When deciding whether to approve a loan, lenders typically evaluate borrowers in four key areas:
* Funds
* Income;
* Qualifying ratios;
* Credit.
In each of these areas, taking a "cookie-cutter" approach can have the unintended effect of reducing access to mortgage money for some credit worthy borrowers. Using one example in each area, the specific steps the lender takes to approve the loan are shown. (Wayne et al 1995)

Funds Saved During Loan Processing
Sam and Terry have $5,000 of the $9,000 they will need to close the loan on the new home they are buying. They have arranged to move in with relatives for the five months it will take the builder to complete their home. This arrangement will give Sam and Terry time to save more money, including the $800 they had been paying in rent and the $200 they had been paying for utilities and parking at their apartment. When they apply for their loan, they are confident that they will save at least the remaining $4,000 they will need to close.

In this case, Freddie Mac's guidelines provide the flexibility to consider the funds that will be accumulated, as long as the lender documents the pre-closing savings. Specifically, if the borrower has opened an account within 90 days of the date of verification of a deposit, or if an account has a balance significantly greater than it has had during the preceding months, the loan file must show that the funds are not borrowed. For example, if a portion of the borrower's funds were saved by the borrower between the date of loan application and the date of loan closing, then the file should document that funds were accumulated prior to closing. (Avery et al 1996)

Lenders apply this flexibility by having Sam and Terry provide additional account statements before closing. These statements must show the series of deposits totaling the additional $4,000 needed to close. Because an account statement showing a single deposit of $4,000 would not satisfy the requirement, the lender should counsel Sam and Terry to deposit the funds they are saving as they are received.

To avoid questions later, the lender could obtain a statement from Sam and Terry at the time of the application stating that the funds will be saved.

In underwriting a loan, income stability is more important than job stability in determining whether or not a borrower can be expected to repay a mortgage. A borrower who changes jobs frequently is not necessarily a greater risk. The job changes may be for advancement, and some borrowers may change jobs frequently as opportunities change. Also, job changes are recognized as normal in some lines of work. (Bernanke 1993)

A borrower who has demonstrated income stability in the past, who is in a line of work in continued demand, or who has a likelihood of advancement should receive favorable consideration if he or she has demonstrated the ability to manage financial affairs. Education or training that would strengthen job opportunities and earnings capacity should also be viewed favorably.

Stable income means income expected to continue for the next three years. Unless the), have evidence to the contrary, lenders can presume that an income level that has been stable for at least the past two years will continue.

In qualifying the borrower, lenders can use all sources of stable income that can be verified by either the payer or the borrower's income tax returns. Lenders verify the borrower's employment and income history for the two years preceding the application. It is good practice to look back three years or more if the borrower's income has fluctuated widely in the past year; this longer perspective may help establish an acceptable pattern for the borrower's earnings. (Caplin et al 1993)
Taco House has employed Steve for three months earning $1,800 per month.

Higher Qualifying Ratios
Ratios of a borrower's monthly housing expenses or total debt payments to income are one indication of a borrower's ability to meet mortgage payments and other obligations on time. Some borrowers can comfortably handle more debt than others, however, so that lenders should underwrite each loan on an individual basis.

In Hyung's case, the lender could use the rent verification as justification for the higher ratio and include an explanation that the borrower's proven ability to pay rent equal to 32 percent of income supports the 32 percent housing expense ratio.

Credit Alternatives
Melanie has been living with her mother for five years while attending community college. Her credit report reveals no lines of credit or credit accounts. She works part-time and makes regular deposits to her savings account. Melanie also pays her tuition on a budget payment plan and makes the utility payments at her mother's residence.

Despite the lack of information on Melanie's credit report to establish an acceptable credit reputation, Freddie Mac guidelines allow the lender to sell this loan if the lender can document the borrower's ability to make payments using situations in which regular payments (not necessarily monthly) have been made.
To sell this loan, the lender would obtain:
• Copies of bank statements, passbooks, dated deposit receipts or a letter signed by a bank official outlining Melanie's deposit history of savings deposits;
• Proof that Melanie made regular tuition and utility payments, such as canceled checks or money order receipts.

Evaluation Criteria For A Secondary Marketing System
One of the critical decisions a selection team must make is whether the secondary marketing unit will use the loan origination systems (LOS) as well or will continue to rely on a separate or specialized system. The advantage to an integrated system is that secondary marketing becomes fully integrated into the workflow, with the ability to exchange real-time information with the originators in negotiating final prices/locks. Similarly, changing credit/financial data obtained in underwriting can be used to re-price the loan. If the shipping function is part of secondary marketing, the loan information can be transferred seamlessly to the new area of responsibility--no redundant data entry. (Cunningham & Capone 1990)

However, the disadvantage of using an integrated system is that the secondary marketing component rarely offers full-fledged functionality, especially some of the more advanced features needed for true risk management. Some evaluation criteria for a secondary marketing system are offered.

