Printer Friendly

Factors affecting borrower choice between fixed and adjustable rate mortgages.

The choice between fixed and adjustable rate mortgages can be a very confusing one for borrowers. It can also be a very costly one that could result in much financial distress if a less than optimal choice is made. Recent mortgage borrowers were surveyed to determine whether they could calculate the financial cost of mortgages and which factors influenced their choice. Insights in these areas may produce a better understanding of mortgage choice and be of use to consumer educators in helping consumers select a mortgage. Previous studies of mortgage choice (Albaum and Kaufman 1979; Brueckner and Follain 1986; Colton, Lessard, and Solomon 1979; Dhillon, Shilling, and Dirmans 1985; Horowitz 1985; Turner 1979) focused only on the economic aspect of the choice or were limited to the use of secondary data in estimating their models.

CONCEPTUAL MODEL

The conceptual model of this study is based on an economic environment in which lenders expect mortgage interest rates to rise over borrowers' average residencies in homes. These rates vary year to year, but the overall trend is upward. Hence, contract rates on adjustable rate mortgages are below those on fixed rate mortgages. The choice between fixed and adjustable rate home loans is hypothesized to be influenced by lender constraints, financial cost of mortgages, compatibility of mortgages to financial plans, and riskiness of mortgages.

Lender Constraints Regarding Mortgage Choice

Lenders limit mortgage choice by the type of loans they offer and the eligibility requirements they establish. Typically, lenders only offer fixed and adjustable rate mortgages. If lenders do not offer a choice between these mortgages, then borrowers have no choice if they want to purchase a home. Mortgage eligibility requirements set by lenders primarily concern the maximum loan for which borrowers may qualify. Lenders employ various methods to determine this figure, but one of the more important methods is the monthly mortgage payment to borrower income ratio. Beginning mortgage payments cannot exceed a certain percentage of borrowers' income. Because this method for determining eligibility is often the same for fixed and adjustable rate mortgages, in the economic environment of the model, borrowers can qualify for larger loan amounts with adjustable as opposed to fixed rate mortgages. Adjustable rate mortgages have lower initial interest rates and, consequently, lower beginning payments. Hence, if borrowers choose the maximum loan amount, they may be constrained by eligibility requirements and be more likely to choose adjustable rate mortgages.

Financial Cost Aspect of Mortgage Choice

For economically rational borrowers, differences in expected financial costs between fixed and adjustable rate mortgages are determined by these borrowers' expectations of future interest rates and length of residencies in homes. From the lenders' viewpoint, at the time of loan origination, the expected present values of payment streams on fixed and adjustable rate mortgages are equivalent, although the actual payment streams on the two loans differ (adjustable rate mortgages have lower mortgage payments in the earlier years and higher payments in later years, assuming an economic environment where interest rates rise over the average loan term). The expected present value equivalency is due to lenders incorporating their expectations of future interest rates and average residencies in homes in determining contract rates on fixed rate mortgages. Hence, if borrowers are economically rational and share these lender expectations, they would believe both fixed and adjustable rate mortgages have the same expected financial cost and be indifferent between the two loans.

However, if borrowers are economically rational and do not share lender expectations of future mortgage interest rates or length of residencies in homes, they would believe one mortgage type has a higher expected cost. For example, if borrowers believe current mortgage rates will rise by a lower percentage than lenders predict over the loan term, then it will take longer for these rates to equal the rate on fixed rate mortgages. In this case, if borrowers share lender expectations of residencies in homes or plan shorter residencies than lenders expect, then they would believe fixed rate mortgages have a higher financial cost than adjustable rate mortgages. If borrowers plan longer residencies than lenders expect, they would be indifferent between the two home loans.

There are nine possible borrower interest rate and residency scenarios, each implying that fixed rate mortgages or adjustable rate mortgages have the higher expected cost or that both mortgages have the same expected cost for economically rational borrowers. These scenarios are empirically defined later in this paper.

Financial Planning Aspect of Mortgage Choice

When choosing between fixed and adjustable rate mortgages, borrowers are likely to consider the financial planning aspect of this choice. Although some may see financial planning and financial cost aspects as identical components of mortgage choice, here they are seen as separate, but integrated components. Financial cost aspect involves estimating overall expenses that will be incurred with fixed and adjustable rate mortgages. In financial planning, borrowers reconcile expected future mortgage payments, income, and expenses for other than housing to ensure that difficulties in meeting mortgage payments do not arise.

