Track the credit override rate
Paste or drop your decisions with the system decision, final decision and underwriter. See the override rate, who overrides most and in which direction, and whether overridden approvals went bad more often.
Credit decisions
3,202 decisions by 12 underwriters were read, 272 referred by the system and 0 rows with a decision that could not be read were left out.
Your decisions are read in this browser and never uploaded. Only counts go anywhere, and only if you ask to receive the result by email.
These are 3,600 made-up decisions by 12 underwriters over one month, with two underwriters planted far above the rest.
Override rate
7.3% of 3,202 decisions were overridden, with 141 low-side and 93 high-side. 2 underwriters are above the flag rate of 10%.
Override rate by underwriter
Each bar shows one underwriter’s override rate, split into low-side and high-side, and names in red are above the flag rate.
Bad rate of approved loans
Low-side overrides went bad at 14.2% against 4.3% for policy approvals, 3.3× the rate.
Override reasons
Underwriters
| Underwriter | Decisions | Overrides | Override rate | Low-side | High-side | Low-side bad rate | Flag |
|---|---|---|---|---|---|---|---|
| UW09 | 243 | 50 | 20.6% | 4 | 46 | 0.0% | Above rate |
| UW04 | 254 | 45 | 17.7% | 44 | 1 | 14.7% | Above rate |
| UW05 | 247 | 17 | 6.9% | 14 | 3 | 0.0% | Not flagged |
| UW07 | 303 | 18 | 5.9% | 13 | 5 | 0.0% | Not flagged |
| UW08 | 274 | 15 | 5.5% | 11 | 4 | 0.0% | Not flagged |
| UW12 | 295 | 16 | 5.4% | 9 | 7 | 37.5% | Not flagged |
| UW10 | 285 | 15 | 5.3% | 6 | 9 | 16.7% | Not flagged |
| UW03 | 273 | 14 | 5.1% | 10 | 4 | 33.3% | Not flagged |
| UW06 | 245 | 12 | 4.9% | 8 | 4 | 28.6% | Not flagged |
| UW02 | 289 | 14 | 4.8% | 11 | 3 | 16.7% | Not flagged |
| UW01 | 254 | 10 | 3.9% | 5 | 5 | 0.0% | Not flagged |
| UW11 | 240 | 8 | 3.3% | 6 | 2 | 16.7% | Not flagged |
Get the result by email
Record each override and its reason inside the credit decision. Then the rate, underwriter and outcome sit on the same record each week. Start your trial
Rate calculations
System approve or decline = System decision ∈ {Approve, Decline}
Override = Final differs from system
Override rate = Overrides ÷ Decisions
Low-side = System decision Decline → Final decision Approve
High-side = System decision Approve → Final decision Decline
Policy approval = System and underwriter approved
Low-side bad rate = Bad loans ÷ Low-side
Policy bad rate = Bad loans ÷ Policy approval
Above rate = Override rate > Flag rate & Decisions ≥ Fewest decisions
Rows the system referred have no system decision to override, so they are counted apart. Bad rates use only rows with a known outcome, since declined applications have none. A flag is a reason to look at the reasons and outcomes, and the flag rate is your own.
Credit override rate explained
A credit override is when the final decision differs from the system decision. The override rate is overrides divided by decisions where the system approved or declined.
The rate shows how often underwriters change system decisions. Compare low-side and high-side overrides, review rates by underwriter and check whether low-side approvals went bad more often than policy approvals.
Common questions
What is a credit override rate?
It is the share of decisions where the final decision differs from the system decision. The calculation counts system approvals and declines.
How do you calculate an override rate?
Divide overrides by decisions where the system approved or declined. An override is counted when the final decision differs from the system decision.
What are low-side and high-side overrides?
A low-side override is when the system declined and the underwriter approved. A high-side override is when the system approved and the underwriter declined.
How can I track underwriter overrides?
Compare the override rate for each underwriter and review the low-side and high-side split. The tool also flags underwriters above the flag rate you choose.
How do I compare bad rates for overridden loans?
Compare the bad rate of low-side overrides with the bad rate of policy approvals. The calculation uses only approvals with a known outcome.