Pricing: Pricing is becoming increasingly specific and variable, based on location (state/county/Zip), loan amount and specific loan parameters such as credit score and source (including partnerships with other mortgage lenders offering specialized products). The system should allow interest rates to be quoted in 16ths, and there should be an unlimited number of interest rates per product. It should be possible to do pricing adjustments across the board (e.g., up 4 points) with a single change.
The system should support yield spread calculations (above-par pricing). In addition, the ability to incorporate pricing based on credit risk factors such as LTV and credit score (risk-based pricing) is becoming increasingly important.

The system should allow customized price sheets to be composed for different originators (e.g., in-house loan officers, brokers and correspondents) and should allow these sheets to be faxed from directly within the system. Whether or not a lender is currently doing business on the Web, there should be an interface (HTML, XML, FTP) that would allow posting to a Web site. (DiPasquale & Somerville 1995)
Pipeline Tracking: Loans should be tracked by product type, location, channel (e.g., telemarketing versus retail versus wholesale), and originator. Management reports should provide a pricing history by product and by loan. Reports should also include a mark-to-market analysis, meaning the relationship between each loan's price and its current market value.

Rate Lock: At a minimum, these systems should offer real-time rate locking information so that the secondary marketing unit can respond to changing market conditions in a timely fashion. This becomes more crucial in a highly volatile market environment. Rate lock histories for at least six locks should track the time and date of the change, the person requesting/approving the change, and they should link to fallout statistics. The mortgage lender should have complete control over the lock periods.

Fallout: Statistics should be available by product, state, sales district, property time and interest rate change. They should track the type of fallout (e.g., withdrawn, denied). (Estrella 1997)

Funding: The system should have automated links to the general ledger, supporting HUD-1 reconciliation. The system should also include all fees, including those supporting purchased loan calculations.

Shipping: A sophisticated document-tracking system should not only note the current completeness of the loan file, but should also incorporate bar-coding or imaging that allows several people to work the file at once.

Automated Pool Allocation: As loans are locked, it would be ideal to have the loans automatically assigned to outstanding pools based on criteria defined by the secondary marketing specialist. It should also be very simple to move a loan or a block of loans from one pool to another and instantly assess the financial impact.
Loan Execution: The system should incorporate factors ranging from price, guaranty fee, buy-up/buy-down grids, excess servicing value, warehouse spread and float value in comparing various cash and mortgage-backed security (MBS) execution alternatives. (Fazzari et al 1988)

Hedge Optimization: A hedge is the creation of a position through the purchase or sale of financial instruments with a profit/loss profile equal but opposite to that of the mortgage pipeline across rate changes. Popular hedge instruments include cash and other forms of liquid assets, as well as MBS. No system can tell a secondary marketing analyst the ideal hedge strategy. However, it can support this effort by performing a continuous best-execution based on changes in spreads between hedge instruments. When the spread narrows or widens, the lender may execute swaps between hedge instruments, thereby profiting from risk-free arbitrage opportunities.

Best Execution: Best execution is the ability to maximize proceeds among several alternatives for secondary market sale. The evolution of the secondary markets for both loan and servicing assets, combined with a growing sophistication of the lender's ability to capitalize on arbitrage opportunities (i.e., the ability to profit from market inefficiencies), has made best execution an increasingly complex challenge. At the same time, given an increasingly competitive lending environment, it has never been more important. The added complexity exists because the decision to hedge loans, deliver loans and retain or sell servicing (once a single decision) is now essentially three separate but interrelated decisions. (Follain 1992)

To further explain, consider that traditionally the instrument used as a hedge became the delivery vehicle for the loan, which in turn determined the disposition of the servicing. Thus, lenders who hedged with a forward mortgage-backed security (MBS) sale delivered into that sale and kept the servicing. Although the result is often the same today, the secondary marketing manager needs to ensure he or she is maximizing net proceeds. That is because, for any given loan, the ideal hedge instrument is not necessarily the best loan sale execution. What's more, for that same loan, the best execution for servicing involves a different set of criteria.
A good best-execution model is one that can incorporate a variety of factors related to hedge efficiency, maximization of loan proceeds and maximization of servicing proceeds. These factors include:
• Monitoring spreads between various securities
• Guarantee fees
• Buy-up/buy-down grids
• Excess servicing value
• Warehouse spread

Servicing: The system incorporates differences in servicing-released premiums specific to targeted lenders, based on loan size, geographic location, escrow versus non-escrow, whether for Fannie Mae versus Freddie Mac versus third-party investors, etc. All these features can increase the productivity of the secondary marketing staff. However, the true value of these systems lies in their ability to help manage risk--the exposure associated with handling interest rate-sensitive financial instruments such as mortgages. (Follain & Tzang 1988)

Mortgages
To get one, lenders require that you have enough down payment and earn a sufficient income. What happens if lenders loosen these requirements a bit? In his research, the author finds that it would not significantly affect decisions on whether to buy a home, but it might encourage people buying a home to purchase a larger one.

