An alternative lender using fintech in the Canadian market had experienced strong growth and was looking to improve how they evaluate the credit quality of applicants. They had been using their own internally built score, which leveraged a number of data sources, to determine borrower credit quality and drive the deal price and structure. Over time the lender experienced a number of delinquencies and losses and realized that their score needed to be re-evaluated and optimized to better predict borrower default. In addition, the lender was looking to utilize new information to approve more deals. PayNet identified a solution to assist them in re-vitalizing their score, and increasing approvals.
Build additional data into their scoring model to improve predictiveness, mitigate losses, and increase approvals to grow revenue
Factor in low hit rates, due to the nature of their business, which hindered evaluating businesses with limited or no borrowing history
Implement a revised and improved credit score with the ability to automate adjudication on select application segments
Provided a data set allowing the lender to complete an analysis of the predictiveness of the PayNet MasterScore®
Compared the PayNet MasterScore and its underlying variables to the lender’s custom score to evaluate predictiveness gains
Completed retro analysis to demonstrate the predictiveness of PayNet MasterScore and identify potential revenue gain and loss reduction
The lender found that PayNet MasterScore registered a very predictive alternative lender ROC of 0.662
PayNet MasterScore identified higher risk borrowers more accurately than the lender’s internal score
The additional predictive value on just 96 deals alone made the business case to invest in the PayNet MasterScore and include it in their revised credit score
A Canadian Commercial Finance lender, who was also a PayNet Member, was undergoing an internal audit. One of the Key Findings/Action Items identified in the audit was to address the apparent lack of loan review and/or risk assessment process for commercial loans/leases on their books. The front-end credit adjudication process (leveraging data and predictive scores to evaluate and approve or decline credit) was deemed to be sound and high quality in nature, however audit identified a business risk because there was no loan review/assessment process once the account was booked.
Ensuring that the business had appropriate measures in place to evaluate the ongoing risk of existing accounts
Satisfying the audit by addressing action items
This PayNet Member had implemented the PayNet AbsolutePD® (APD) solution
APD risk rates their entire commercial portfolio: >Providing probability of defaults (PD’s) up to 8 quarters into the future >Without the need for financial statements >With the PD’s updated each quarter
APD was actively used within the organization to: >Provide early warning signs of credit deterioration >Identify accounts with positive migration for up sell opportunities >Identify accounts with negative migration for closer review >Assist in enhancing internal risk ratings
Lender educated its auditors about the PayNet Ratings that could be assessed for each commercial account
Auditors recognized the ratings as independent 3rd party risk assessments and strongly endorsed the frequent quarterly delivery of the PD Rating
A Captive commercial lender in the Canadian market was looking to improve credit decisioning by leveraging a predictive score during adjudication. Ultimately the lender aimed to improve credit performance, advance decision automation and reduce turnaround times with the end goal of increasing close rates and raising overall customer service levels.
Ensuring the Captive’s Credit Management team was comfortable with a true commercial credit score
Assessing PayNet MasterScore® to statistically validate its predictiveness (Power score/KS/ROC)
Implement a revised and improved scorecard with future potential to fully automate select application segments
Run a trial of the PayNet MasterScore for manual credit reviews, supported by training and review sessions delivered by PayNet’s in-house statistical modeling team. Perform custom analysis to substantiate the impacts and fully satisfy the Credit Management team
Run a retroswap analysis to determine the predictiveness of the PayNet MasterScore on the Captive’s portfolio and graphically show the rank ordering of risk
Provide a Dataset with quarterly archive data and scores to allow independent testing by the member of the data/score's predictiveness
The retroswap analysis allowed PayNet to show how highly predictive our data/score are. In this case our MasterScore generated an ROC greater than 0.7 and a KS over 0.39
The Captive’s in-house risk team found PayNet’s Dataset (data/score) to be 10-20% more predictive than alternatives
The PayNet MasterScore was included in the revised scorecard and is now leveraged on all commercial credit applications
A Canadian national commercial finance company was re-assessing its existing scorecard, which included multiple 3rd party data sources, with the goal of improving credit performance and automated processes.
The finance company was not interested in leveraging a 3rd party score.
The finance company decided to develop its own internal score utilizing both in-house and predictive 3rd party data variables.
PayNet requested and received an application file from the lender for a fixed historical period. PayNet matched that file against its database and returned a quarterly archive data set and scores to allow independent testing of the predictiveness of the data and scores within the datasets.
PayNet had previously run a retro/swap dataset to show the high predictiveness of PayNet data and scores specifically for this member’s data.
The finance company’s in-house credit risk teams found PayNet’s dataset (data/score) to be 10-15% more predictive than alternatives in the market. The PayNet MasterScore® was determined to be highly predictive, but given that the internal goal/direction was to build an in-house score, PayNet MasterScore was not leveraged.
The finance company analysis identified select PayNet variables it found to be highly predictive of default, and incorporated those into their automated scorecard.
Evaluation of the portfolio’s performance is ongoing.
A General Manager at a Major Bank Equipment Finance Company, which lends to the industrial and construction equipment sectors, was concerned that some good deals were rejected while some bad deals were approved.
Thin commercial information requires manual evaluation.
Credit policy requires underwriting to a 1.6% default rate.
PayNet analyzed 15,000 credit applications from a 12 month period to determine if credit decision practices were consistent with the company's business plan and CCO's objectives.
PayNet conducted retro and swap set analysis of 12,420 deals booked during the same period and developed two credit decision scenarios.
Scenario 1 revealed bookings could be increased 6.0% or $8.8 million with unchanged losses.
Scenario 2 revealed the average loss rate of 2.37% could be reduced to 1.73% representing a decrease of losses from $3.4 million to $2.7 million, or 22.3%, while bookings remain constant.
The CEO of a major Canadian leasing company was losing volume as the vendor relationship weakened. The vendor was dissatisfied with excessive application processing time and was considering other sources.
50% of through the door applications had thin credit information on applicant and requests for additional financial information from the vendor were met with resistance.
Pending application decisions took three days to process at significant expense.
PayNet's Credit History Report provided comprehensive analysis of applicant's repayment history eliminating the need to request additional information from the vendor.
The additional information from the Credit History Report enabled the leasing company to increase decisions within vendor's 2 hour turnaround objective.
The decisioning rate within 2 hours jumped from 42% to 57% of through the door applications, representing a 36% increase in approvals.
Applications declined within 2 hours also increased by 25%.
The leasing company was able to achieve $528,000 in added volume per month.