A lender scorecard is a structured way to rate lenders against a specific deal. For a commercial broker, it means scoring each candidate on fit (property type, loan size, leverage, geography), on execution (responsiveness, time to term sheet, closing rate) and on terms, then sending the deal to the top few instead of the whole list. The phrase also means other things in US finance, and mixing them up wastes time, so we separate them first and then show you how to build the broker version.
TL;DR
- "Lender scorecard" means three different things: HUD's FHA TOTAL algorithm, public-data grades of mortgage lenders, and a broker's own ranking of lenders for a commercial deal. Only the last one helps you place CRE loans.
- Banks have no single regulatory scorecard for commercial real estate. Their lending policy, with measurable limits on DSCR, LTV and debt yield, plays that role, and it tells you what a bank will say yes to.
- A useful broker scorecard has two layers: hard criteria that qualify or disqualify a lender, then weighted scores that rank the survivors.
- Keep the scoring rule-based and explainable. You have to tell a client why a lender was in or out.
- Track outcomes after every submission, so the scorecard improves with each deal.
Table of Contents
- Three meanings of "lender scorecard"
- Why commercial brokers need their own
- What regulators tell you about bank scoring
- Building the scorecard: two layers
- Match scores and matching algorithms
- Keeping the scorecard honest
- Our perspective
- Putting it into practice with BrokersConnect
- FAQ
- Sources
Three meanings of "lender scorecard"
Search results for this term are a mix, and the mix is the first problem.
The FHA TOTAL Scorecard. HUD describes the FHA TOTAL Mortgage Scorecard as a statistically derived algorithm that evaluates borrower credit history and application information. It is reached through an automated underwriting system and returns one of two results: Accept, or Refer to a Direct Endorsement underwriter. HUD is also clear that a mortgagee may not approve or deny a loan on the TOTAL result alone. The loan still has to be underwritten under the FHA Single Family Housing Policy Handbook 4000.1. It scores borrowers, not lenders, and it applies to FHA-insured one-to-four family mortgages, so it has no bearing on a retail strip center or a bridge loan.
Public-data lender grades. Consumer sites build scorecards from public records. Bankrate's mortgage lender ratings, for example, compare more than 1,900 lenders on pricing, complaints and related measures, and state that lenders cannot pay to rank higher. These draw on Home Mortgage Disclosure Act filings and complaint data. They are useful for a homebuyer choosing a retail lender. They say almost nothing about whether a debt fund will bridge a 60% LTV hotel acquisition in Phoenix.
The broker's working scorecard. This is the one you build yourself: a ranked view of lenders for a particular deal, based on what each actually does. The rest of this article is about that.
Why commercial brokers need their own
Commercial lending has no universal rating system and no TOTAL equivalent. Each lender, whether a community bank, credit union, life company, debt fund or private lender, writes its own box. Appetite shifts with capital availability, concentration limits and the market cycle. A lender that closed three multifamily bridge loans last quarter may be full this quarter.
The cost of ignoring this is time. Sending a deal to twenty lenders and waiting is slow, and the replies tend to be "not a fit" for reasons you could have predicted. A scorecard moves that work to before the submission.
What regulators tell you about bank scoring
Banks do not publish a scorecard, but they are required to have something that works like one. The OCC's Commercial Real Estate Lending handbook lists the information examiners review when overseeing a bank's CRE lending, including loan risk ratings, rating migration, concentrations of credit and supervisory LTV exceptions. In practice a bank's board-approved lending policy is its scorecard: it sets measurable standards by property type for DSCR, LTV, pricing and debt yield.
Three points follow for brokers.
- Policy limits are real. A bank that sets a maximum LTV for a property type cannot stretch it casually, because examiners look at how often exceptions are granted.
- Concentration matters. The Interagency Statement on Prudent Risk Management for Commercial Real Estate Lending notes that institutions with weak risk management and high CRE credit concentrations face a greater risk of loss and failure. A bank near its concentration comfort level will be choosy about what it adds, however good your deal looks.
