# Second Level Underwriting Review: When Files Need Two

- Author: TrueBankFlow (https://www.truebankflow.com)
- Published: 2025-05-31
- Updated: 2025-05-31
- Category: Loan desk operations
- Canonical: https://www.truebankflow.com/blog/second-level-underwriting-review

> When a second level underwriting review is required, what the second underwriter checks in income and statements and how to keep it fast.
A second level underwriting review is a check by a second underwriter, before the decision is final, on files above a set risk: large loans, exceptions, thin documentation or a newer first underwriter. The reviewer re-tests the income calculation, the bank statement findings and any exception, then signs or returns the file with written reasons.

## What is a second level underwriting review?

A second level underwriting review is a pre-decision check in which a second underwriter, usually senior to the first, reviews a file before the approval is issued. It differs from quality control, which samples closed loans afterwards. The second reviewer can still change the outcome, so problems are fixed before the borrower is told yes.

### Three reviews that get confused

- **First-level underwriting.** The assigned underwriter reads the file, calculates income, writes conditions and recommends a decision.
- **Second-level review.** A second underwriter checks that recommendation on selected files before it goes out. The approval carries both names.
- **Quality control.** A separate function samples closed or funded loans to test whether the process worked. It finds patterns; it rarely changes a single loan.

### Why lenders use it

A second signature catches arithmetic errors, missed statement items and exceptions that were granted without enough support. It also spreads judgment: newer underwriters learn from the notes a reviewer writes, and the lender gets a consistent standard across the desk.

### What it is not

A second review is not a full re-underwrite. The reviewer does not reorder documents or repeat every step. The reviewer tests the points where a mistake would matter most and the evidence that supports them.

## Which files trigger a second level underwriting review?

Files trigger a second level underwriting review when they cross a threshold the lender's credit policy sets in writing. Common triggers are loan amount, an exception to a guideline, alternative documentation such as bank statement income, a first underwriter below a set authority level, and layered risk such as high leverage with thin reserves.

### Typical triggers

The list below is an example of triggers a small lender might write into policy. Your own credit policy decides the actual list and limits.

- **Loan amount** above the first underwriter's individual authority.
- **Any exception** to a program guideline, such as a credit score, reserve or loan-to-value limit.
- **Alternative income documentation**, including bank statement income, profit and loss only and asset depletion.
- **Underwriter experience**, where files from an underwriter in their first months always get a second look.
- **Layered risk**, such as a high debt-to-income ratio combined with low reserves and recent credit events.
- **Unusual transactions**, such as a business purchase, a non-arm's-length sale or a borrower who is related to staff.

### Authority levels

Many lenders tie triggers to individual authority. An underwriter might approve up to one amount alone, need a second signature above it, and need a credit committee above a higher amount. Writing the levels down makes it obvious when a file needs a second reviewer and stops the question being argued file by file.

### Training files

New underwriters often have every file reviewed for a set period, then only triggered files once their error rate drops. How lenders structure that ramp, from shadowing to independent authority, is covered in [training new underwriters](/blog/training-new-underwriters).

## What the second underwriter checks in income and bank statements

The second underwriter checks that qualifying income was calculated from eligible sources with the right method, that bank statement findings were handled, and that any exception has a documented offset. On a bank statement file, the reviewer re-adds a sample of months, checks exclusions and confirms the expense factor or business percentage matches the program.

### Income checks

- **Method.** The averaging period, expense factor and ownership percentage match the program.
- **Exclusions.** Transfers between the borrower's own accounts, refunds, loan proceeds and one-time deposits were removed, and each removal has a reason.
- **Trend.** Recent months were compared with earlier ones. A declining trend may call for a shorter average or a lower figure.
- **Consistency.** Income on the application, the statements and any CPA letter or profit and loss statement tell the same story.

### Bank statement checks

- **Completeness.** Every month and page the program requires is present.
- **Large deposits.** Each one is sourced or excluded.
- **NSF and overdraft events.** The count is within the program's tolerance or has an approved explanation.
- **Recurring debits.** Payments to creditors appear in the debt-to-income ratio.