There is no established theory to explain the precise mechanics of this financial planning process. Some borrowers may estimate future income and expenses and choose the mortgage type with expected payments that best fits the gap between the two, whereas other borrowers may select a mortgage type and then tailor future expenses to expected payments. Although the precise mechanics of the financial planning process are unknown, the presumed outcome is that, at initiation, borrowers' expected future incomes, expenses, and mortgage payments are reconciled for expected residencies in homes so that potential difficulties in ability to pay are minimized.

Risk Aspect of Mortgage Choice

Borrowers also consider risk when choosing between fixed and adjustable rate mortgages. Risk involves the variability of mortgage payments. As the model assumes an economic environment where interest rates are expected to rise over the average term of a loan, adjustable rate mortgages are more risky than fixed rate mortgages. Borrowers will select the type of mortgage that is in best accord with the level of risk they are willing to accept. Borrowers may employ numerous decision-making models. Borrowers who consider mortgage risk to be undesirable would prefer fixed rate mortgages.

DATA COLLECTION AND ANALYSIS

Based on the model of this study, data were collected to answer two questions. First, does the mortgage type borrowers believe has the highest expected cost correspond to the mortgage type that would have the higher expected cost, based on borrowers' interest rate and residency expectations? Second, how do the four factors (lender constraints, financial cost, financial planning, and risk) of the conceptual model affect mortgage choice?

To answer these questions a mail survey was designed and implemented in accordance with Dillman's Total Design Method (1978). The sample consisted of 358 fixed rate mortgage borrowers and 358 adjustable rate mortgage borrowers who acquired mortgages between January and November 1986, in Tompkins County, New York. Names of borrowers were obtained from photocopies of their mortgage documents on public record at Tompkins County Courthouse. Type of loan could be identified by the attachment of a special rider to adjustable rate mortgage documents. Only people who acquired a new mortgage because they bought a home or refinanced a previous mortgage were surveyed. People who acquired a second mortgage or a residence with more than three living units were excluded because they were thought to be unrepresentative of the average mortgage borrower.

To determine whether borrowers in Tompkins County, NY could judge which mortgage type had the higher expected cost, lender expectations of future interest rates, and average borrower residency in a home needed to be empirically defined. The contract rate on fixed rate mortgages is a measure of lender expectations of future mortgage interest rates over the average borrower residency in a home. The national average contract rate on fixed rate mortgages was between ten and 12 percent during the time borrowers in the sample acquired their mortgage (Freddie Mac Reports 1986). Contract rates on fixed rate mortgages available from lending institutions in Tompkins County during this period were similar. Lender expectations of borrower residency in a home are based on actual borrower residency for fixed rate mortgages--between eight and 12 years, with a median of ten years (Hendershott, Hu, and Villani 1983). These lengths of time are typically used by lending institutions in setting contract rates on fixed rate mortgages.

Borrowers were asked their expectations of mortgage interest rates (response options: under ten percent, between ten and 12 percent, or over 12 percent) and residency in a home (response options: under eight years, between eight and 12 years, or over 12 years). Based on their responses, borrowers were classified as having lower (interest rates under ten percent and residency under eight years), the same (interest rates between ten and 12 percent and residency between eight and 12 years), or higher expectations (interest rates over 12 percent and residency over 12 years) than lenders with regards to these two factors. Nine categories, measuring borrowers' interest rate and residency expectations relative to lenders and implying one mortgage type had the higher expected cost or they both had the same cost, resulted. Borrowers were also asked the type of mortgage they thought had the higher cost taking into account mortgage payments, points, and fees (response options: fixed rate mortgage, adjustable rate mortgage, or both mortgages had the same cost). To determine whether the mortgage type borrowers believed had the higher expected cost matched the mortgage type that would have the higher cost based on their interest rate and residency expectations, cross-tabulations were performed.

To determine how the four factors of the conceptual model affected mortgage choice, logit analysis was used (Maddala 1983). For the lender constraint of availability, fixed rate mortgage borrowers were asked if adjustable rate mortgages were available at lending institutions they went to, and adjustable rate mortgage borrowers were asked if fixed rate mortgages were available at lending institutions they went to. If a mortgage type was not available to borrowers, then they could not acquire it. For mortgage eligibility, borrowers were asked whether they acquired the maximum loan amount for which they were eligible. Borrowers acquiring the maximum loan amount were thought to be constrained by the eligibility requirement set by lenders. Because borrowers desiring the greatest loan amount possible would be more likely to choose adjustable rate mortgages, it was hypothesized that borrowers who acquired this maximum amount were less likely to choose fixed rate mortgages.

