Capitalization of above market financing: Condos and co-ops

Journal of Real Estate Research, The, 1998 by Kelly, Austin

Austin Kelly*

Abstract. Prices and characteristics were collected for two similar, adjacent buildings. One building, a co-op, has a master mortgage with a prepayment lock-out, while the other building, a condo, has no master mortgage. They provide a natural experiment to isolate the capitalization of financing terms. The research provides the clearest demonstration to date of the impact of financing terms on sales price. The value of the prepayment lock-out is estimated, using a stochastic simulation, as a function of the level of interest rates, rate volatility and time remaining on the lock-out provision. Prices for co-op units are found to fluctuate with the value of the prepayment lock-out. The value of the lock-out is overcapitalized in the price of co-op units. Co-op status reduces the value of apartments by about 9%.

Introduction

In 1967 a complex of three buildings, in the same architectural style, was built in northwest Washington DC. In 1980-81 two of the buildings were converted from rental property to owner-occupancy. One became a co-op, the other a condo. The coop obtained a 30-year master mortgage, at a then attractive 10% interest rate, with a 15-year prepayment lock-out, followed by a declining prepayment penalty. The condo had no master mortgage.

Since purchasing a unit in the co-op requires assuming a share of the master mortgage's payments, economically rational, fully informed buyers and sellers will capitalize the value of any unusual financing provisions in the sale price of the units. Previous researchers, examining below-market rate mortgage terms, calculated the value of the financing under the assumption that interest rates were not stochastic. This line of research usually found incomplete capitalization-sales prices were increased by the financing, but by less than 100% of the financing's value. This article suggests an alternative method for calculating the value of unusual financing arrangements attached to the purchase of owner-occupied housing. A dynamic stochastic simulation of the time path of interest rates is used to calculate an expected value of the difference in payment streams between the co-op's mortgage and a standard condo mortgage. This estimated value is then used in a hedonic regression to predict the selling price of units in the co-op. By adding data from the adjacent, virtually identical condo building, a sharper estimate of the value of the financing terms can be obtained, as well as an estimate of the effect of co-op status vs. condo status.

The remainder of the article is laid out as follows. The first section reviews the literature on pricing mortgage features. This is followed by a discussion of the model used to value the terms of the co-op's mortgage. Then the hedonic model used to test for capitalization effects is presented. The next section discusses the characteristics of the two VanNess buildings and describes the multiple listing service data used to test the market's value of the blanket mortgage. Finally, an empirical section tests the market's pricing of the co-op's financing.

Previous Research

A series of papers written in the mid-1980s, many collected in a 1984 special issue of Housing Finance Review, found incomplete capitalization effects for below-market assumable mortgages. Most researchers calculated the value of the below-market financing using the Cash Equivalence method, which assumes that the current difference between the mortgage rate and the market rate will continue for the remaining term of the mortgage. Smith, Sirmans and Sirmans (1984) and Strathman, DeLacy and Dueker (1984) made some adjustments to the cash equivalence measure, assuming that the below market mortgage would refinance in five years, and include an estimate of the tax consequences of the financing.

Cash Equivalence (or some variant) is then included in a hedonic regression with property characteristics, and the coefficient on the value of the financing terms is interpreted as a measure of capitalization. By and large, these articles find incomplete capitalization; the value of the property increases by some percentage of the cash equivalence value of the financing, with "adjusted" values coming closer to 100% capitalization. An interesting alternative method is that of Schwartz and Kapplin (1984), who match sales of Florida condominiums with concessionary mortgage terms to similar units with market rate mortgages, in place of using a hedonic regression to standardize the sales prices. In some ways this method yields a more precise estimate of capitalization than does the approach presented here, as the functional form of the hedonic regression is no longer an issue. However, their method breaks down if there is any unobserved systematic difference between units offered with and without concessionary financing.

Options models of mortgage prepayment suggest an alternative to Cash Equivalence for calculating the value of the financing in the face of volatile interest rates. The contractual provisions in the co-op's mortgage, which preclude refinancing for fifteen years (a prepayment lock-out), and assess a prepayment penalty afterwards, can be thought of as an option whose value is driven by a stochastic variable, the current rate of interest. Given the complexity of the declining prepayment penalty in the contract, a simulation strategy can be employed to "price the option." Similar techniques have been used by Berry and Gehr (1985), Hall (1985) and Leung and Sirmans (1990) to value simpler mortgage terms.