Abstract: Many approved credit deals fail to fund not because a competitor offered a better rate, but because the file was not ready to move quickly after approval. Missing or unresolved items—such as employment verification, insurance documentation, or seller confirmation—create delays while a competing lender advances directly to documents and funding. Yet these losses are often classified as “pricing pressure” without anyone verifying the competing offer. This misdiagnosis hides the real operational problem: lenders are incurring the full cost of adjudication on incomplete files that ultimately fund elsewhere. Credit shops should review unfunded approvals, verify whether rate was genuinely decisive, and measure how many post-approval conditions could have been resolved before submission.
Every credit shop has a number nobody owns, approvals that go out and never fund. Ask where those deals went and you will hear the same answer in every post-mortem. Pricing pressure. The client found a better rate.
Here is the version I watched from inside the funder, and later from the brokerage side. The client at the center of a competitive deal is holding more than one approval. From the moment the second one lands; the contest is not about price. It is about which approval becomes money first, and that is decided by how much touching the file needs after the credit decision.
An approval that goes out with three conditions enters a loop. The employment letter that cannot be verified through a bureau. The insurance certificate under a related company name. The seller nobody confirmed. Each item is a day, sometimes several. Meanwhile another lender’s approval, on a file that arrived needing nothing, is already at docs. The client does not wait out of loyalty. The machine needs to work. They sign whatever funds first without surprises.
Then the accounting happens. Your shop paid the full cost of adjudication: the analyst hours, the bureau pulls, the committee time. The asset sits on a competitor’s book. And because the client politely said they went with a better rate on the way out, the loss is booked as pricing pressure, and the response is a rate conversation about a deal that was never lost on rate.
Two details give the truth away. First, in file after file, nobody on your side ever saw the winning offer. The rate story travels unverified because it is convenient for everyone: the client avoids a conversation, the referral source stays blameless, and the loss fits a familiar column. Second, look at how fast post-approval questions get answered. When a condition is resolved the same day, it is asked, the information existed at submission. The file simply arrived without it, and your process discovered that after the decision instead of before it.
None of this is an argument that adjudication is doing anything wrong. The queue is deep and a fast first read is rational. It is an argument about where the loss actually lives. Approvals that age and die are not a pricing problem. They are a readiness problem that reaches your desk dressed as one.
If you run a credit shop, one exercise will show you your own version of this. Pull the last ten approvals that never funded. For each, two questions. Did anyone verify the competing rate. And how many conditions were resolved within a day of being asked. The first answer tells you how much of the pricing pressure is real. The second tells you how much of your adjudication spend is buying deals for your competitors.
The approvals your team issues are among the most expensive things your shop produces, and the ones that never fund cost exactly as much as the ones that do. Until the post-mortem stops taking the rate story at face value, that cost will keep hiding in a conversion number, filed under a market condition nobody can manage, instead of a file condition somebody can.
Author: Tejas (Teyjes) Virmani is the founder of Daixta and spent about a decade across the banking, funder, and brokerage seats of Canadian commercial finance. He writes Diagnostic Notes, a series on the operations of the channel.