What happened.
Home Capital identified failures to perform required income verification and the falsification of information within its underwriting process. Broker terminations reduced originations, but public disclosure did not promptly explain the underlying investigation. The OSC settlement recorded admissions about the company’s disclosure and governance failures.
The episode then moved beyond individual mortgage files. A loss of depositor confidence created a funding crisis, followed by emergency liquidity and recapitalization. The C$12.5 million regulatory settlement is distinct from the C$29.5 million class-action settlement; C$11 million from the regulatory settlement was directed toward that class recovery. These amounts should not be added as independent losses.
A verification record is only useful if it proves that the underlying check happened. When exceptions are concentrated in a distribution channel, file quality, broker oversight, escalation, and funding confidence become connected risks.
From individual facts to a fuller picture.
- Terminated brokers formed an identifiable origination cohort.
- Internal investigation findings and the related decline in originations preceded full public disclosure.
- Does independent income evidence support the application, and is its source retained?
- Do documentation exceptions cluster by broker, employer, branch, or underwriter?
- Who receives a material investigation finding, and how is its resolution recorded?
The review questions are analytical prompts. They do not establish that particular data was available to the institution at the time, or that a different process would have changed the outcome.
A different decision starts here.
Make unsupported income visible before approval, then connect repeated exceptions across the broker channel.
Authorized account transactions, income documents, employment evidence, broker identifiers, and the verification record available at each review date.
01 Reconcile the income claim
- Evidence to connect
- Application income, payroll deposits, employer evidence, and recurring debt payments.
- Amlok output
- A source-linked income and debt-service assessment with missing evidence and contradictions retained.
Decision it could supportRoute unsupported income to a verification exception before the institution clears the file.
02 Find the connected cohort
- Evidence to connect
- Broker IDs, employers, account owners, underwriters, and repeated documentation exceptions.
- Amlok output
- An evidence-backed relationship graph with uncertain links left for review.
Decision it could supportExpand a file-level investigation to the affected broker cohort without treating every linked borrower as fraudulent.
03 Turn exceptions into accountable action
- Evidence to connect
- Open findings, investigation dates, policy thresholds, and the original approval record.
- Amlok output
- A versioned exception history that shows which checks remain unresolved and who reviewed them.
Decision it could supportGive risk and governance teams a traceable basis to restrict a channel, re-review exposure, or escalate material findings.
Put numbers behind the opportunity.
These scenarios apply Amlok’s supplied comparison rates to the review described above. They are projections, not measured results at Home Capital / Home Trust. Count-based scenarios use an assumed population, not the institution’s actual records.
Debt-service coverage assessability
Assumed population: 1,000 eligible mortgage files.
More reviews could support an evidence-based affordability decision instead of proceeding with an incomplete debt-service picture.
Calculation & assumptions
Comparison inputs: 70% → ≈85%.
Population calculation: 1,000 × 70% = 700; 1,000 × 85% ≈ 850. Difference: ≈150 additional.
Relative change: (85 − 70) ÷ 70 × 100 ≈ +21.43%.
15 percentage points higher than the input baseline. Assumes the supplied assessability rates transfer to mortgage files with comparable income and obligation evidence.
View the supplied comparison inputsGraph relationship precision
Assumed population: 1,000 accepted broker, employer, and account links.
A more accurate cohort gives investigators a better basis for finding repeated verification exceptions.
Calculation & assumptions
Comparison inputs: 80% → ≈95%.
Population calculation: 1,000 × 80% = 800; 1,000 × 95% ≈ 950. Difference: ≈150 additional.
Relative change: (95 − 80) ÷ 80 × 100 ≈ +18.75%.
15 percentage points higher than the input baseline. Holds the number of accepted links constant. Correct links require an independently reviewed relationship label; precision is not the share of all relationships discovered.
View the supplied comparison inputsFrom a projection to a validated result.
Replay files using only documents available before approval. Independently label income contradictions and broker relationships; compare assessable-file coverage, false links, and unresolved exceptions at approval.
The model addresses verification and escalation. It does not estimate a deposit run avoided or settlement costs saved. The supplied comparison rates are approximate scenario inputs; no institution-specific replay or confidence interval is available for these estimates. The two metrics describe separate dimensions and should not be added together.
Read the source material.
This case draws on an admissions-based regulatory settlement. It does not establish that any particular technology would have prevented the deposit run.
- 01OSC: Home Capital settlement agreementRegulatory settlement
- 02Norton Rose Fulbright: summary of the OSC settlementLaw-firm analysis of a regulatory document
- 03
Amlok’s interpretation is independent of the institutions discussed. Public records are limited to their stated scope and period. Historical figures describe the source events; the separately labeled Amlok scenarios are calculated projections. Institution names and logos do not imply affiliation or endorsement.