Borrowers were also asked which mortgage type they thought had the highest cost, best fit in their financial plan, and was more risky (response options: fixed rate mortgage, adjustable rate mortgage, or both mortgages). Borrowers who believed that fixed rate mortgages had the higher expected cost, adjustable rate mortgages best fit in their financial plan, and fixed rate mortgages were more risky were hypothesized to be less likely to acquire fixed rate mortgages. Likewise, borrowers who believed that adjustable rate mortgages had the higher expected cost, fixed rate mortgages best fit in their financial plan, and adjustable rate mortgages were more risky were hypothesized to be more likely to acquire fixed rate mortgages.

The survey was mailed in November 1986 to the sample of 716 fixed and adjustable rate mortgage borrowers. A postcard followup was sent one week later to those who had not returned the questionnaire. A total of 352 usable questionnaires were returned, a response rate of 49 percent. Of these, 180 were from fixed rate mortgage borrowers and 172 from adjustable rate mortgage borrowers.

Descriptive statistics of the socioeconomic characteristics of borrowers in the sample are in Table 1. Nationwide statistics regarding the socioeconomic characteristics of mortgage borrowers were not available so the representatives of the sample to the overall population of mortgage borrowers could not be determined. However, one may reasonably conclude that because the sample was drawn from a county where some of the major employers are educational institutions, college educated borrowers are overrepresented in the sample as compared to the general population of mortgage borrowers.

TABULAR DATA OMITTED

DETERMINING MORTGAGE COST

Cross-tabulation results of the type of mortgage with the higher expected cost (or that both mortgages have the same cost) based on each of the nine possible borrowers' interest rate and residency expectations versus the type of mortgage borrowers believed had the higher expected cost are shown in Table 2. For those instances where borrowers' interest rate and residency expectations indicate fixed rate mortgages have the higher expected cost (cases 1, 2, and 4), a clear majority of borrowers in the first case recognized the higher cost mortgage type based upon these expectations and a smaller majority did so in case 4. However, in both cases a substantial minority of borrowers did not recognize the higher cost mortgage type. Although case 2 had few respondents, it does indicate consumers were economically rational in this instance.
TABLE 2
Cross-Tabulation Results Regarding Expected Financial Cost of a
Mortgage
 Borrowers' Belief of Mortgage with
 Higher Cost
 Mortgage with
Case n Higher Cost FRM ARM Same
 (percent)
1 20 FRM 70 0 30
2 2 FRM 50 0 50
3 10 FRM/ARM 30 20 50
4 93 FRM 53 27 20
5 39 FRM/ARM 23 44 33
6 71 ARM 23 52 25
7 30 FRM/ARM 37 50 13
8 21 ARM 14 67 19
9 41 ARM 10 71 19
Case Definitions, borrower expects average mortgage interest
rates to be:
1 = |is less than~ 10% for next 10 years, plans to reside in
home |is less than~ 8 years
2 = |is less than~ 10% for next 10 years, plans to reside in
home between 8 and 12 years
3 = |is less than~ 10% for next 10 years, plans to reside in
home |is greater than~ 12 years
4 = 10-12% for next 10 years, plans to reside in home |is less
than~ 8 years
5 = 10-12% for next 10 years, plans to reside in home between 8
and 12 years
6 = 10-12% for next 10 years, plans to reside in home |is
greater than~ 12 years
7 = |is greater than~ 12% for next 10 years, plans to reside in
home |is less than~ 8 years
8 = |is greater than~ 12% for next 10 years, plans to reside in
home between 8 and 12 years
9 = |is greater than~ 12% for next 10 years, plans to reside in
home |is greater than~ 12 years


Of those cases in which borrowers' interest rate and residency expectations indicate adjustable rate mortgages have the higher expected cost (cases 6, 8, and 9), the majority of borrowers in these three cases recognized the higher cost mortgage type based upon these expectations. However, as before, a substantial minority of borrowers in all three cases did not recognize the higher cost mortgage type.

