Portfolio IQ
Turn every loan outcome into better risk intelligence.
Connect historical and ongoing loan performance with financial data and risk intelligence. Validate Default IQ or build a bespoke model against the portfolio you already have, then keep learning as new outcomes develop, with Inverite’s experts alongside your team.
The decision is only the beginning
Predict risk before the decision.
Scores each applicant to help guide the decision.Learn from what happens after it.
Analyzes actual outcomes during the initial launch of Default IQ and thereafter.A risk score helps guide the decision. The Loan Performance API transfers the raw loan and repayment data to Inverite, and Portfolio IQ analyzes it, connecting what happens after the decision back to what was known before it. The result is the evidence needed to validate and refine your risk strategy and guide future decisions.
Start with the portfolio you already have.
You don’t need to be using Default IQ today to get started. Historical loan performance can be used to validate Default IQ retrospectively. That retrospective data is provided through the Loan Performance API and Portfolio IQ, and can also be used to evaluate score-band performance, identify potential cut-offs and support bespoke model development. After implementation, the same connection supports ongoing review.
New to Default IQ?
- 1Historical portfolio provided via the Loan Performance API
- 2Validate Default IQ or build a bespoke model with Portfolio IQ
- 3Implementation of Default IQ
Already using Default IQ?
- 1New outcomes
- 2Portfolio IQ
- 3Monitor, review & refine
What’s possible depends on the historical outcomes, identifiers and financial data available for your portfolio.
Retrospective risk score validation
See what an existing portfolio can tell you.
In a retrospective analysis of one lender’s portfolio, Inverite compared Default IQ Flex scores with observed repayment outcomes for 11,720 originated loans. For this analysis, a loan was counted as in default if it was past due and had an outstanding balance greater than $0. The observed default rate was 3.2% in the lowest-risk score bin and 48.3% in the highest-risk score bin.
The same data shows the trade-off behind every cut-off: a lower threshold keeps fewer loans at a lower default rate, and a higher one keeps more loans with more risk.
Real customer analysis · single merchant portfolio · retrospective
Try it: Move the slider to select the highest score to include. See how many historical loans fall within that range and how many defaulted.
Above the cut-off: 2,959 loans at a 27.6% observed default rate.
Originated loans only. The slider shows observed outcomes for historical loans within the selected score range. Applicants the lender declined are outside the analysis.
| Score | Loans | Observed defaults | Observed default rate |
|---|---|---|---|
| 0 | 1,152 | 37 | 3.2% |
| 1 | 994 | 55 | 5.5% |
| 2 | 811 | 42 | 5.2% |
| 3 | 724 | 57 | 7.9% |
| 4 | 760 | 64 | 8.4% |
| 5 | 662 | 69 | 10.4% |
| 6 | 710 | 74 | 10.4% |
| 7 | 623 | 69 | 11.1% |
| 8 | 622 | 90 | 14.5% |
| 9 | 613 | 82 | 13.4% |
| 10 | 541 | 94 | 17.4% |
| 11 | 549 | 99 | 18.0% |
| 12 | 499 | 110 | 22.0% |
| 13 | 445 | 111 | 24.9% |
| 14 | 425 | 100 | 23.5% |
| 15 | 424 | 108 | 25.5% |
| 16 | 456 | 123 | 27.0% |
| 17 | 393 | 132 | 33.6% |
| 18 | 259 | 106 | 40.9% |
| 19 | 58 | 28 | 48.3% |
Hover or tap a bar for loans, observed defaults and observed default rate in that score bin.
Not perfectly linear, and that’s expected. Neighbouring bins sometimes move slightly out of order (1–2, 8–9, 13–14). Real portfolios behave this way.
This type of retrospective validation can be the starting point. Once a lender implements a score or risk strategy, ongoing performance data continues the feedback loop.
Results shown are from a retrospective customer portfolio analysis. Default IQ Flex was applied retrospectively to historical loans using only information available at origination. The resulting scores were compared with observed repayment outcomes. Every loan in this dataset was originated, so declined applicants aren’t represented. The cut-off view illustrates the analysis Inverite supports, not an automated decisioning tool; your team sets risk appetite and credit policy. Results vary by portfolio and historical performance does not guarantee future outcomes.
The lifecycle
One connection, before and after implementation.