- Measurable criteria are the language. DSCR, LTV and debt yield are how lenders talk about a deal. If your scorecard does not use them, you are not speaking to the credit committee. One capital advisory firm lays out the measurable underwriting criteria banks use on commercial real estate loans, including how actual and underwritten DSCR differ, which is worth reading before you quote a ratio to a lender.
Our own guides cover the main ratios in detail: debt yield versus LTV on submissions and DSCR versus NOI.
Building the scorecard: two layers
The mistake most people make is putting everything into one weighted score. A lender that cannot lend in your state should not be "78 out of 100". It should be out. So split the work.
Layer one: hard filters
These are pass or fail. If a lender fails any, drop it before you score anything.
- Property type. Multifamily, retail, office, industrial, hospitality, mixed-use, land, special purpose. Many lenders exclude one or more.
- Loan amount. Both a floor and a ceiling. A lender with a $5 million minimum is no use for a $1.8 million refinance.
- Geography. States, and sometimes markets or metro populations.
- Loan purpose. Acquisition, refinance, cash-out, construction, bridge to permanent.
- Maximum leverage. If the deal needs 75% LTV and the lender stops at 65%, it fails.
- Borrower profile. Minimum credit, experience, liquidity or net worth, and whether the lender accepts recourse or non-recourse structures for the deal.
Layer two: weighted scores
Once you have a survivors list, rank it. Our suggested categories, with starting weights you should adjust to your own book:
| Criterion | Suggested weight | What to score |
|---|---|---|
| Box fit | 30% | How comfortably the deal sits inside the lender's stated range, not just at its edge |
| Recent activity | 20% | Whether the lender closed similar deals lately |
| Responsiveness | 15% | Time from submission to first substantive reply |
| Terms | 15% | Rate, fees, amortization, prepayment, recourse |
| Closing reliability | 15% | How often their term sheets close at the quoted terms |
| Relationship | 5% | Your history and the quality of your contact |
Score each from 1 to 5, multiply by the weight, and total it. The numbers matter less than the habit: you write down why you chose these lenders, and the writing exposes weak assumptions.
A worked example
Say you have a $4.2 million refinance of a stabilized 24-unit apartment building, requesting 70% LTV, with a 1.30x DSCR and a borrower with two similar properties.
Start with ten candidates. Two lend only on commercial property over $10 million, so they fail on loan size. One does not lend on multifamily in that state. One caps leverage at 65%. That leaves six. You then score them: Lender A is squarely in its box and closed two similar deals recently, scoring high on fit and activity. Lender B sits at the edge of its maximum leverage, so it scores low on fit even though it passed the filter. Lender C is slow to respond, which costs it points on responsiveness, but its pricing is the best of the six.
You submit to A, C and a third lender, and you hold B in reserve. That is three submissions with reasons, instead of ten with hope.
For a DSCR-focused version of this process, see our lender shortlist criteria for DSCR and LTV. If you place investor loans, the DSCR loan guide for CRE brokers and the list of DSCR lenders for real estate investors apply the same logic to one product.
Match scores and matching algorithms
Some readers arrive asking about lender matching algorithms, match scores and "AI" matching. The idea is the scorecard automated: the software applies the hard filters, then ranks what is left.
There are two broad approaches.
Rule-based (deterministic) matching. Each lender's criteria are stored as rules, and a deal either meets them or it does not. The strength is that every exclusion has a reason. One open-source lender-matching project, Sanctioned, is built around this idea: it computes which lenders will fund a borrower, the maximum amount, the indicative rate and the binding constraint, deterministically and without machine learning. Knowing the binding constraint tells you what to change. If leverage is the constraint, you might add a mezzanine piece or shrink the loan. It is a developer's project for a different lending context, so treat it as an illustration of the design, not a CRE product.
Statistical or machine-learning ranking. A model predicts which lenders are likely to respond or close, based on past data. It can find patterns you did not think of, but a broker cannot defend "the model said so" to a client, and it needs a lot of clean outcome data to be trusted.
