Latest Headlines

How Fannie Mae Desktop Underwriter Works: Findings, Approvals, and Referrals in the DU Decisioning Engine

A detailed market analysis of the Fannie Mae Desktop Underwriter platform, examining how DU processes loan data, generates findings, and produces the approval and referral decisions that shape conforming loan origination

Fannie Mae Desktop Underwriter is the most widely used automated underwriting system in the conforming mortgage market, processing millions of loan applications each year through its proprietary risk models and decisioning engine. This market analysis examines how DU works from data submission to decision output, covering the risk model architecture, the Approve and Refer decision pathways, the findings system, the representation and warranty framework, and the practical considerations for lenders and borrowers who rely on DU for conforming loan eligibility decisions.

Ebonie Beaco, Publisher & Mortgage Strategist
Ebonie Beaco

Mortgage Strategist | Senior Mortgage Loan Officer

Nov 13, 2026 8 minutes
How Fannie Mae Desktop Underwriter Works: Findings, Approvals, and Referrals in the DU Decisioning Engine

Fannie Mae Desktop Underwriter stands as the most widely used automated underwriting system in the conforming mortgage market, processing millions of loan applications each year and serving as the primary decisioning engine for loans that are intended for sale to Fannie Mae. Since its introduction in the mid-1990s, DU has evolved from a relatively simple rule-based system into a sophisticated risk assessment platform that incorporates trended credit data, automated valuation models, and continuously refined risk algorithms that draw on the massive Fannie Mae loan performance database. The platform evaluates borrower credit, capacity, and collateral through a proprietary risk model that estimates the probability of default based on the comprehensive set of loan and borrower data that the lender submits, and the risk estimate is compared to Fannie Mae established thresholds to determine whether the loan receives an Approve, a Refer, or a Refer with Caution decision. Understanding the internal workings of DU, from the data submission process through the decision generation and the findings output, is essential for mortgage professionals who use the platform to originate conforming loans.

The DU submission process begins when the lender enters the loan data into the loan origination system and transmits the data to the DU platform through an integrated interface that connects the lender system to the Fannie Mae servers. The data that is submitted to DU includes the borrower personal information, the loan amount, the loan term, the interest rate, the property type, the occupancy status, the loan purpose, the borrower income, the borrower assets, the borrower liabilities, and the credit report data that is pulled from the credit reporting agencies. The credit report data is a critical input, as DU uses the credit score and the credit account histories to assess the borrower creditworthiness, and the credit data is transmitted directly from the credit reporting agency to DU to ensure data integrity and to prevent the lender from altering the credit data before it reaches the AUS. The submission process is typically completed in seconds, and the DU response is returned to the lender origination system almost immediately, allowing the loan officer to review the decision and the findings while the borrower is still on the phone or in the office.

The DU risk model is the proprietary algorithm that evaluates the submitted loan data and estimates the probability that the loan will default, and the model has been developed and refined by Fannie Mae using the extensive loan performance data that the enterprise has accumulated over decades of mortgage purchasing. The risk model considers dozens of variables simultaneously, including the credit score, the debt-to-income ratio, the loan-to-value ratio, the reserve adequacy, the property type, the occupancy status, the loan purpose, the number of borrowers, the income source and stability, and the trended credit data that shows the borrower payment behavior over time. The model weights these variables according to their predictive power, as determined by the statistical analysis of historical loan performance, and the weighted combination of variables produces a risk score that reflects the overall probability of default for the specific loan. The risk score is compared to the Fannie Mae risk thresholds, which are the probability cutoffs that separate loans that are eligible for an Approve decision from loans that require a Refer decision, and the thresholds may vary depending on the loan program and the specific risk characteristics of the loan.

The Approve decision from Desktop Underwriter is the most favorable outcome that the system can produce, and it indicates that the loan, based on the data submitted, meets the Fannie Mae risk thresholds and is eligible for purchase by Fannie Mae subject to the satisfaction of the findings that accompany the decision. The Approve decision is the outcome that lenders and borrowers seek, as it provides the clearest path to loan closing and the broadest representation and warranty relief for the lender. The findings that accompany an Approve decision typically include requirements to verify the borrower income with appropriate documentation, to verify the borrower assets with bank statements, to verify the borrower employment with a verbal verification of employment, to obtain an appraisal or a property inspection waiver, and to satisfy any additional conditions that are specific to the loan. The Approve decision is not a guarantee that the loan will close, as the verified data must match the submitted data and all findings must be satisfied, but it provides a strong indication that the loan is on track for approval and closing.

The Refer decision from Desktop Underwriter indicates that the loan, based on the data submitted, does not meet the Fannie Mae risk thresholds and must be reviewed by a human underwriter who will determine whether the loan can be approved based on the underwriter judgment and the Fannie Mae manual underwriting guidelines. The Refer decision does not mean that the loan is denied, but it does mean that the automated system cannot approve the loan based on the data provided, and the loan must proceed through a manual underwriting review that involves a more detailed and subjective evaluation of the borrower financial profile and the loan characteristics. The human underwriter reviews the loan data, the credit report, the income and asset documentation, and the property information, and applies the Fannie Mae manual underwriting guidelines to determine whether the loan is eligible for approval. The manual underwriting process is more time-consuming than the automated process, and the outcome depends on the underwriter judgment and the specific circumstances of the loan, including the compensating factors that may offset the elevated risk profile that triggered the Refer.

