Mortgage loan approvals, denials, and pricing issues have become among the most volatile regulatory issues facing the lending industry.

Though the body of law known as Fair-Lending has been the law of the land for many years, disclosure of loan approvals and denials, required since 1991 by the Home Mortgage Disclosure Act (HMDA), has focused intense public attention on the subject.

There is virtually no disagreement on the goal of fair-housing legislation - to eliminate bias in loan pricing and underwriting. Equal access to credit and equal pricing of credit products on the basis of equal criteria are principles universally supported.

This also makes good business sense. Mortgage lending is a volume business, and the goal is to make every qualified loan. Mortgage lenders repeatedly state that denying a loan on any basis other than concern about timely repayment is a bad business decision.

But though the goal is widely supported, disagreement exists on how compliance should be measured and evaluated.


The source of much of the debate centers around the information released and made public under the disclosure act. Numerous groups have used HMDA data to draw conclusions about the "fair-lending compliance" of the lenders.

Such analyses make assumptions about the loan underwriting criterion in order to reach their conclusions.

Our experience indicates that a number of such generic assumptions may be wrong.

Each institution's lending criteria are unique. Even with the relatively consistent secondary market criteria providing some market guidelines, we have found no two lenders with identical lending criteria.

We have, however, found several common threads. Some of them surprised us.

One of the biggest surprises was the lack of correlation between income and the likelihood of loan approval. Repeated discussions with loan underwriters has indicated a minimal correlation between the two.

Statistical analyses of loan files indicate that income has a limited role or none in explaining loan decisions when other underwriting factors are also considered. Though HMDA-level data might seem to suggest that income affects the loan decision, additional loan information shows the effect to be negligible.

Though the reasons for denial vary, it is not uncommon to find that higher-income individuals applying for larger loans have virtually as many credit- and closing-related issues in their applications as do borrowers with less income.

Another surprise is the lack of correlation between loan approvals and net worth.

Intuitively, we expected the correlation to be very tight. It seems logical that higher-net-worth individuals are more likely to be approved.

What we have found is that net worth becomes a factor only in limited instances. A high net worth, when illiquid, does not offset the need for liquid assets to close the loan, sufficient monthly income to service the loan, or an unacceptably high loan to value ratio.

A third surprise is the level of scrutiny given to different categories of consumer debt.

In many instances some tardiness in revolving credit payments can be offset by consistent, regular payments on a home loan or monthly rent. The opposite is not true. Sterling charge card credit can rarely offset sloppiness in servicing a current mortgage loan or making rent payments on time.


The strongest correlating factors are less surprising, although their relative priority might be.

Overwhelmingly, the strongest correlation with loan denial is the lack of liquid assets to close the loan.

Though it is quite logical that this would be problematic, its emergence as a top priority occasionally surprises some lending institutions.

Another strong correlating factor is negative information in public records. Though all public records factors are detrimental to loan approval, the relative importance of these factors varies.

For instance, our analyses suggest that lenders view judgments more severely than collection problems. A collection can be initiated by any creditor; a judgment involves a finding after proper adjudication of the issue.

Debt-to-income ratios offer a better explanation of loan decisions than the absolute income levels.

The underwriting policies of most lenders include both the housing debt and the total debt ratios. However, our analyses indicate that for most lenders, the total debt ratio becomes the critical driving variable. Statistical tests indicate that for many lenders, a high total debt ratio can be compensated by a low loan-to-value, although the reverse is generally not true.


How the various criteria for loan approvals and denials are evaluated is of more than just passing interest. Perhaps of primary importance is the benefit to potential mortgage applicants.

Potential borrowers, particularly marginal borrowers interested in developing a credit record that will enhance mortgage approvals, will benefit greatly by understanding the relative importance of underwriting criteria.

Mortgage originators have multiple reasons for wanting more precise understanding of approval/denial standards. As lending criteria differ from one institution to another, mortgage lenders need to constantly monitor and quantify the critical success factors in their underwriting process.

Monitoring fair-lending compliance requires lenders to develop a methodology for comparing how underwriting criteria are administered to all applicants, including all protected classes.

Finally, a clear understanding of real-life underwriting will result in more clarity of public-policy interpretation of fair-lending laws and regulations.

Subscribe Now

Access to authoritative analysis and perspective and our data-driven report series.

14-Day Free Trial

No credit card required. Complete access to articles, breaking news and industry data.