Examining Price Appreciation in Condominiums for the Benefit of Low-income Households?

Transcription

1 June 2017 Examining Price Appreciation in Condominiums for the Benefit of Low-income Households? Darren K. Hayunga and R. Kelley Pace Dr. Hayunga is in the Department of Insurance, Legal Studies, and Real Estate at the University of Georgia, Dr. Pace is the LREC Endowed Chair of Real Estate in the Department of Finance at Louisiana State University, We thank the Housing and Demographics Research Center at the University of Georgia for financial support.

2 Background Home ownership being part of the American Dream is so prevalent in the United States that we seem to be a nation of homeowners and aspiring homeowners. The belief in home ownership comes from potential societal, familial, and financial benefits. Homeownership is said to encourage neighborhood involvement by residents, which builds communities. 1 Homeownership is also thought to create positive environments for families with children. 2 One of the most researched benefits of homeownership is its impact on wealth. Homeownership can contribute to financial prosperity by offering forced savings through mortgage payments, income tax benefits, and appreciation in the property value that can build wealth and act as a hedge against inflation and rising housing costs. Unfortunately, homeownership benefits have not been universally realized, in part because lower-income and minority households simply do not own homes to the extent of other households. Figure 1 shows that the ownership rate for households below the median income level is significantly and consistently less than the rate of the upper portion. This difference of approximately 30 percentage points has been true for several decades. Further, the homeownership rate is less for the lowest income households. Approximately 30% of the households in the lowest quintile own their own home. There are multiple reasons lower-income households do not own homes to the extent of others. As can be expected, a number center on financial considerations. A foremost concern is finding affordable housing. Stegman, et al. (2000) observe a severe shortage of housing units priced at levels affordable for working families. Another major issue is maintaining ownership. When considering whether homeownership is an effective means of building wealth for low-income households, Herbert, et al. (2013) conclude that ownership continues to be an important opportunity but the foremost 1 A 2011 survey for the National Association of Realtors found that, compared to renters, homeowners are more satisfied with various aspects of their community such as healthcare, educational opportunity, transportation, and arts and culture. The homeowners who responded also considered their communities to be safer and more stable than did renters. 2 In a 2012 Fannie Mae survey, the most cited reasons for wanting to own a home are to have a good place to raise children, a safe place to live, and more space for family. Additionally, homeownership has been associated with higher graduation rates, improved math and reading scores, and fewer behavioral problems (see Green and White (1997), Haurin, et al. (2001), and Herbert and Belsky (2008)). 1

3 requirement for these households is to maintain ownership. This can be a challenge for low-income earners due to less employment stability and the transient nature of the jobs they often hold. Other financial issues include flat-fee transaction costs that are a greater relative burden for poorer buyers and the loss of the tax deductibility of mortgage interest because some low-income taxpayers do not itemize their deductions. 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% Higher income Lower income 0% Figure 1. Percentage of homeownership split at national median income Source: US Census We examine in this study whether condominium can address some of these issues. Despite copious attention paid to low-income homeownership over more than three decades, there has been a paucity of consideration of condominium ownership. Affordable condominiums (not the deluxe high-rise units) tend to be smaller and lower in price than single-family residences in the same area. The lower cost will consequently help with the issues of affordable housing stock, transaction costs that are a percentage of price, and the mortgage and property taxes commitments. We are particularly interested in the maintenance aspect of condominium ownership. The ability to pay for repairs is an important factor towards maintaining homeownership. Due to the shortage of affordable units, many lower-income 2

4 homebuyers purchase properties in declining or distressed neighborhoods (Belsky and Duda, (2002), Van Zandt and Rohe, (2006) and Reid (2006)). Lower-income households thus typically own older homes in poorer condition requiring higher maintenance expenditures. Despite this fact, Louie, et al. (1998) find that lower-income owners repair their homes less often, spend less on housing repairs, and experience problems at more than twice the rate of higher-income owners. These findings apply largely to single family residences and not necessarily to condominiums. We find that affordable condominiums can be found in desirable neighborhoods with better schools, transportation, and safety. 3 Instead of neglecting maintenance, condominium fees should pay for the upkeep of the foundation, roof, common parking and roads, exterior walls, windows, and doors as well as landscaping and seasonal expenses such as snow removal. The fees at some units may pay for interior maintenance of common areas, utilities, elevators, and security. Overall, since condominium owners are required to pay for maintenance, the fees should help curb property depreciation, which may increase low-income homeowners equity wealth levels. Data We examine broker-assisted condominium transactions within the Dallas-Fort Worth (DFW) Metroplex from 2003 to We thank the Housing and Demographics Research Center at the University of Georgia for financial support to obtain three years of additional transactions ( ), which increases the sample by almost 70%, a total of 6,910 observations. Since our tests examine price appreciation and thus only properties in which we can observe both purchase and sale prices, the additional records prove integral when we split the sample into multiple cost levels. Findings To initially document condominium price moments, we use the Bailey, Muth, and Nourse (1963) method that models condominiums purchased in one period and sold in 3 In fact, the national median condominium price often is higher than the median single-family home value because condominiums are concentrated in higher-cost housing markets. Note, though, as shown in the National Association of Realtors metro area price reports, single-family homes typically sell for more than condominiums in a given area. 3