Our view is that the best approach combines both: rules for eligibility, where reasons matter, and learned signals such as responsiveness for ranking, where small differences are fine. When you evaluate any tool that promises a match score, ask three questions:
- Can I see why a lender was excluded?
- Where does the lender's criteria data come from, and who keeps it current?
- Does the score change when I change the deal inputs, so I can test scenarios?
Keeping the scorecard honest
A scorecard goes stale. Lender appetite moves faster than any spreadsheet. Build in three habits.
- Log every outcome. For each submission, note the date sent, date of first reply, response type (quote, pass, silence), and result. After 20 or 30 deals you have your own data on who answers and who closes.
- Record pass reasons. "Out of box" tells you the scorecard failed. "Too much leverage" tells you to tighten that filter. Our piece on matching lender appetite and avoiding concentration scrutiny explains why appetite changes.
- Review quarterly. Remove lenders that went quiet and re-check the ones whose boxes changed.
For construction work the filters change: the lenders' requirements for equity, guarantees and draw procedures dominate. Our guides to construction lender requirements and construction lenders for brokers show which criteria to add.
A clean package improves every score because it shortens the lender's first look. Use our loan document checklist and the DSCR submission checklist before you send.
Our perspective
We built BrokersConnect because we saw the same pattern in broker workflows: hours go to finding out who might lend, not to advising the client. A spreadsheet scorecard is a good start, and we encourage every new broker to build one by hand first, because the exercise teaches you how lenders think.
But a hand-built scorecard has limits. It lives in one person's head or file, it goes stale, and it cannot see what the rest of the market is experiencing. So we designed the platform around the same two layers described above: matching a deal scenario against a database of 289+ verified lenders on property type, loan amount, location, leverage, credit profile, loan purpose and structure, then giving you a lender responsiveness leaderboard so you can see who actually replies. We would rather you understand why a lender is on your shortlist than trust a black box, and that is how we think about every feature.
We also think pricing should not punish success. That is why we charge a flat monthly fee and take no commission or transaction fees on your deals.
— Theron
Putting it into practice with BrokersConnect
If you want the scorecard without maintaining it by hand, BrokersConnect is a commercial real estate broker platform with AI-powered lender matching, batch lender outreach, a deal pipeline CRM, a secure document vault, AI tools for deal packaging and underwriting, and a fee calculator. It costs $50 per month, with a free trial, and no commissions. You enter a scenario, see the lenders more likely to consider it, and track the conversations in one place.
FAQ
What is a lender scorecard?
It depends on context. At HUD it is the FHA TOTAL Mortgage Scorecard, which scores a borrower's FHA application and returns Accept or Refer. On consumer sites it is a grade assigned to a mortgage lender from public data. For a commercial broker, it is a ranking of lenders against a specific deal.
Is the FHA TOTAL Scorecard used for commercial loans?
No. HUD developed it to evaluate borrowers on FHA-insured mortgage transactions, and it is accessed through automated underwriting systems. Commercial real estate lenders set their own criteria, usually around DSCR, LTV and debt yield.
What criteria should a lender match score include?
Start with pass or fail filters: property type, loan size, geography, loan purpose, maximum leverage and borrower requirements. Then score the remaining lenders on fit, recent activity, responsiveness, terms and closing reliability. Weights should reflect what has actually decided your past deals.
Are AI lender matching tools better than a manual scorecard?
They are faster and can cover far more lenders, but only if they show why a lender matched or did not. Prefer tools with explainable rules and current lender data, and check their results against your own experience.
Sources
- HUD: FHA TOTAL Scorecard
- Bankrate: Mortgage Lender Ratings
- OCC Comptroller's Handbook: Commercial Real Estate Lending
- Interagency Statement on Prudent Risk Management for Commercial Real Estate Lending
- Sanctioned: a deterministic lender-matching engine (GitHub)