Skip to content
Alternative Finance

Underwriting · Funders and lenders

Underwriting decision matrix builder

This tool builds an underwriting decision matrix from criteria you define, one per line with a name, weight, minimum pass value and the file's actual value, then scores the file out of 100, marks each criterion pass or fail, applies your approve and refer thresholds and lets you download the matrix as a CSV template. It is for funders and lenders writing or testing a credit policy. Watch the fail column alongside the score: the weighted total can clear your approve line while a criterion you consider essential has failed, which is the signal to make that criterion a hard knockout.

Weighted score
84.6 / 100
Decision
Approve
Criteria failed
0
Total weight
100

Passing a criterion earns 60 percent of its weight plus up to 40 percent for margin above the minimum; failing earns a reduced share that falls with the shortfall. Edit the lines to build your own policy and download the matrix.

Decision matrix
CriterionWeightMinimumFile valueResultPoints
Months in business151230Pass15
Average monthly deposits201500042000Pass20
Negative days (inverse)1552Pass12.6
NSF count (inverse)1031Pass8.7
Owner FICO15550640Pass10
DSCR151.11.35Pass10.4
Positions (inverse)1021Pass8

Formula

For each criterion: pass if file value ≥ minimum (or ≤ minimum when the name contains "inverse"). Margin = (value − minimum) ÷ |minimum| for normal criteria, or (minimum − value) ÷ |minimum| for inverse ones.

Points when passing = weight × min(1, 0.6 + 0.4 × clamp(margin, 0, 1)): 60 percent of the weight for passing plus up to 40 percent for margin, maxing out at a margin of 100 percent above the minimum.

Points when failing = weight × clamp(0.6 + margin, 0, 0.5): at most half the weight, falling to zero when the shortfall reaches 60 percent of the minimum.

Weighted score = (sum of points ÷ total weight) × 100. Decision: Approve at or above the approve threshold, Refer at or above the refer threshold, otherwise Decline.

Worked examples

Example 1: the default seven-criterion matrix scores 84.6 and approves

Inputs: seven lines with total weight 100; approve at 75, refer at 55. Months in business, weight 15, minimum 12, value 30: margin = (30 − 12) ÷ 12 = 1.5, clamped to 1, so points = 15 × (0.6 + 0.4) = 15. Average monthly deposits, weight 20, minimum 15,000, value 42,000: margin 1.8, clamped to 1, points = 20. Negative days (inverse), weight 15, minimum 5, value 2: margin = (5 − 2) ÷ 5 = 0.6, points = 15 × (0.6 + 0.24) = 12.6. NSF count (inverse), weight 10, minimum 3, value 1: margin 0.667, points = 10 × (0.6 + 0.267) = 8.7. Owner FICO, weight 15, minimum 550, value 640: margin = 90 ÷ 550 = 0.164, points = 15 × (0.6 + 0.065) = 10.0. DSCR, weight 15, minimum 1.1, value 1.35: margin = 0.25 ÷ 1.1 = 0.227, points = 15 × (0.6 + 0.091) = 10.4. Positions (inverse), weight 10, minimum 2, value 1: margin 0.5, points = 10 × 0.8 = 8. Total points = 15 + 20 + 12.6 + 8.7 + 10.0 + 10.4 + 8 = 84.6; score = 84.6 ÷ 100 × 100 = 84.6 out of 100. All seven pass, zero criteria failed, and 84.6 is above 75, so the decision is Approve. Note that FICO and DSCR, both passing comfortably in absolute terms, earn only two thirds of their weight because their margins are measured relative to the minimum: 640 is just 16 percent above 550.

Example 2: five failures produce a decline at 40

Inputs: the same seven criteria and thresholds with weaker file values: months 8, deposits 22,000, negative days 9, NSFs 4, FICO 560, DSCR 1.0, positions 3. Months: 8 < 12, fail; margin = (8 − 12) ÷ 12 = −0.333; points = 15 × clamp(0.6 − 0.333, 0, 0.5) = 15 × 0.267 = 4.0. Deposits: pass; margin 0.467; points = 20 × (0.6 + 0.187) = 15.7. Negative days: 9 > 5, fail; margin = (5 − 9) ÷ 5 = −0.8; 0.6 − 0.8 is negative, clamped to 0; points 0. NSFs: 4 > 3, fail; margin −0.333; points = 10 × 0.267 = 2.7. FICO: 560 ≥ 550, pass; margin 0.018; points = 15 × 0.607 = 9.1. DSCR: 1.0 < 1.1, fail; margin −0.091; 0.6 − 0.091 = 0.509, clamped to 0.5; points = 15 × 0.5 = 7.5. Positions: 3 > 2, fail; margin −0.5; points = 10 × 0.1 = 1.0. Total = 4.0 + 15.7 + 0 + 2.7 + 9.1 + 7.5 + 1.0 = 40.0; score 40.0, five criteria failed, and 40 is below the refer line of 55, so Decline. The DSCR line shows the cap in action: a narrow miss still earns half the weight, which is deliberate, since a 1.0 coverage against a 1.1 minimum is an exception a human might approve, while the negative-day line, at nearly double the limit, earns nothing.