5 another. Figure 1 shows the price indexes by low-, medium-, and high-cost properties. Unless noted, all of our returns are in natural logarithms (continuous-time compounding) and deflated by the annual Consumer Price Index to produce real holding period returns (HPRs) net of inflation. As a reference, we also include over 55,000 real HPRs for all single-family transactions in the DFW Metroplex during the sample period. 4 Figure 1 shows that high-cost condominiums increase the most during the expansionary period from 2003 to During the same time, low-income condominium appreciate the least. All three cost levels of condominium experience considerably decreases after 2007 and then rebound beginning in 2011 or Each of the three condominium price levels ends the sample period at about 10% total reduction in value compared to the base year of 2003, which is not necessarily surprising given the loss in value during the so-called Great Recession, the price levels are in real terms, and the values include depreciation typical to durable goods. In comparison, single family residences did not increase as much from 2003 to 2007 but then experienced less loss during the contractionary period. By 2013, the price level of single family homes ends at about the same magnitude as the three condominium price levels. While the price trends exhibit losses of approximately 1% per year over the economic cycle, they are composite measures comprised of thousands of real HPRs of specific properties. To gain better insight we report summary statistics of each transaction grouped by the three cost levels in Table 1. For low-cost properties, we note that the mean of the real HPRs over the sample period is actually a positive return of 0.86% with a median value of 0.78%. While some properties experience large negative returns as indicated by the minimum value of % there are also numerous high positive returns, the maximum being 89.27%. The positive mean and median are especially noteworthy because the same measures are negative for the medium- and highcost properties. Medium-cost condominiums experience an average loss of 3.75% and high-cost properties realize a mean reduction in value of 6.51%. These measures of central tendency for low-cost condominium are encouraging and reinforce the thought that each property is a specific transaction. But are these 4 The sample data extends to 2016 but since we will want to model condominium fees and do not observe the data past 2013, we use the sale prices from 2014 to 2016 to compute the HPR but do not report post price movements. 4

6 properties suitable for low-income homeowners? The first metric we check is the purchase price. The statistics in Table 1 report that the low-cost units are affordable. The lowest price is $20,000 and the highest is $85,000. Low-cost condos High-cost condos Medium-cost condos Single-family homes Figure 2. DFW price indexes. 5

7 Table 1. Descriptive statistics of condominium grouped into three cost categories ( ) Standard Variable Obs. Mean Deviation Median Minimum Maximum Low cost Real HPRs (%) 2, Purchase price ($) 2,330 60,622 16,511 63,000 20,000 85,000 Square feet 2, ,120 No. of bedrooms 2, Home age (years) 2, Condominium fees ($annual) 2,330 2, , ,924 Fees/purchase price (%) 2, Medium cost Real HPRs (%) 2, Purchase price ($) 2, ,114 17, ,000 85, ,000 Square feet 2,289 1, , ,258 No. of bedrooms 2, Home age (years) 2, Condominium fees ($annual) 2,289 2,841 1,205 2, ,384 Fees/ purchase price (%) 2, High cost Real HPRs (%) 2, Purchase price ($) 2, ,103 99, , ,100 1,329,000 Square feet 2,291 1, , ,763 No. of bedrooms 2, Home age (years) 2, Condominium fees ($annual) 2,291 4,041 2,824 3, ,124 Fees/ purchase price (%) 2,

