> ## Documentation Index
> Fetch the complete documentation index at: https://docs.amlok.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Personal lending benchmarks

> Evaluate personal-loan affordability, delinquency and household cash-flow changes.

Evaluate unsecured installment loans separately from revolving credit and secured lines. Use one borrower-application snapshot for underwriting and one borrower-period for monitoring. Record joint obligations without counting the same payment twice.

## Financial-risk measures

| Measure | Definition and comparison rule |
| - | - |
| Debt-service burden | Scheduled recurring debt payments divided by supported recurring income for the same period. State gross or net income and include the proposed payment. |
| Residual cash flow | Supported inflows less essential outflows and scheduled obligations. Exclude loan proceeds and internal transfers from income. |
| Payment performance | Report first-payment default and 30-, 60- and 90-day arrears separately by origination vintage and seasoning. |
| Liquidity resilience | Measure available liquid resources against essential outflows; distinguish an observed balance from an available credit limit. |
| Loss | Report charge-offs and recoveries against an explicitly defined balance or origination denominator. |

## Public benchmark sources

The [Bank of Canada's financial stability indicators](https://www.bankofcanada.ca/rates/indicators/financial-stability-indicators/) distinguish borrowers with an account at least 60 days in arrears from product-level loans at least 90 days in arrears. Use the matching series; do not combine their rates. [CMHC's mortgage industry report](https://www.cmhc-schl.gc.ca/professionals/housing-markets-data-and-research/housing-research/research-reports/housing-finance/residential-mortgage-industry-report) also includes non-mortgage credit trends.

These sources provide market context. They do not establish a universal personal-loan affordability cutoff or expose the transaction histories needed to validate income recognition.

## AMLOK evaluation

Test income recognition, transfer exclusion, recurring-obligation extraction and residual-cash-flow calculations against adjudicated evidence. Score field accuracy and calculation error separately from credit-outcome prediction. Compare applicants with stable payroll, variable income and interrupted income as distinct cohorts.

For Signal, measure confirmed payment or income deterioration, false alerts per borrower-month and lead time before a defined arrears event. Retain data-connection failures as evidence-quality events rather than treating missing transactions as zero income.

Use the shared [evaluation design](https://docs.amlok.io/research/industries/index.md#amlok-evaluation-design). Completed results are listed in [measured results](https://docs.amlok.io/research/public-results.md); no personal-loan model performance is inferred from public aggregate rates.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.