Example 3: a three-criterion matrix approves a file with a hard fail

Inputs: Years in business, weight 40, minimum 2, value 6; DSCR, weight 40, minimum 1.25, value 1.25; Guarantor FICO, weight 20, minimum 680, value 650; approve at 70, refer at 50. Years: margin = (6 − 2) ÷ 2 = 2, clamped to 1; points = 40. DSCR: 1.25 ≥ 1.25, pass with zero margin; points = 40 × 0.6 = 24. FICO: 650 < 680, fail; margin = −30 ÷ 680 = −0.044; 0.6 − 0.044 = 0.556, clamped to 0.5; points = 20 × 0.5 = 10. Total = 40 + 24 + 10 = 74; score 74.0; one criterion failed; 74 is above the approve threshold of 70, so the decision is Approve. This is the result the fail column exists to catch. If a 680 guarantor score is a genuine requirement of the policy, a weighted matrix will not enforce it, because a strong showing elsewhere can outscore the shortfall. The fix is to treat FICO as a knockout checked before the matrix, or to raise its weight until a fail cannot be outscored, or to set the policy so that any fail caps the decision at Refer. The matrix is a ranking device; knockouts are a separate layer.

What a decision matrix is for

A credit policy has to do two things at once: rule out files that fail essential requirements, and rank the rest so that pricing and approval authority can be set consistently. An underwriting decision matrix does the second. It lists the criteria the policy cares about, assigns each a weight that reflects its importance, states the minimum acceptable value, and converts a file's actual values into a single score. The score places the file in an approve, refer or decline band, and the per-criterion pass and points columns show why. The result is a decision that two underwriters would reach identically from the same inputs, which is the property fair-lending compliance, investor reporting and staff training all require.

The scoring rule in this builder is designed to reward margin as well as compliance. Passing a criterion earns 60 percent of its weight outright; the remaining 40 percent is earned in proportion to how far above the minimum the file sits, maxing out when the value is double the minimum (or, for inverse criteria, when it is zero). Failing earns a reduced share that shrinks with the size of the miss: a hairline miss keeps half the weight, and a miss of 60 percent or more of the minimum earns nothing. That shape distinguishes a file that scrapes past every line from one that clears them comfortably, and a near miss from a serious one, without requiring the policy author to write separate point bands for each criterion.

Points as a share of weight
SituationMarginShare of weight earned
Fails by 60% of minimum or more−0.6 or worse0%
Fails by 30% of minimum−0.330%
Fails by a hairjust under 050% (capped)
Passes exactly at minimum060%
Passes by 50% of minimum0.580%
Passes by 100% of minimum or more1.0 or better100%

Writing the criteria lines

Each line has four comma-separated fields: the criterion name, its weight, the minimum pass value and the file's value. Weights can be on any scale because the score is normalized to the total; using weights that sum to 100 makes the points column read directly as percentage points of the final score. Add the word inverse to the name of any criterion where lower is better, such as negative days, NSF count or open positions; the builder then passes the file when its value is at or below the minimum and measures margin as the distance below. The default matrix mixes four normal and three inverse criteria and is a reasonable starting point for a daily-debit product; a term-loan matrix would weight DSCR and leverage more heavily and might drop the statement behaviour lines.

Thresholds, knockouts and the refer band

The approve and refer thresholds define three bands. Above the approve line, the file may be approved within the underwriter's authority at the pricing tier the score implies. Between refer and approve, the file needs a second review and a documented reason; this is where exceptions live, and the per-criterion table tells the reviewer exactly which lines caused the shortfall. Below refer, the file is declined on policy. Setting the thresholds is a calibration exercise: run a sample of past files through the matrix and place the lines so that the files you would have approved land above the approve line and the files that defaulted land below it. If they do not separate, the weights need work before the thresholds do.

Knockouts belong outside the matrix. Any requirement that must be met regardless of other strengths, an active bankruptcy, an open tax lien above a size, a returned financing debit, fraud indicators, a prohibited industry, should be checked first and should end the review if failed. The third example shows why: a weighted score can clear the approve line while a criterion the policy regards as essential has failed. The merchant risk score calculator and the probability of default calculator are also ranking tools and have the same property. A simple rule many funders adopt is that any fail in the matrix caps the decision at Refer, which the builder does not enforce automatically but the fail count makes easy to apply.