There are three cases in which borrowers should be indifferent between fixed and adjustable rate mortgages based on their interest rate and residency expectations (cases 3, 5, and 7). Only in case 3 were the majority of borrowers indifferent between the two instruments. In the other two cases, most borrowers believed one or the other mortgage type had the higher expected cost.

Of the six cases where the mortgage type with the higher expected cost is clear, based on borrowers' interest rate and residency expectations, the majority of borrowers sampled in Tompkins County correctly determined the higher cost mortgage type in five instances, even though a substantial minority did not. The number of observations in case 2 was believed to be too small to make any judgment. In the three cases where borrowers should be indifferent between fixed and adjustable rate mortgages, a large number of borrowers were not.

These results were similar by educational level of borrowers. The above analysis was repeated for borrowers whose highest educational level was a high school diploma or less and those with some college or more. Of the six cases where the mortgage type with the higher expected cost is clear based on borrowers' interest rate and residency expectations, the majority of borrowers in both educational groups correctly determined the higher cost mortgage type. Compared with borrowers with a high school diploma or less, a somewhat larger majority of borrowers with some college or more correctly determined the higher cost mortgage type. However, in both groups a substantial minority of borrowers incorrectly determined the higher cost mortgage type based on interest rate and residency expectations.

FACTORS AFFECTING MORTGAGE CHOICE

In the logit model determining how the factors of the conceptual model affected mortgage choice, the dependent variable specified that borrowers acquired the fixed rate mortgage (FRM). To gauge the impact of the four overall factors (lender constraints, financial cost, financial planning, and risk), dummy variables were created. Regarding lender constraints, mortgage availability was measured by whether fixed rate mortgages were available to adjustable rate mortgage borrowers (FRMAV) and whether adjustable rate mortgages were available to fixed rate mortgage borrowers (ARMAV). Borrowers acquiring the maximum loan amount for which they were eligible was used as the mortgage eligibility measure (ELIG).

For the expected financial cost of a mortgage, whether borrowers thought fixed rate mortgages had a higher total expected cost (FCFRM) or adjustable rate mortgages did (FCARM) was measured; the omitted category was that they thought both mortgage types had the same cost. Regarding the financial planning aspect, whether borrowers thought fixed rate mortgages best fit in their financial plan (FPFRM) or adjustable rate mortgages did (FPARM) was measured; the omitted category was that they thought both mortgage types fit equally well. Lastly, mortgage risk was gauged by whether borrowers thought fixed rate mortgages were more risky (RKFRM) or adjustable rate mortgages were (RKARM); the omitted category was that they thought both mortgage types had the same risk.
TABLE 3
Logit Analysis Results Regarding Choice of a FRM (N = 352)
 Estimated
Independent Variables Coefficient t-statistic
INTERCEPT -2.15(*) -3.54
ELIG 0.41 1.27
FCFRM -0.14 -0.31
FCARM 0.68 1.58
FPFRM 2.35(*) 6.01
FPARM -3.17(*) -4.07
RKARM 1.27(*) 2.16
Log likelihood -110.17
Chi-squared 267.46
Prediction success rate (percent) 87
*Significant at the .05 level.
Variable definitions:
FRM = 1, borrower acquired FRM, 0 otherwise
ELIG = 1, borrower took maximum loan amount eligible for, 0
otherwise
FCFRM = 1, borrower believed FRM had greater total expected
cost, 0 otherwise
FCARM = 1, borrower believed ARM had greater total expected
cost, 0 otherwise
FPFRM = 1, borrower thought FRM payments best fit with expected
future income and expenses, 0 otherwise
FPARM = 1, borrower thought ARM payments best fit with expected
future income and expenses, 0 otherwise
RKARM = 1, borrower thought ARM was more risky, 0 otherwise


The logit analysis results are presented in Table 3. It should be noted that because of a lack of variability, three variables had to be dropped from the model. These variables were fixed rate mortgages not being available to adjustable rate mortgage borrowers, adjustable rate mortgages not being available to fixed rate mortgage borrowers, and borrowers believing fixed rate mortgages were more risky. For the estimation, the log likelihood ratio is small enough to indicate a good fit. The chi-squared test also suggests the model fits the data well. Lastly, the prediction success rate indicates the model does a good job at predicting borrowers' choice of fixed rate mortgages. Whereas the floor value percentage of borrowers who acquired fixed rate mortgages was 51 percent, mortgage choice was correctly predicted 87 percent of the time.