8 Eighty percent of local median income is often used by economists to designate a household as low income. Within the DFW Metroplex during the sample period, the median household income varied from approximately $55,000 to $60,000; therefore, 80% equates to an income of $48,000 or less. Let us consider a household with annual income of $42,000 or $3,500 per month. To purchase the highest priced condominium of $85,000, the monthly mortgage payment on 95% of the purchase price is $ using a 30-year fully-amortizing loan at 5% interest rate. Lenders of conventional financing generally allow 28% of the monthly income to be allocated for principal, interest, property taxes, and insurance. The 28% equates to $980 per month for our representative household. The mortgage amount is thus less than half of the maximum percentage and should allow comfortably for the additional expenses of private mortgage insurance, property taxes (should not more than $70 per month), condominium fees, and interior upkeep. Note that this example uses the highest priced units within the low-cost group and lower priced properties will further help the financial commitment. Given the reasonable monthly expenses, the bigger challenge for a low-income household may be obtaining a sufficient down payment: 5% of $85,000 equals $4, Select low-cost condominium should also be sufficient size to accommodate couples and small families. As can be expected, the square feet is less f0r the lower priced units but the median number of bedrooms is 2 and the median size is only about 100 square feet less than the medium-cost properties. One cost that will significantly impact low-income owners is the condominium fees. The average annual fees for the low-cost units is $2,339 or slightly less than $100 per month. We note that this amount fits within the budget in the above example. Real HPRs To better understand the real HPRs of the three cost groups, we bin the returns into categories for large loses, moderate decreases, increases, and substantial positive returns. Figure 3 details the percentage of real HPRs within each bin across the cost groups. From this breakdown, we see that the real HPRs for the high-cost group 5 The Federal Housing Administration has programs for qualifying buyers that allow for lower down payments with the trade-off being higher monthly mortgage costs compared to conventional financing. 7

9 experienced more large losses and materially less large positive gains than the other groups. At the other end of the spectrum, low-cost properties experienced about the same percentage of large losses as medium-cost units as well as less moderate losses and substantially more gains of 25% or more than the other two groups. Obviously the gains are dependent on the specific property but these percentages demonstrate that low-cost properties can provide opportunities to build equity wealth for low-income homeowners. We also compute the gains and losses across the economic cycle, which are not detailed for brevity and because there is no major difference across the economic cycle. We find that low-cost properties perform as well as higher priced units in both the expansionary and contractionary periods % Low cost Medium cost High cost 35.00% 30.00% 25.00% 20.00% 15.00% 10.00% 5.00% 0.00% More than 10% Loss Less than 10% Loss 0-10% Gain 11-25% Gain 25-50% Gain 50%+ Gain Figure 3. Real HPRs grouped by cost and return magnitude 8

10 Condominium Fees Structural maintenance is an important component of property ownership that helps counteract the natural depreciation from aging and wear and tear. We are interested in the condominium price changes in conjunction with the fees because, while the fees may seem high, they probably are a better measure of the true soft and hard costs necessary to mitigate depreciation. For example, respondents to the American Housing Survey estimate their median annual expenditure on maintenance is 0.64% of the purchase price of their home (Harding, et al. (2007)). We suspect homeowners of single family residences forget about some of the trips to the hardware store as well as the soft costs of their own time in owner-provided efforts involved minor carpentry, plumbing and so forth. In contrast, condominium fees should include the total costs of labor and material to maintain the common complex features along with certain services. The condominium fees will not generally include interior maintenance of a specific property so even the total annual fees will undercount the full maintenance costs to an owner. In Figure 4, we detail total annual fees for the DFW sample. Panel A reports the fees per unit, Panel B reports the fees scaled by square feet, and Panel C details the fees scaled by purchase price. The mean (median) total condominium fees scaled by purchase price for the entire sample period is 3.28% (2.63%), magnitudes much higher than those reported on the American Housing Survey and generally attributed to maintenance expenditures by homeowners. 6 The mean and median total annual fees are consistently increasing over the economic cycle. The average (median) percentage increase in Panel A is 2.92% (3.14%). Similar yearly increases are depicted in Panel B, 2.35% using means and 2.58% using medians. The main difference in Figure 4 is in Panel C. The mean and median ratios are largely stable from 2003 to 2008 and then increase due to the reduction in purchase prices from 2009 to An Internet search of annual maintenance costs for housing indicates that the standard rule of thumb is 1% of home value per year or $1 per square foot annually. 9

11 $4,000 Mean Median $3,500 $3,000 $2,500 $2,000 $1,500 $1, Panel A. $3.50 Mean Median $3.00 $2.50 $2.00 $1.50 $ Panel B. 10