The Refer with Caution decision from DU is an intermediate outcome that indicates the loan is borderline and that the risk profile is elevated but not sufficient to trigger a full Refer. The Refer with Caution finding instructs the lender that the loan requires human underwriter review, similar to a Refer, but the implication is that the loan is closer to the approve threshold and may be approvable with additional documentation or explanation. The interpretation of this intermediate decision requires the mortgage professional to understand the specific risk factors that DU identified and to work with the underwriter to address those factors through additional documentation, compensating factors, or loan structure adjustments that may improve the risk profile and support an approval decision. The Refer with Caution decision is particularly common for loans with debt-to-income ratios near the Fannie Mae threshold, for borrowers with credit scores in the mid-600s to low 700s, and for loans with high loan-to-value ratios combined with marginal credit profiles. The experienced mortgage professional can often identify the specific factors that triggered the Caution and can take steps to address those factors before resubmitting the loan to DU.

The DU findings system is the mechanism through which Desktop Underwriter communicates the specific conditions, documentation requirements, and verification steps that the lender must complete for the loan to be eligible for sale to Fannie Mae. The findings are generated by the DU risk model based on the specific loan data and the risk profile, and they are organized into sections that correspond to the major underwriting components, including the borrower eligibility, the property eligibility, and the loan eligibility. Each section contains a list of specific findings that the lender must satisfy, and the findings vary from loan to loan depending on the borrower characteristics, the loan structure, and the property type. The findings are presented in a standardized format that includes a description of the condition, the documentation that is required, and any specific instructions or limitations that apply. The mortgage professional must carefully review every finding, must ensure that the documentation collected satisfies the specific requirements stated in the finding, and must track the satisfaction of each finding through the condition clearing process to ensure that no finding is overlooked.

The income documentation findings that DU generates are among the most detailed and the most important, as the qualifying income is a primary driver of the debt-to-income ratio and the loan eligibility. For base employment income paid through an employer on a regular schedule, the DU findings typically require pay stubs covering 30 days and W-2 forms for the prior one to two years, along with a verbal verification of employment. For variable income such as bonus, commission, or overtime, the findings typically require a two-year history of the variable income, documented through W-2 forms and employer verification, to establish the stability and the average of the variable income. For self-employment income, the findings typically require two years of personal and business tax returns, and may require a year-to-date profit and loss statement and business bank statements to corroborate the income reported on the tax returns. The income findings may also include specific instructions for the treatment of non-traditional income sources, such as rental income from investment properties, boarder income, or income from a family member who is not a borrower on the loan, and the lender must follow these instructions precisely to ensure that the income calculation meets the Fannie Mae requirements.

The asset documentation findings that DU generates specify the documentation that the lender must collect to verify the borrower funds for the down payment, the closing costs, and any required reserves. The findings typically require two months of statements for all accounts from which funds will be drawn, and the statements must show the account balance, the account ownership, and the transaction history. The findings may also require documentation of the source of any large or unusual deposits that appear in the account history, with a large deposit typically defined as a deposit that exceeds 50 percent of the borrower monthly qualifying income or a fixed dollar threshold. The source of funds documentation must demonstrate that the large deposit came from an acceptable source, such as a transfer from another account owned by the borrower, the sale of an investment asset, or a gift from a family member, and the findings may specify the exact documentation that is required to verify the source. For borrowers who are using gift funds for the down payment, the findings will specify the gift letter requirements, the donor documentation requirements, and the transfer verification requirements that must be satisfied for the gift funds to be acceptable.

The property findings that DU generates specify the property valuation and inspection requirements that the lender must satisfy, and the findings may range from a property inspection waiver, which eliminates the need for a full appraisal, to a requirement for a full interior and exterior appraisal with specific appraisal review conditions. When DU issues a property inspection waiver, the finding indicates that the DU automated valuation model has sufficient confidence in the estimated property value based on the data available to the system, and that a full appraisal is not required for the loan to be eligible for sale to Fannie Mae. When DU requires a full appraisal, the findings will specify the type of appraisal that is required, any specific appraisal conditions that must be addressed, and any appraisal review procedures that the lender must follow. The property findings also specify any property-specific conditions that must be satisfied, such as requirements for condominium project approval, requirements for flood hazard determination and flood insurance, requirements for property repairs or inspections, and requirements for additional property documentation such as a survey or a homeowners association certification.

The DU version updates and model refinements that Fannie Mae releases periodically are an important consideration for mortgage professionals, as each new version of DU may incorporate changes to the risk model, the eligibility criteria, the findings logic, or the feature set that can affect the decision outcome for a given loan. Fannie Mae typically announces DU updates several weeks before the update takes effect, giving lenders time to review the changes, to test their loan origination systems against the new version, and to train their staff on any new findings or requirements. The DU update announcements include release notes that describe the changes in the new version, and the release notes are an important resource for mortgage professionals who need to understand how the updates will affect their loan origination process. Significant DU updates may include changes to the minimum credit score requirements, the maximum debt-to-income ratio thresholds, the property inspection waiver eligibility criteria, or the documentation requirements for specific borrower or property types, and mortgage professionals should review the release notes carefully and should adjust their origination practices to align with the updated requirements.