Mortgage lenders typically require that potential borrowers satisfy minimum down payment and income requirements to qualify for a loan. Although there has been little empirical research on the effect of such requirements, it is widely felt that they significantly discourage homeownership and the housing expenditures of homeowners. The focus of my research is to determine as precisely as possible the impact of mortgage market constraints on households' housing decisions.

Interest in this research stems from the widely held belief that these constraints significantly affect households' housing decisions. For example, there is great interest in affordability indices that purport to measure the vitality and equity of housing markets. These indices are typically based on the "average" households' ability to qualify for a mortgage sufficient to purchase some minimum standard or "average" quality house. Thus these indices, and consequently peoples' perception of the housing market, depend critically on the extent to which mortgage market constraints are binding.

Concern over mortgage market requirements has focused primarily on their potential role in discouraging renters from becoming homeowners. For example, graduated payment mortgages have been introduced that, by offering low initial monthly payments on a fixed rate mortgage, is designed to help first-time homeowners overcome the 28 percent payment-to-income ratio hurdle. Legislation has also been proposed that would allow first-time homeowners to use their Individual Retirement Accounts (IRAs) as down payments.

Given the long-standing concern for homeownership in this country, this public-policy focus on first-time homeowners is probably appropriate. Once households obtain ownership status, generally they are financially able to remain homeowners. However mortgage constraints may affect housing expenditures as well as tenure choice, and existing as well as first-time homeowners. For example, the rapid appreciation of house prices during the later half of the 1970s encouraged existing homeowners to "trade up" (increase their housing expenditures). Existing homeowners generally had sufficient capital gains that they had little difficulty in meeting their down payments, but increasing interest rates and low-income growth combined to create problems with the 28 percent payment-to-income ratio. (Goodman & Ittner 1992)

There is casual evidence to suggest that mortgage market constraints play an important role in households' housing decisions. For example, although most U.S. households rent early in their life cycle, the majority become owners (approximately 64 percent of all U.S. households currently), and once they do they rarely revert to renting. There are several potential explanations for such behavior. Clearly one important factor is that housing capital gains are taxed if households revert to renting, while they are generally deferred if they remain owners. Also, households early in their life cycle are likely to be more mobile, and owning is relatively less desirable (more costly) for such households due to the high costs of buying and selling owner-occupied housing.

However, the typical pattern of housing tenure is also consistent with the view that most households simply prefer to own; they prefer to become homeowners once they can meet the burden of mortgage qualification. Specifically, households purchase a home once they earn income and amass a down payment (savings) sufficient to qualify for their desired/required-sized mortgage. To the extent that this appropriately characterizes household behavior, binding mortgage requirements may significantly deter home-ownership as well as consumption. (Green 1993)
The ratio of fixed to adjustable rate mortgage originations provides an additional indication of the role of mortgage market constraints. In particular, adjustable rate mortgages (ARMs) are originated far more frequently in higher interest rate environments. When interest rates are higher, ARMs represent as much as 60 percent of all mortgage originations, while they account for as little as 20 percent of originations when interest rates are low. Here again, household mobility is likely to play an important role in explaining this phenomenon-more mobile households will find the lower, short-term rates of ARMs more attractive, and this attraction will increase with interest rates. However, the lower rates of ARMs also offer a means of qualifying for a larger mortgage than would be possible with fixed-rate instruments. Likewise, the value of this option will increase as interest rates increase. (Maddala 1983)

The focus of my current research is to provide a more rigorous assessment of the importance of mortgage market constraints on housing decisions.

Do Lenders Impose Mortgage Requirements?
The primary reason lenders impose down payment and income constraints is a concern over default. Lenders typically lose money when borrowers default on their mortgages, so they try to limit this possibility.