12 4.50% 4.00% 3.50% 3.00% 2.50% 2.00% 1.50% Mean Median 1.00% Panel C. Figure 4. Mean and median annual condominium fees. Panel A details total fees per unit. Panel B details total fees scaled by square feet per unit and Panel C details total fees scaled by purchase price per unit. To investigate whether the attention to maintenance helps property values, we model real HPRs as a function of fees and many aspects of the specific properties, owners, and overall economy in multivariate regressions. Table 2 details the parameter estimates and standard errors using the full sample of condominium properties. In addition to the reported covariates, Model (1) includes economic variables that control for local conditions at the ZIP code level. These are the income and age levels of residents, percentages of African-American, White and Asian races, numbers of housing units and percentages of vacant or renter-occupied properties, levels of new construction, and changes in the months supply of housing. All of the models also include month and year binary variables (fixed effects). The parameter estimate on condo fees in Model (1) is a statistically significant positive value of Since the real HPRs and condo fees are both in log form, the interpretation is a 0.17 percent increase in real HPR for a ten percent increase in fees. 11

13 While not economically overwhelming, this results indicates that fees increase real HPRs when valued at the mean. Models (2) (4) replace the economic controls with ZIP code fixed effects, which saves degrees of freedom and does not reduce model precision. Models (3) and (4) also include a variable to measure the properties qualities. We note that purchase price and square feet are inverse collinear functions of real HPRs in Models (1) (3) so in Model (4) we replace the fee levels with the ratio of fees to square feet and consequently remove the square feet measure from the models. In all specifications, fees have a positive impact on real HPRs. Table 2. Regression models of log real HPRs (1) (2) (3) (4) Log annual condo fees ($) * * * (0.007) (0.007) (0.007) Log condo fees ($)/square feet # (0.007) Log purchase price ($) ** ** ** ** (0.009) (0.010) (0.031) (0.008) Log square feet ** ** ** (0.013) (0.013) (0.033) Quality ** ** (0.033) (0.012) Log holding period (days) ** ** ** ** (0.004) (0.004) (0.004) (0.004) Log TOM at purchase ** ** ** ** (0.002) (0.002) (0.002) (0.002) Log TOM at sale ** ** ** ** (0.002) (0.002) (0.002) (0.002) Home age at purchase * ** ** ** (0.004) (0.004) (0.004) (0.005) Pool on site ** (0.005) (0.005) (0.005) (0.005) Log garage spaces ** ** ** (0.006) (0.006) (0.008) (0.006) Cash purchase * (0.006) (0.006) (0.006) (0.006) FHA financing at purchase ** (0.007) (0.007) (0.007) (0.007) VA financing at purchase (0.033) (0.033) (0.036) (0.032) REO seller at purchase ** ** ** ** 12

14 (1) (2) (3) (4) (0.011) (0.011) (0.011) (0.011) Relocation seller at purchase * ** ** ** (0.018) (0.018) (0.017) (0.020) Corp seller at purchase (0.022) (0.022) (0.022) (0.022) Vacant at purchase ** (0.005) (0.005) (0.005) (0.005) Tenant occupied at purchase (0.012) (0.012) (0.012) (0.012) Old home ** * * (0.025) (0.029) (0.028) (0.031) New home (0.011) (0.011) (0.011) (0.012) Many bathrooms (0.017) (0.017) (0.017) (0.017) Many bedrooms (0.068) (0.053) (0.055) (0.054) Real T-bill rate ** ** * ** (0.001) (0.001) (0.002) (0.001) Constant ** ** ** ** (0.626) (0.093) (0.146) (0.107) Observations 6,851 6,903 6,851 6,851 Adjusted R Measured at the ZIP code level, Model (1) includes additional unreported covariates that are income and age levels, percentages of African-American, White and Asian races, number of housing units and percentage of vacant or renter-occupied, amount of new construction, and the change in the months supply of housing within the ZIP code. Models (2) (4) uses binary variables (fixed effects) at the ZIP code level in place of the economic measures. All of the models also include month and year dummy variables. Standard errors in parentheses are robust to heteroscedasticity. **, *, and # denote p-values <0.01, <0.05, and <0.10, respectively. To examine the impact of fees, we next model real HPRs for just the low-cost condominiums. Table 3 reports the findings. Model (1) uses the condo fee levels and Model (2) uses the ratio of fees to square feet. In both specifications, fees for low-cost condominiums have a greater impact on real HPRs than on the full sample. The parameter estimates in both models are approximately Overall, the results indicate that higher condominium fees are associated with higher real HPRs over the recent economic cycle. 13