### How much to re-test

The reviewer does not need to re-add every deposit. Re-adding two or three months, including the largest and the most recent, usually shows whether the first calculation was done carefully. If those months do not tie, the reviewer goes further.

## A second level review worksheet for bank statement files

A second level review worksheet is a one-page record of what the second underwriter checked, what they found and what they decided. It turns the review into evidence. The example below shows a bank statement purchase file with one finding sent back to the first underwriter.

### Example worksheet

The entries below are illustrative values for an example file.

| Item | First underwriter | Second reviewer finding | Result |
| --- | --- | --- | --- |
| Program and trigger | 12-month business bank statement, exception for reserves | Trigger confirmed | Reviewed |
| Months present | 12 of 12, all pages | Spot-checked months 1, 6 and 12 | Agree |
| Average monthly deposits | $21,400 | Re-added months 6, 11 and 12; ties | Agree |
| Exclusions | $18,000 owner transfer in month 4 | Supported by savings statement | Agree |
| Large deposit, month 11 | Counted as income | No invoice in file; above pattern | Return: source or exclude |
| Expense factor | Program default | Matches guideline | Agree |
| NSF events | 2 in 12 months | Within program tolerance | Agree |
| Reserve exception offset | Strong credit, low DTI | Offsets documented | Agree |
| Decision | Approve with conditions | Approve after month 11 item resolved | Signed |

### What makes the worksheet useful

Each row names the item, what the first underwriter did, what the reviewer did to test it and the result. A later reader, whether quality control, an investor or an examiner, can see what the second signature covered. A bare initial on the approval cannot show that.

### Where the evidence lives

The reviewer should be able to open each document the worksheet cites without searching. A file indexed by document type and date, as described in the guide to [loan document indexing](/blog/loan-document-indexing), lets the reviewer jump straight from a worksheet row to the statement page behind it.

## Keeping second level underwriting review turn times short

Second level underwriting review turn times stay short when the first underwriter hands over a file that is ready to test: a clear summary, the calculation with its exclusions listed, the statement findings and the exception request with its offsets. A reviewer who has to reconstruct the reasoning takes days; one who only has to test it takes hours.

### What the first underwriter sends

1. A short summary: program, loan amount, trigger and recommended decision.
2. The income worksheet, with each excluded deposit listed and the reason.
3. A list of statement findings: large deposits, NSF events, transfers and how each was handled.
4. For an exception, the guideline it breaks and the offsets that support it.

### Set a target and a queue

Give second reviews a turn time target, for example one business day, and a visible queue. Assign a backup reviewer so a vacation does not stop every triggered file. Track how often files come back, because a high return rate points to a training need rather than a slow reviewer.

### Share the file safely

Second reviews often mean sending a file to someone in another office or working from home. Share it through the loan system or a controlled shared workspace, not by emailing statements as attachments. A misdirected email with a borrower's statements is a privacy incident, and the steps that follow one are set out in [data breach notification for lenders](/blog/data-breach-notification-for-lenders).

Reports that tie each total to the statement lines behind it make the re-test faster. [TrueBankFlow](/) produces monthly totals, averages and flags with links back to the source lines, so a reviewer can check a month without re-adding it by hand.

## When the second underwriter disagrees with the first

When the second underwriter disagrees with the first, the file goes back with a written reason, and the two resolve it by discussion or by escalation to whoever holds higher authority. The resolution is recorded on the worksheet. Neither underwriter should quietly overwrite the other's work.

### Kinds of disagreement

- **Factual.** A deposit was miscounted or a page was missed. The first underwriter corrects the figure and the reviewer signs.
- **Interpretation.** The two read a guideline differently, such as whether a deposit counts as business income. The credit policy owner settles it, and the answer becomes guidance for the whole desk.
- **Risk appetite.** Both agree on the facts but differ on whether an exception is acceptable. This goes to the next level of authority.

### Record the outcome

Write what was disputed, who decided and why. Over a few months these notes become the most useful training material the desk has, because they show exactly where judgment varies.

### Watch for patterns

If one reviewer returns far more files than the others, or one guideline causes most disputes, the problem is likely in the policy wording or the training, not in individual files. Fix the source, and the second review gets faster for everyone.