Validation can begin with an existing historical portfolio. After implementation, the same connection keeps adding outcomes, so each review starts with more evidence than the last.
- 01
Validate
Historical outcomesTest Default IQ or a bespoke model against past performance.
- 02
Implement
Risk strategySet score bands and cut-offs aligned with your objectives.
- 03
Monitor
New outcomesKeep sending outcomes so they link back to the original decision.
- 04
Review & refine
Performance over timeReview results with Inverite and decide what, if anything, should change.
Portfolio strategy
Minimizing defaults isn’t the goal. Optimizing the portfolio is.
A lending business still needs to lend. Tightening underwriting may reduce observed defaults, but it can also reduce originations and funded volume and leave capital undeployed. Broadening the approval population can increase originations and deploy more capital, but may introduce additional risk.
Strategy changes as conditions change.
A lender may choose to broaden its approval population within its risk tolerance.
The same lender may become more selective and prioritize lower-risk originations.
Bring marketing into the same conversation.
Cost per application and cost per approval tell you how much it costs to attract applicants and turn them into approvals. Linking acquisition spend to repayment outcomes lets you also compare how much you spend to acquire each loan that repays.
When you provide acquisition channel data with your loan records, Portfolio IQ can compare observed repayment outcomes by channel. Use those comparisons alongside your acquisition costs to guide where you spend.
- 1Score applicants as they arrive
Default IQ scores each applicant, so risk is visible by source from the start.
- 2Track outcomes by channel
Portfolio IQ links repayment outcomes back to the channel recorded at origination.
- 3Use repayment outcomes to inform acquisition spend
Compare repayment outcomes and acquisition costs by channel to decide where to focus your budget.
More defaults with more originations isn’t necessarily bad, and fewer of both isn’t necessarily good. What matters is whether the incremental risk is worthwhile given your risk appetite, origination goals, available capital, acquisition costs and portfolio economics. Portfolio IQ brings that evidence into one conversation and Inverite helps you weigh the trade-offs; it doesn’t manage capital or make treasury decisions.
Monitor & review
Keep learning as new outcomes develop.
As ongoing loan performance becomes available, Inverite can work with your team to evaluate how the portfolio is behaving. Changes may indicate that deeper analysis or a model review is warranted; your team decides what to do about them.
Benchmark your portfolio against itself.
Compare repayment performance across vintages, products, channels, score bands, time periods and borrower segments to see where performance is improving and where it warrants a closer look.
What a periodic review can cover
- Score-band performance
- Portfolio mix
- Delinquency & days past due
- Default
- Losses & recoveries
- Vintages
- Products & channels
- Borrower characteristics
- Approval & decline populations
- Funded volume & loan size
What we can review depends on the data you provide. Comparing approved and declined applicants requires application and decision data for both groups.
Bespoke models
Turn historical outcomes into a model built for your portfolio.
Your loan outcomes, combined with financial and borrower data from origination, let Inverite’s data science team validate Default IQ or develop a bespoke model with you.
Questions we can help analyze
Bring us the questions your risk and finance teams are asking.
- How would Default IQ have performed on our existing portfolio?
- What cut-off historically aligns with our target default rate?
- Are we originating enough for the capital we have available?
- What incremental risk comes with expanding the approval population?
- Is portfolio performance changing from what we observed at validation?
- Should we revisit our model, cut-offs or strategy?
The Loan Performance API
The connection that starts — and closes — the loop.
A structured way to send loan and repayment outcomes to Inverite. Start with historical performance to validate Default IQ or build a bespoke model, then keep the connection in place so new outcomes link back to the data, score and decision behind them.
Expert partnership
Your data. Your portfolio. Our risk expertise.
Portfolio IQ combines connected data, risk intelligence and actual outcomes with Inverite’s expertise. The technology creates the connection; our team works alongside yours on the analysis. Your team makes the credit decisions.
We work alongside
- Risk
- Credit
- Finance
- Data science
- Portfolio management
To help
- Validate models
- Analyze score bands
- Evaluate cut-offs
- Investigate portfolio changes
- Support bespoke modelling
- Analyze portfolio strategy
Go further with Inverite
Related products
Your portfolio already contains the evidence.
Start with the loans you’ve already originated. Validate your risk strategy against actual outcomes, then keep the connection in place to learn from every loan that follows.