Free Transunion Credit Based Insurance Scores thru CK

[Thomas_Thumb]

---

@beutiful5678 wrote:

Very helpful step-by-step instructions!

The dip in May was due to a 30-day late reporting (went Apr 852 --> May 789 --> Jun 852) and the score seems to have bounced back quickly. I think the two things helping my score (per the screenshot and link below) are the thickness and length of my history. I have 40 accounts, AAoA is 7.8, and AoOA is 12.1 years. (Unfortunately, along with that thick history I also have 70+ lates payments reporting - having multiple Stafford loans can be a blessing and a curse.)

My home insurance score "needs work" - it's probably a good thing that I don't own a home, LOL.

https://www.creditkarma.com/auto/i/insurance-scores-affect-car-insurance-rates/

---

I appreciate the info so far. More would be better. Hopefully SJ, BBS, CGID and others are willing to share.

From what I can tell so far TU CBIS:

1) Does not consider # accts reporting balances (my score stayed the same month to month regardless of # reporting.

2) Does consider time since most recent credit ap rather strongly (this includes HPs for CLIs)

3) Does consider recent BC openings with none in the last 24 months being optimal

4) Does consider AAoA  (optimal value > 4.5 years) but not AoOA in general.

5) Uses a 5 year timeframe for charge offs, collections.

6) Looks at 60+ day lates only with focus on # 60+ day lates in the last 12 months and AoOA with a 60 + day late. Settled 30 day lates do not have a lingering impact.

7) Considers average CL but does not consider aggregate UT%.

8) Considers # of bank installment loans with < 6 being preferred

9) Does not appear to look at credit mix and file thickness (other than too many accts)

Provided below is a summary table of data from above posts. I'll add to it as more data becomes available. New posters can include as little or as much information as they are comfortable with. [last update 6/5/2018]

Poster ID	* T_T *	* T_T *	*T_T*	B5678	Gmood1	HOhio	SubE	-BBS-	Sarge
TU Auto CBIS	902	869	892	852	774	860	820	903	912
TU Home CBIS	950	917	931	698	847	897	850	912	823
File Description	clean	clean	Clean	derog	clean	clean	clean	clean	clean
Average Age of Accounts (mo)	186	205	212	92	35	76	14	91
Average card CL (optimal > $9k)	20,100	23,600	24,100	360	17,900	10,200	6100	23,900
AG UT%	2	3	3	62	2	3	1	1
# installment loans (optimal <6)	1	1	1	1	3	0	1	1
Most recent BC opened (mo)	46	79	86	4	7	15	7	13
# card aps < 12 mo	0	0	0	2	6	0	3	0
# card aps < 24 mo	0	0	0	?	14	5	5	6
Most recent credit AP/HP (mo)	46	6	13	3	2	7	7	13
# collections < 5 yr age	0	0	0	3	0	0	0	0
# Charge offs < 5 yr age	0	0	0	6	0	0	0	0
# +60 day lates <12 mo age (opt < 5)	0	0	0	0	0	0	0	0
most recent delinquency retail acct	n/a	n/a	n/a	1	n/a	n/a	n/a	n/a
AoOA 60+ days late	n/a	n/a	n/a	23	n/a	n/a	n/a	n/a

Based on the TU reason statements 30 day lates are not considered and focus for 60+ day lates is the last 12 months. Clearly there are differences between Auto and Home - otherwise no need for two models. Question is how to differentiate. B5678's much higher Auto score vs Home coupled with GMOOD1s higher Home score vs Auto might help shed some light on this but, more data is needed

[HeavenOhio]

TT, what's BC? And does AP/HP stand for application/hard pull?

[Thomas_Thumb]

BC = Bank Card (likely includes credit cards in general as TU CBIS does not appear to differentiate store cards)

AP/HP = credit application with a hard pull. This includes CLI and other types of requests that receive a hard inquiry (pull).

[HeavenOhio]

Thanks for the clarification.

I've seen it posted numerous times in the scoring forum that retail cards ding insurance scores. Care to comment?

Here's my TU info:

• TU Auto CBIS: 860
• TU Home CBIS: 897
• File Description: clean
• Average Age of Accounts (mo): 76 (Pulled from myFICO rather than Karma due to their omission of closed accounts. Karma has me at 96 months including only my open accounts.)
• Average card CL: $10,150 ($9,400 if closed accounts are included)
• AG UT%: 3%
• # installment loans: 0 (open or closed)
• Most recent BC opened: 15 months
• # card aps < 12 mo: 0
• # card aps < 24 mo: 5
• Most recent credit AP/HP (mo): 7 months (CLI)
• # collections < 5 yr age: 0
• # Charge offs < 5 yr age: 0
• # +60 day lates <12 mo age (opt < 5): 0
• most recent delinquency retail acct: n/a
• AoOA 60+ days late: n/a

I see my auto score bumped up by 20 points in May. It's possible that it's due to accounts aging. Or maybe it's because a November inquiry reached six months in age.

