止,我discover的FICO分数是769,lender pull到的TU分数是720多,相差
40多分
C******n
发帖数: 9204
3
各金融公司会有自己的model,FICO只是一个input。Amex听说是ANN的算法。。。所以
没法说FICO多高就必然能如何。。。
1 (共1页)
进入Money版参与讨论
相关主题
FICO和真的score差挺远啊震惊:我在于17%比例的人口里。。。
三大信用公司的 FICO 分数都能免费看了要FICO,不要FAKO,有兴趣的看看。。。
为什么Citi和Discover提供的FICO分数不一样?discover 申卡给的 credit score 准么?
包子求问怎么看3局详细信用分数+记录?付费的也行,有deal更好discover fico 正式上线了
FICO信用分数进阶介绍 (转载)myfico vs creditkarma
哪里能查三大信用分数?请问到底有几个信用分数?
不同版本的FICO可以差近100分关于credit karma的分数
TransUnion的信用分数Discover的FICO挺准的啊?
相关话题的讨论汇总
话题: fico话题: 分数话题: 信用话题: __话题: score