There are two main theories explaining borrowers' incentives to default, the "equity" and the "ability-to-pay" theories. The equity theory views default as an option that borrowers will exercise only when it is to their financial advantage. In an options framework, default is viewed as "putting" the house to the lender. To illustrate, consider the situation of typical borrowers with an outstanding mortgage. They own a house (on which the lender has a lien), and owe lenders their current out-standing balance. In the simplest view, default is simply giving the house to the lender in return for declaring the debt paid. Obviously default is more complicated than this. In some states lenders can get deficiency judgments to make up for any difference between the value of borrowers' houses and their outstanding balances; defaulting on a loan requires borrowers to move; and there can be significant transaction costs to defaulting, including the loss of a good credit rating. (Peristiani et al 1996)
Regardless of these complications, default will be financially disadvantageous to borrowers if their houses are worth more than the outstanding balances on their mortgages. Consequently, the probability of default increases as the value of the house decreases relative to the outstanding loan balance. This provides the motivation for the down-payment constraint--the greater the down payment, the greater the difference between the value of the house and the outstanding loan balance at origination, and the smaller the probability that house prices will drop sufficiently in the future to encourage default.

In contrast, the ability-to-pay theory argues that borrowers default when mortgage payments are too burden-some, or more specifically when mortgage payments become too high a proportion of disposable income. Therefore, requiring that borrowers' payment-to-income ratios be less than 28 percent reduces the possibility that future fluctuations in borrower income (and interest rates in the case of ARMs) will cause a payment burden.

Empirical research on mortgage default suggests that lenders are wise to consider both factors when issuing mortgage credit. In general, the evidence is that, while it is extremely rare for borrowers to default when their house is worth more than their unpaid balance, not all borrowers with a negative net equity default. Default seems to require a combination of negative net equity (as suggested by the equity theory) and some precipitating event that makes the burden of mortgage payments too high (as suggested by the ability-to-pay theory). So by requiring a relatively large down payment (over 20 percent) and a relatively low mortgage payment-to-income ratio (under 28 percent), lenders can significantly reduce their risks from default. (Richard & Roll 1989)

As an aside, it is interesting to note that mortgage qualification requirements are a non-price method of rationing credit. That is, rather than charge a higher interest rate for borrowers who are a higher credit risk (are more likely to default); lenders simply make mortgages unavailable to them by using credit constraints.
A variant of this is exactly what is happening currently with construction loans--lenders are rationing credit to builders/developers because of their increased concern with default.

Economists have argued that credit rationing might make more sense than increasing interest rates due to asymmetric information between borrowers and lenders. In particular, borrowers differ in their propensities to default because, for example, they vary in their attitudes toward risk and their ethical standards. While borrowers may know this about themselves, these differences are unknown to lenders. Further, when credit is rationed, borrowers who are more likely to default will be more willing to pay higher interest rates for credit, since they value the default option more highly. As a result, those borrowers to whom lenders would least prefer to extend credit are the ones most likely to be willing to pay more for it. This potential for "adverse selection" discourages the allocation of credit solely through interest rates. (Schorin 1992)

Preliminary Empirical Evidence
I am currently studying the impact of mortgage qualification criteria on households' housing tenure choice and consumption using data from a nation-wide sample of approximately 800 households, collected during May of 1986. The data include information on characteristics of households as of May 1986, as well as changes in tenure status and housing expenditures since September 1983.

Taken together, these figures suggest that relaxing mortgage qualification requirements is likely to have a bigger impact on the housing expenditures and tenure decisions of renters.

Getting a better assessment of these impacts requires a more sophisticated analysis. Specifically, a multivariate analysis of households' mobility, tenure, and housing expenditures decisions is undertaken. In doing so, care is taken to allow for differences in preferences across households, and for measurement error, both in observed housing expenditures and in determining the largest mortgage households can borrow. (Avery et al 1996)

The results of this analysis suggest that households in the United States have, on average, approximately a 46 percent chance of being constrained by mortgage qualification criteria. Another way of viewing this is that approximately 46 percent of United States households are likely to be constrained by down payment or payment-to-income criteria. Being constrained means that, if households choose to purchase a house, they would prefer to obtain a larger mortgage than lenders make available to them. As expected, the analysis shows that renters and owners are differentially affected by mortgage market constraints--renters have on average a 65 percent probability of being constrained, while owners have a 3 $ percent probability.
To get a better understanding of the impact of mortgage market constraints, estimates of the effect of relaxing these constraints on households' homeownership and housing expenditures decisions are estimated. Again there are different impacts. Allowing existing homeowners to obtain a 10 percent larger mortgage than they would otherwise qualify for increases the probability of their purchasing a new house (during a 32-month period) only approximately 0.12 percent. The effect on existing renters is significantly larger, increasing their probability of purchasing a house by 2.36 percent. In neither case, however, is the effect particularly large.
The impact of relaxing mortgage constraints on households' housing expenditures is larger. A 10 percent increase in the size of the mortgage that existing owners can qualify for causes their expenditures on owner-occupied