Turning to the independent variables, the choice by borrowers in Tompkins County to take the maximum loan amount was not significantly related to choice of fixed rate mortgages. This finding is somewhat biased as only borrowers who were eligible for some type of mortgage were included in the sample. To the extent the variable reflects whether borrowers selected one type of home loan because they were ineligible for another, this aspect of the eligibility requirement set by lenders does not appear to have an effect on borrowers' mortgage choice. Given that the availability of fixed and adjustable rate mortgages also did not affect this choice (for the most part both mortgage types were available to borrowers), it seems that lender constraints as measured in this study were not really constraints at all.

Believing either fixed or adjustable rate mortgages had higher expected costs had no significant effect on borrowers' mortgage choice. Borrowers who believed fixed rate mortgages best fit with their expected future income and expenses were significantly more likely to choose fixed rate mortgages, whereas borrowers who believed adjustable rate mortgages best fit with their expected future income and expenses were significantly less likely to choose fixed rate mortgages. Borrowers who thought adjustable rate mortgages were more risky were significantly more likely to choose fixed rate mortgages.

SUMMARY AND IMPLICATIONS

Two overall findings result from this study. First, many borrowers are unable to correctly assess which type of mortgage had the higher expected cost, based on their interest rate and residency expectations. Second, lender constraints do not have a significant effect on mortgage choice nor does expected financial cost of a mortgage. Factors that do have a significant effect on mortgage choice are the financial planning aspect of this choice and the risk associated with a home loan.

Given that the financial planning aspect of a mortgage has such a large effect on mortgage choice, more research may be needed in this area. All that is known is that borrowers adjust expected future expenses, income, and mortgage payments in some way that mortgage payment difficulties are not expected to arise. Findings also imply that consumer education programs designed to help borrowers make a mortgage choice are needed. These programs should focus on estimating the expected financial cost of a mortgage. Many borrowers, regardless of educational level, could not determine which mortgage type had the higher financial cost. Borrowers need to become more knowledgeable about the financial cost aspect of a mortgage choice.

Mark Lino is an Economist, Family Economics Research Group, U.S. Department of Agriculture, Hyattsville, MD.

REFERENCES

Albaum, Gerald and George Kaufman (1979), "The Mortgage Acquisition Process: A Comparison of VRM and FRM Borrowers," American Real Estate and Urban Economics Association Journal, 7(2, Summer): 253-264.

Brueckner, Jan and James Follain (1986), "The Rise and Fall of the ARM: An Econometric Analysis of Mortgage Choice," Report Number 33-R (November), Urbana-Champaign, IL: University of Illinois, Office of Real Estate Research.

Colton, Kent, Donald Lessard, and Arthur Solomon (1979), "Borrower Attitudes Toward Alternative Mortgage Instruments," American Real Estate and Urban Economics Association Journal, 7(4, Winter): 581-609.

Dhillon, Upinder, James Shilling, and C. F. Dirmans (1985), "Choosing Between Fixed and Adjustable Rate Mortgages," Working Paper, Baton Rouge, LA: Department of Finance, Louisiana State University.

Dillman, Don A. (1978), Mail and Telephone Surveys: The Total Design Method, New York: John Wiley and Sons.

Freddie Mac Reports (1986), 4(January-November), Washington, DC: Federal Home Loan Mortgage Corporation: 1-11.

Hendershott, Patric, Sheng Hu, and Kevin Villani (1983), "The Economics of Mortgage Terminations: Implications for Mortgage Lenders and Mortgage Terms," Housing Finance Review, 2(2, April): 127-142.

Horowitz, Marvin (1985), "Economic Determinants of Residential Mortgage Choice," doctoral dissertation, Portland State University, Portland, OR.

Maddala, G. S. (1983), Limited Dependent and Qualitative Variables in Econometrics, Cambridge, England: Cambridge University Press.

Turner, Lloyd (1979), Consumer Preferences for Alternative Mortgage Instruments, Working Paper 19 (April), Iowa City, IA: Institute of Urban and Regional Research, University of Iowa.
COPYRIGHT 1992 American Council on Consumer Interests
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1992 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Lino, Mark
Publication:Journal of Consumer Affairs
Date:Dec 22, 1992
Words:3899
Previous Article:Consumption theories and consumers' assessments of subjective well-being.
Next Article:Homeless children: coping with material losses.
Topics:

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters