Детальний розбір
Покроковий розбір
Крок 1. VantageScore Consortium Origins and Model Architecture
VantageScore was created in 2006 as a joint venture among Equifax, Experian, and TransUnion. Unlike FICO, which is developed by an independent analytics company, VantageScore is owned by the bureaus through VantageScore Solutions LLC. This ownership structure gives the bureaus a proprietary model competing with FICO's market dominance.
VantageScore 1.0 used a 501-990 range, deliberately different from FICO's 300-850. VantageScore 3.0 (2013) adopted the 300-850 range for compatibility. VantageScore 4.0 (2017) introduced machine learning and trended credit data, making it the most technically advanced version. The machine learning architecture captures non-linear interactions between predictor variables without explicit feature engineering.
VantageScore can score consumers with as little as one month of credit history and one active account, compared to FICO's requirement of six months and recent activity. This lower threshold allows approximately 33 million additional consumers to be scored who are unscorable under FICO's criteria, including recent immigrants, young adults, and credit-inactive consumers.
- VantageScore Solutions LLC is jointly owned by Equifax, Experian, and TransUnion
- VantageScore 4.0 uses machine learning instead of traditional logistic regression
- VantageScore can score consumers with as little as one month of credit history
- Approximately 33 million more consumers can be scored by VantageScore than by FICO
- VantageScore 1.0 used 501-990 range; versions 3.0+ use the standard 300-850 range
Крок 2. Factor Weight Hierarchy: Six Categories vs. FICO's Five
VantageScore 4.0 publishes six factor categories: payment history (41%), age and type of credit (20%), utilization (20%), total balances (6%), recent credit behavior (11%), and available credit (2%). These differ from FICO's five factors in both number and weight distribution.
The most notable divergence is payment history at 41% vs. FICO's estimated 35%. Delinquency events have a proportionally larger impact on VantageScore. When VantageScore's three credit-usage categories are aggregated (utilization + total balances + available credit = 28%), the total approaches FICO's 30% utilization weight, but the granular breakdown reveals more nuanced evaluation.
VantageScore also incorporates rental, utility, and telecom payment data when reported to bureaus. These alternative data sources supplement thin files and provide signals for consumers with limited conventional credit history. FICO has been slower to adopt alternative data types, though UltraFICO and FICO XD represent experimental efforts.
- VantageScore weights payment history at approximately 41% vs. FICO's estimated 35%
- VantageScore breaks credit usage into three separate categories totaling approximately 28%
- Rental, utility, and telecom data can be incorporated into VantageScore when reported to bureaus
- The six VantageScore categories do not map one-to-one to FICO's five factors
- VantageScore publishes influence levels alongside percentage ranges for each category
Крок 3. Trended Data and 24-Month Behavioral Profiling
VantageScore 4.0 was the first widely adopted scoring model to incorporate trended credit data by default. It captures 24 months of historical balance, payment, and credit limit data per tradeline, analyzing trajectory rather than just the current snapshot. This enables distinction between consumers whose balances are declining versus those steadily increasing.
Under trended data, two consumers with identical 25% current utilization score differently if one has been paying down from 50% while the other has been climbing from 5%. The declining-balance consumer demonstrates lower risk trajectory. Traditional snapshot models cannot make this distinction because they only see the current month's balance.
The model also penalizes minimum-payment revolvers more aggressively. Consumers who consistently pay only the minimum while carrying near-limit balances exhibit a behavioral pattern correlating with higher default risk. VantageScore 4.0 captures this through the 24-month payment history, applying a penalty that snapshot models cannot assess.
- VantageScore 4.0 analyzes 24 months of balance, payment, and credit limit data per tradeline
- Declining balance trajectories score higher than static balances at the same current utilization
- Minimum-payment revolvers receive a behavioral penalty under trended data analysis
- FICO 10T also uses trended data but was released three years after VantageScore 4.0
- Trended data models reward consumers demonstrating consistent debt reduction
Крок 4. Inquiry Handling: Universal 14-Day Deduplication
VantageScore uses a 14-day rolling deduplication window that applies universally across all inquiry types. All hard inquiries within 14 days, regardless of product type (mortgage, auto, credit card, personal loan), count as a single inquiry. This is fundamentally different from FICO's product-specific approach.