15 Table 3. Real HPRs of low-cost condominiums (1) (2) Log annual condo fees ($) * (0.020) Log condo fees ($)/square feet * (0.020) Log purchase price ($) ** ** (0.059) (0.022) Log square feet ** (0.064) Log holding period (days) ** ** (0.008) (0.008) Quality ** ** (0.062) (0.027) Log TOM at purchase (0.004) (0.004) Log TOM at sale * * (0.004) (0.004) Home age at purchase * (0.023) (0.024) Pool on site (0.009) (0.009) Log garage spaces * (0.020) (0.019) Cash financing at purchase (0.011) (0.011) FHA financing at purchase * (0.013) (0.013) VA financing at purchase (0.063) (0.055) REO seller at purchase ** ** (0.018) (0.017) Relocation seller at purchase (0.059) (0.072) Corp seller at purchase (0.036) (0.037) Vacant at purchase (0.010) (0.010) Tenant occupied at purchase * (0.022) (0.022) Old home (0.158) (0.154) New home (0.099) (0.096) Many bathrooms ** ** (0.066) (0.067) 14

16 (1) (2) Constant ** * (1.670) (1.722) Observations 2,325 2,325 Adjusted R The models include binary variables (fixed effects) for the ZIP code as well as the month and year. Standard errors in parentheses are robust to heteroscedasticity. ** and * denote p- values <0.01 and <0.05 respectively. Conclusion While the lack of low-income and minority homeownership has been studied and government programs proposed for decades, the rate of homeownership by these groups has stubbornly stayed at approximately 30 percent less than higher-income households. This study puts forth condominium as an affordable alternative. Condominiums are generally lower cost than single family residences in the same geographical area and can be found in locations with good schools, transportation, and safety. Select properties also are sufficiently large to accommodate couples and small families. Condominium ownership involves the payment of fees for exterior maintenance and certain services, which we investigate as a possible advantage that helps restrain property depreciation and can mitigate the finding that low-income households do not maintain their homes as well as higher-income groups. For our sample across the economic cycle from 2003 to 2013, the results demonstrate that the holding period returns net of inflation are positive for low-cost condominiums while higher cost properties experience significant negative returns. Regression models indicate that higher fees increase the real returns of condominiums and more so for low-cost units. 15

17 References Bailey, Martin J., Richard F. Muth, and Hugh O. Nourse A Regression Method for Real Estate Price Index Construction, Journal of the American Statistical Association 58: Belsky, Eric S. and Mark, Duda Anatomy of the Low-income Homeownership Boom in the 1990s. In Low-income Homeownership: Examining the Unexamined Goal, Edited by: Retsinas, Nicolas P. and Belsky, Erick S Washington, DC: Brooking Institution press. Fannie Mae National Housing Survey: Future of Homeownership: Own-Rent Analysis. Green, Richard K. and Michelle J. White Measuring the Benefits of Homeowning: Effects on Children, Journal of Urban Economics 41: Harding, John P., Stuart S. Rosenthal, and C. F. Sirmans Depreciation of Housing Capital, Maintenance, and House Price Inflation: Estimates from a Repeat Sales Model. Journal of Urban Economics 61.2: Haurin, Donald R., Toby L. Parcel, and R. Jean Haurin The Impact of Homeownership on Child Outcomes, Joint Center for Housing Studies of Harvard University. Herbert, Christopher E. and Eric S. Belsky The Homeownership Experience of Low-Income and Minority Households: A Review and Synthesis of the Literature, Cityscape 10:2, 7 9. Herbert, Christopher E., Daniel T. McCue, and Rocio Sanchez-Moyano Is Homeownership Still an Effective Means of Building Wealth for Low-income and Minority Households? Homeownership Built to Last: Balancing Access, Affordability, and Risk after the Housing Crisis. Louie, Josephine, Eric, S. Belsky and Nancy, McArdle The Housing Needs of Low- Income Homeowners, Joint Center for Housing Studies, Harvard University W98 8. National Association of Realtors American Attitudes about Homeownership, Reid, Caroline K Locating the American Dream: Where Do Low-Income Homeowners Live? In Chasing the American Dream: Multidisciplinary Perspectives on Affordable Homeownership, Edited by: Rohe, William M. and Watson, Harry. Ithaca, N.Y.: Cornell University Press. 16

18 Stegman, Michael A., Roberto G. Quercia, and George William McCarthy Housing America's Working Families. Center for Housing Policy. Van Zandt, Shannon and William M. Rohe Do First-Time Home Buyers Improve their Neighborhood Quality? Journal of Urban Affairs, 28(5):