The integration of trended credit data into the DU risk model is one of the most significant enhancements that Fannie Mae has made to the platform in recent years, and it represents a substantial advancement in the ability of the AUS to assess borrower creditworthiness. Trended credit data provides a historical view of the borrower credit account balances and payment behavior over the prior 24 months, allowing DU to assess not just the current level of debt but the direction and the magnitude of changes in the borrower credit utilization over time. A borrower who has been actively paying down credit card balances over the prior 24 months presents a lower risk profile in the trended data analysis than a borrower whose credit card balances have been increasing, even if both borrowers have the same current credit score and the same current credit utilization. The trended credit data integration can improve the decision for borrowers who have been actively managing their credit, as the trended data shows a pattern of declining utilization that the traditional credit score snapshot does not capture, and this can result in an Approve decision for a borrower who might have received a Refer under the previous non-trended model.

The Fannie Mae Day One Certainty program is a value-added service that supplements the core DU decisioning and that provides lenders with additional certainty and efficiency in the origination process. The Day One Certainty program provides representation and warranty relief on certain loan components, fraud protection services, and validated third-party data services that can streamline the verification process and reduce the lender risk. The program includes services such as the validated income and asset data service, which allows the lender to obtain income and asset verification through approved third-party data providers and to receive representation and warranty relief on the verified data; the fraud protection service, which provides enhanced fraud detection and prevention capabilities; and the project standards service, which provides condominium project review and representation and warranty relief on the project eligibility determination. The Day One Certainty services are optional, and lenders can choose which services to use based on their origination model, their risk tolerance, and their cost-benefit analysis of each service.

The DU resubmission process is a common and important aspect of the conforming loan origination workflow, as the initial DU run is typically based on application data that has not yet been verified, and changes to the data are frequently identified during the verification process. When the verified data differs from the submitted data in a material way, the lender must update the loan file and resubmit the loan to DU with the corrected data, and DU will generate a new decision based on the updated information. The resubmission process ensures that the final DU decision reflects the verified borrower and loan data, and it is a critical step in ensuring that the loan is eligible for sale to Fannie Mae. Common reasons for DU resubmission include changes to the qualifying income after verification of the pay stubs and W-2 forms, changes to the asset balances after verification of the bank statements, changes to the property value after the appraisal is completed, and changes to the loan amount or the loan structure based on borrower decisions or market conditions. The resubmission process is typically quick, as DU can process the updated data and return a new decision in seconds, but the lender must ensure that the corrected data is accurately entered and that the new findings are reviewed and satisfied.

In conclusion, Fannie Mae Desktop Underwriter is a sophisticated and powerful automated underwriting platform that has transformed the conforming mortgage market through its data-driven risk assessment and its efficient decisioning process. The platform processes millions of loan applications each year, evaluating borrower credit, capacity, and collateral through proprietary risk models that have been refined over decades of development and testing, and producing decisions that determine whether conforming loans are eligible for sale to Fannie Mae and what conditions must be met for the loans to close. Mortgage professionals who understand the DU submission process, the risk model architecture, the decision outputs, and the findings system can use the platform more effectively to originate conforming loans, can anticipate the documentation requirements that DU will impose, and can guide borrowers through the underwriting process with greater efficiency and confidence. The continuous evolution of DU through version updates, trended credit data integration, and value-added services like Day One Certainty ensures that the platform will remain at the forefront of automated underwriting technology, and mortgage professionals who stay informed about the DU enhancements will be best positioned to leverage the platform capabilities for the benefit of their borrowers. This article is for educational purposes only and does not constitute a loan offer or financial advice. Borrowers should consult a licensed mortgage professional for guidance specific to their individual situation.

Ebonie Beaco — Mortgage Strategist

Have Questions About Your Mortgage Options?

Speak with Ebonie Beaco, Senior Mortgage Loan Officer (NMLS #2389954).

Mortgage Financing & Mortgage Education

Explore Mortgage Financing Options

Visit MortgageStrategist.net for professional services and resources.

Visit Site

This is a promotional link to an affiliated business of the publisher, separate from the magazine's editorial content.

Share

Editorial Disclosure: This article is published by Mortgage & Real Estate Strategist Magazine for informational and educational purposes. It does not constitute legal, tax, financial, or investment advice. Mortgage and real estate professionals should consult appropriate experts and verify current program guidelines before making decisions based on this content.

Ebonie Beaco, Publisher & Mortgage Strategist
Ebonie Beaco

Publisher & Mortgage Strategist

Ebonie Beaco is the founder and publisher of Mortgage & Real Estate Strategist Magazine. She is a licensed mortgage loan officer (NMLS #2389954) affiliated with NEXA Lending (NMLS #1660690), providing lending services in multiple states. She covers mortgage strategy, lending programs, and real estate investing.

Contributing AuthorProfessional Website