FICO 8+ uses a 45-day window limited to mortgage, auto, and student loan inquiries. Credit card inquiries are never deduplicated under FICO. Applying for three credit cards in one day generates three separate inquiry hits under FICO but may count as one under VantageScore. This difference is a common source of score variance between the two models.
The impact of inquiries on VantageScore is somewhat lower than on FICO 8. VantageScore's machine learning model determined that inquiry count is less predictive of future default than other factors, assigning proportionally less weight. The model can also analyze inquiry patterns, distinguishing rate-shopping from diffuse credit-seeking behavior.
- VantageScore deduplicates all inquiry types within a 14-day window, including credit cards
- FICO 8+ uses a 45-day window for mortgage, auto, and student loans only; never deduplicates credit card inquiries
- VantageScore assigns less overall weight to inquiries than FICO based on machine learning analysis
- Both models ignore soft pulls from existing creditors for account management
- The universal deduplication is a common reason VantageScore exceeds FICO for consumers with recent card applications
Крок 5. Collection Account Treatment and Medical Debt Handling
VantageScore 4.0 excludes all paid collections from scoring regardless of original balance or debt type. This applies universally: any collection showing a zero balance is removed from the scoring calculation. FICO 8 still penalizes paid collections above $100 at reduced weight, making this a primary driver of score divergence.
Medical collections receive special treatment even when unpaid. VantageScore weights medical collections less heavily than non-medical collections, recognizing that medical debt often results from insurance processing delays rather than irresponsible borrowing. An exponential time-decay function also reduces collection impact as the account ages, with the steepest reduction in the first 24 months.
These collection handling differences produce the largest systematic score variance between VantageScore and FICO for consumers with collection tradelines. A consumer with two paid collections and one unpaid medical collection might score 40-60 points higher on VantageScore than FICO 8 purely from these algorithmic differences.
- All paid collections are excluded from VantageScore 4.0 regardless of original balance or type
- Medical collections carry reduced coefficients compared to non-medical collections when unpaid
- Collection aging follows exponential decay with steepest impact reduction in first 24 months
- Collection treatment is the largest driver of VantageScore-FICO variance for affected consumers
- FICO 8 still penalizes paid collections above $100; only FICO 9+ ignores them
Крок 6. Market Adoption and Consumer Exposure
VantageScore adoption has grown to over 2,400 lenders and financial institutions. However, approximately 90% of the largest U.S. lenders still use FICO as their primary underwriting score. FHFA's mandate for FICO 10T and VantageScore 4.0 in conforming mortgage underwriting represents the first time VantageScore will be used in mortgage origination at scale.
Free credit monitoring programs (Credit Karma, Capital One CreditWise, Chase Credit Journey) have made VantageScore the most consumer-visible model, providing VantageScore 3.0 or 4.0 to hundreds of millions of consumers. This creates a paradox: consumers monitor and optimize for a VantageScore that may not be the model their lender uses for underwriting.
The consumer-lender model mismatch is one of the most important concepts in credit scoring literacy. A consumer who sees a 740 VantageScore on Credit Karma and assumes equivalent creditworthiness under FICO may be surprised when a lender evaluates them at 700 on FICO 8 due to paid collections or recent credit card inquiries that VantageScore handles more favorably.
- Over 2,400 lenders use VantageScore for some purpose; 90% of largest lenders still use FICO primarily
- FHFA mandated that GSEs accept both FICO 10T and VantageScore 4.0 for mortgages
- Free monitoring programs have made VantageScore the most consumer-visible model
- The model a consumer monitors is often not the model their lender uses for underwriting
- Consumer-lender model mismatch creates false expectations about creditworthiness assessment