From matrix to policy document

The download button produces a CSV of the matrix as scored: criterion, weight, minimum, file value, result and points. Strip the file-value and result columns and what remains is the policy: the criteria, weights and minimums that define the product. Version it, date it internally, and attach it to the credit policy so that every underwriter scores from the same sheet. When the policy changes, the diff between versions is the change log that auditors and investors ask for. The loan data CSV validator checks the exported file if it is to be loaded into a loan origination system, and the loan agreement generator and MCA agreement template generator cover the documents that follow an approval.

A written matrix is also the foundation of fair-lending compliance in business credit. The Equal Credit Opportunity Act and Regulation B, whose text the CFPB publishes, apply to commercial applicants and require that adverse action be explained on request. A matrix that records which criteria failed produces those reasons mechanically and consistently. The CFPB's Section 1071 rule will require covered lenders to report application outcomes and denial reasons for small-business credit, and a scored matrix is the simplest system that generates the data. The banking agencies' model risk guidance, collected on the Federal Reserve's supervision pages, describes the documentation, validation and monitoring expectations for any quantitative decision tool, and a matrix is one.

Calibrating and maintaining the matrix

A matrix is a hypothesis about what predicts repayment, and it should be tested. Score a sample of funded files at origination values and compare scores with outcomes twelve months later: the default rate should fall steadily as the score rises, and the refer band should contain a mix. Criteria whose pass or fail does not separate outcomes should lose weight; criteria that do should gain it. The step from a matrix to a statistical scorecard is the probability of default calculator, whose logistic form is what a fitted model of the same inputs looks like, and whose weights can be estimated from the same sample once it is large enough.

  1. Define knockouts separately and apply them before scoring.
  2. Draft criteria, weights and minimums; keep the total weight at 100 so points read as percentage points.
  3. Score fifty to a hundred past files and set approve and refer thresholds where the outcomes separate.
  4. Download the matrix, version it and attach it to the credit policy.
  5. Re-score quarterly on new outcomes; adjust weights before thresholds, and record every change.

Keep the matrix short. Seven to ten criteria capture most of the predictive power available from an application and statements; beyond that, criteria overlap and the weights become arbitrary. Keep the names plain so that a broker can be told which line a file failed, and keep the minimums at values an underwriter can verify from documents. The policy that is easiest to apply consistently is the one that is applied consistently, which is the whole point of writing it down.

Questions people ask

What is an underwriting decision matrix?

A table of the criteria a credit policy uses, each with a weight, a minimum pass value and the file's actual value, that converts a file into a single score and a per-criterion pass or fail. The score places the file in an approve, refer or decline band. It makes decisions consistent across underwriters and produces the reasons for an adverse action mechanically.

How are points awarded in this underwriting decision matrix builder?

Passing a criterion earns 60 percent of its weight plus up to 40 percent more in proportion to the margin above the minimum, reaching the full weight when the value is double the minimum. Failing earns a reduced share capped at half the weight for a hairline miss and falling to zero when the shortfall reaches 60 percent of the minimum. The score is total points over total weight, times 100.

What does inverse mean in a criterion name?

That lower values are better, as with negative days, NSF count or open positions. Adding the word inverse to the name tells the builder to pass the file when its value is at or below the minimum and to measure margin as the distance below it. Without the word, the builder treats higher as better, which is right for deposits, tenure, credit score and coverage.

Can a file be approved while failing a criterion?

Yes, if the weighted score clears the approve threshold despite the failed line, as the third example shows. That is why hard requirements should be knockouts checked before the matrix, or why a policy may cap any file with a fail at Refer. The fail count output exists to make that rule easy to apply. The matrix ranks files; it does not enforce essentials on its own.

How do I set the approve and refer thresholds?

Score a sample of past files at their origination values and compare the scores with how the loans performed. Place the approve line where the files you would approve again sit above it and the refer line where the files that defaulted sit below it. If the scores do not separate good from bad outcomes, adjust the weights first; thresholds cannot fix a matrix that does not rank.

Sources and further reading

Related tools and guides

More underwriting

Updated October 12, 2026 · By Alternative Finance editorial team

Next step

Turn the estimate into real offers

One request, several funding partners, side-by-side payback. Nothing is binding until you accept an offer.

  • Soft inquiry to request offers; a hard pull only if you proceed with a funder.
  • Decisions in as little as 24 hours for many products; funding often within days.
  • Use our calculators to check the true cost before you accept anything.

See what you may qualify for

Takes about a minute. Requesting offers does not affect your credit score. Free, no obligation.

We are not a lender or broker. Requests are forwarded to funding partners free of charge. How the site is paid.