Why income validation fails — and how Kirk Donaldson plans to fix it

Kirk Donaldson is no stranger to solving  problems. As the CEO and founder of Halcyon, Kirk has spent over two decades building solutions that help lenders operate with more confidence and less friction. His latest venture brings a solution to one of the industry’s biggest headaches: income validation. In this executive conversation, Kirk explains why so many post-close income validations fail, how Halcyon’s partnership with Fannie Mae tackles that issue at scale, and what lenders can gain from moving away from outdated processes.

HousingWire: Your most recent press release highlights a major milestone: Halcyon being one of the first to launch Fannie Mae’s new Income Calculator using IRS transcripts. But there’s a deeper issue behind this integration — struggles with income validation. How widespread is this problem across the industry, and what made Halcyon decide to tackle it?

Kirk Donaldson: Our story is quite funny, we didn’t set out initially to solve this problem at all, but happened to be in the right place at the right time. Our company initially started out exclusively in the tax preparation space and in doing the development work for automating tax preparation, the transcript inefficiencies in mortgage almost hit us over the head. The traditional route, with its high failure rates and transactional nature, allowed us to leverage the same process we needed for tax preparation and parley into a better user experience for the industry as a whole. Our authorization spans form types and years, facilitating a much more efficient data exchange.

HW: Six out of ten post-close income validations fail, which can lead to repurchases for lenders. How does Halcyon’s approach solve this problem in a way others haven’t?

KD: With our approach, the borrower gives consent for an authorization on file at the IRS, so the post-close QC issues go away. The transcript data can be provided without worrying about the expiration or data quality of the form i.e. the 4506 included in a closing package and whether it’s outside of the 120-day period, all issues that attribute to the validation failure. This provides assurances to the buyers and sellers of the mortgages along with the GSE’s that information can be independently verified with less than a 1% failure rate to adhere to their respective selling guides.

HW: This seems like more than a tech update. It feels like a larger shift in how lenders and GSEs validate income. What implications does this have for the future of underwriting processes and lender-GSE relationships?

KD: Well, for starters, there’s no manual intervention so you are only dealing with source-of-truth data. Human intervention at best leads to manual data entry errors & at worst opens the GSE’s up to potential fraud. Our solution is direct, unaltered, and therefore bedrock.  Underwriting can be simplified, costs can be lowered, and everyone can align on a single source of data.

HW: You’ve integrated Fannie Mae’s Income Calculator directly into the Halcyon platform. What does that look like for underwriters and processors in their day to day  — and how does it differ from traditional processes?

KD: A lot changes for the better. Currently processors are getting tax returns scanned and uploaded by the borrower, then sending them off for OCR, someone manually checks the OCR — it’s messy, takes too long and introduces errors. With the new transcript-only calculator, the data comes directly from the IRS, error-free, and often in minutes. This data automatically populates a handful of calculators, including Fannie Mae’s and customized ones. There’s no fat fingering, no typos. A borrower can land on your website, and potentially, within 15 minutes you can have an income calculation, using validated data with the ability to try different loan products.

HW: There’s growing pressure across the industry to modernize outdated processes. How is Halcyon addressing these needs for both traditional and self-employed borrowers?

KD: We came from the OCR business in a prior life and it is an outdated process.  Starting with the JSON and assessing what is missing that isn’t available on a transcript has started making the industry question the value vs. the effort.  We like to joke that the effort required to try to  “automate” for the day-trading miner with 17 rental properties by OCR’ing the tax return to get each rental address and percent depletion might not be a good use of technology rather than assessing what can be rationally analyzed with minimal effort.  The policies are being questioned for the first time in a very long time because there are real questions about the tradeoff between additional cost and actual value provided.  We’re not suggesting the criteria is any less comprehensive but simply whether the way it’s always been done still makes sense with better modernization.

HW: What is Halcyon doing to stand apart from the other vendors in this space and are there any other innovative products you’re bringing to the market?

KD: We have only scratched the surface. Tax transcripts were a  pain point for lenders, and a cash cow for vendors. Look how fast that has changed. We’ve delivered transcripts in under a minute, and at a far lower price point. We’ve invested heavily in borrower-led approval and our income analyzer, and the industry is starting to discover how powerful these tools are when they are working together; Validated and Calculated self-employed income eligible for Reps and Warrants within minutes and at a fraction of the cost.  This is all a small part of the innovation. We’ve also recently launched SigSign which is our digital signature platform with real-time identity verification. Originally built to support non-mortgage clients lacking disclosure packages for consent authorizations, the product has broad applicability across industries to capture early identification fraud off server for pennies on the dollar. Commoditizing these services to reduce the manufacturing cost of originating a loan is our goal, which ultimately serves the borrowers.

Related