My home score was at 912 last November and dropped to 894 after an inquiry. It's worked its way up by three points since then.

[Thomas_Thumb]

---

@HeavenOhio wrote:

Thanks for the clarification.

I've seen it posted numerous times in the scoring forum that retail cards ding insurance scores. Care to comment?

I see my auto score bumped up by 20 points in May. It's possible that it's due to accounts aging. Or maybe it's because a November inquiry reached six months in age.

My home score was at 912 last November and dropped to 894 after an inquiry. It's worked its way up by three points since then.

---

Thanks for the data.

Neither TU nor Fico CBIS models appear to differentiate between retail store cards and "Bank" cards other than TU mentioning it in a delinquency age metric. However, LexisNexis, the market leader in CBIS, does differentiate retail/store cards from Bank cards as evidenced in their reason code list. LN also calls out gas cards - FWIW.

It is important not to lump all CBIS models together. These models differ one from another. It's not all that different than what we encounter trying to compare VantageScore to Fico.

[Gmood1]

Average CL on 29 BCs reporting is $17,9xx. Total revolving reporting credit limits minus one line of credit is $519,700.

Overall reporting utilization is less than 1%. Closer to 2% with hidden TLs.
No derogs or lates, file completely clean.
Two youngest bank cards reporting, were obtained 7 months ago. Two others 8 months ago. One at 10 months and another at 12 months.
Last HP April 2018.
Succesful card apps last 24 months reporting to TU is 14. Real number is closer to 20 total, with some hidden or not reporting to TU.
I have a total of 19 closed accounts still reporting.
3 active installment loans. 9 installment loans reporting.

Let me know if you need anything else.

[HeavenOhio]

---

@Thomas_Thumb wrote:
Neither TU nor Fico CBIS models appear to differentiate between retail store cards and "Bank" cards other than TU mentioning it in a delinquency age metric. However, LexisNexis, the market leader in CBIS, does differentiate retail/store cards from Bank cards as evidenced in their reason code list. LN also calls out gas cards - FWIW.

It is important not to lump all CBIS models together. These models differ one from another. It's not all that different than what we encounter trying to compare VantageScore to Fico.

---

Thanks for the clarification. In the future, we can specifically mention LexisNexis when we say that retail cards cause a ding.

I've seen it posted here that in addition to gas cards, LN doesn't like retail cards for auto parts stores either.

[Subexistence]

---

@Thomas_Thumb wrote:

I would pay to check my LexisNexis CBIS as well - but those scores are no longer available The takeaway for me is CBIS models appear to be very sensitive to credit seeking relating to HPs alone. I shudder to think how the models would treat a bunch of new accounts (particularly store cards).

---

Why are those scores no longer available?

[beutiful5678]

---

@Thomas_Thumb wrote: 

Poster ID	* T_T *	* T_T *	B5678	Gmood1	HOhio	******	******
TU Auto CBIS	902	869	852	774	860
TU Home CBIS	950	917	698	847	897
File Description	clean	clean	derog	clean	clean
Average Age of Accounts (mo)	186	205	92	35	76
Average card CL (optimal > $9k)	20,100	23,600	363	17,900	10,200
AG UT%	2	3	62	2	3
# installment loans (optimal <6)	1	1		3	0
Most recent BC opened (mo)	46	79	4	7	15
# card aps < 12 mo	0	0	2	6	0
# card aps < 24 mo	0	0	2	14	5
Most recent credit AP/HP (mo)	46	6	3	2	7
# collections < 5 yr age	0	0	3	0	0
# Charge offs < 5 yr age	0	0	6	0	0
# +60 day lates <12 mo age (opt < 5)	0	0	0	0	0
most recent delinquency retail acct	n/a	n/a	1 mo	n/a	n/a
AoOA 60+ days late	n/a	n/a	23 mo	n/a	n/a

Based on the TU reason statements 30 day lates are not considered and focus for 60+ day lates is the last 12 months. Clearly there are differences between Auto and Home - otherwise no need for two models. Question is how to differentiate. B5678's much higher Auto score vs Home coupled with GMOOD1s higher Home score vs Auto might help shed some light on this but, more data is needed

---

Is this just open installment loans or total installment loans? I have 1 open installment loans and 30 total installment loans.

[Thomas_Thumb]

---

@Subexistence wrote:

---

@Thomas_Thumb wrote:

I would pay to check my LexisNexis CBIS as well - but those scores are no longer available The takeaway for me is CBIS models appear to be very sensitive to credit seeking relating to HPs alone. I shudder to think how the models would treat a bunch of new accounts (particularly store cards).

---

Why are those scores no longer available?

---

I suspect LN no longer offers CBIS to consumers because the major insurance companies "suggested" they discontinue the practice. Given the limited revenue from consumers purchasing the score, why antagonize your primary customers? Also, unlike denials for credit, disclosure of scores/reasons for high insurance premiums is not required by law.

I wrote LexisNexis a note about this. Below is their response: