This paper investigates the effects of cash-only transactions on residential property values before, during and after the financial crisis. Using a comprehensive database of residential property sales in Tallahassee, FL from 2006 to 2015 and a propensity score matching model, we find that, on average, cash-only transaction is associated with a 4.9% discount to the overall housing price. Further analyses reveal that the cash-only discount is present only in the lower- price segment after the financial crisis. Although cash-only transactions are more common for distressed, investor-purchased and rental properties, we find no significant cash-only price discount for these property categories except for the lower-priced segment. Overall, we demonstrate that cash-only price discounts are less contingent on the distressed status, but more so driven by the increased supply and attractiveness of lower-priced homes by investors.

The usual view is that households who purchased at the top of the market are those most at risk of foreclosure due to price declines. Our paper shows that the presence of house price appreciation, coupled with liberal lending practices, also leads to higher risk of having negative equity, the primary driver of foreclosure. Using public record data to study Southern California borrowers actually experiencing foreclosure during 2006 - 2008, we show that 41% of them extracted equity. The borrowers who extracted lost their homes despite having, on average, 14% initial equity and 5% home value appreciation during their ownership period. In addition, low loan-to-value, price increase, young age and high income level increase propensity to extract.

Using the MLS and the land registration data from Indiana, this paper identifies and explains price distortions associated with out-of-state sellers and buyers in the housing market. We find that out-of-state buyers pay 20.4% higher prices than local buyers, and the premium is fully explained by the former purchasing larger homes than the latter. On the other hand, out-of-state sellers receive a 21.2% price discount, among which 9.3% is attributable to differences in transactional characteristics, 3.2% is explained by increased motivation and weak bargaining power of out-of-state sellers, and 1.5% is due to differences in agent characteristics and behaviours. The remaining 7.2% discount varies systematically with the informational disadvantage of out-of-state sellers, and with the market condition. Our results are robust to model misspecification.

Canadian and U.S. real estate markets have compared similarly along dimensions such as inflation, mortgage interest rates, population and income growth and other measures. With respect to house prices, however, the series have moved in similar ways at some times, but then significantly diverged by the second quarter of 2007. For example, Canadian and U.S. house price indices reached essentially identical levels in 1987Q2, 1995Q1 and 2007Q2. As a consequence of the U.S. financial crisis and precipitous decline in house prices, the U.S. and Canadian indices have sharply diverged. Our paper examines whether or not the house price indices were driven by fundamentals during these time periods, or whether they diverged from fundamentals. We find that the U.S. house prices closely aligned with fundamentals until the mortgage markets crashed in 2008. We find that Canadian house prices continue to align with fundamentals. However, there have been some significant market changes between the two countries and key housing market measures indicate that Canadian markets are now moving along some paths similar to those taken by the U.S. prior to the crash.

This study examines the feasibility of constructing reliable commercial property price indices using property tax records. We employ the Clapp and Giacotto ( Journal of American Statistical Association, 87 (418), 300–306, 1992 ) assessed-value method to estimate price indices for commercial properties in Florida. The estimated Florida commercial property price index is compared to the Moody’s/REAL Commercial Property Price Index (CPPI) and to the transaction-based index (TBI) produced at MIT. Our results are promising, suggesting that this widely-available data source can be used to produce commercial property price indices for a variety of precise market locations and specific investor segments. A secondary but interesting objective of this paper is to use our rich and comprehensive database to examine the price performance of two specific subsets of properties in more detail. First, we narrow our range to focus on just the office sector for Florida. We compare price movements for the Florida office sector with the comparable CPPI. Estimates produce very similar price movements providing support to both methods. Second, we contrast the price performance of higher- and lower-valued properties and reject the hypothesis that their periodic price index levels are equal. The mean price changes of Florida commercial properties assessed at $2.5 million and above are observed to be slightly higher than for properties assessed below $2.5 million, although not statistically different. In particular, higher-valued properties had higher mean price changes relative to lower-valued properties during periods of economic expansion. This economic difference represents an important contribution toward beginning to understand the relative performance of smaller and investment-grade commercial properties.

The usual view is that households who purchased at the top of the market are those most at risk of foreclosure due to price declines, upward re-sets of adjustable rate mortgage instruments, the economic downturn, and other factors. Here we use public record data to study Southern California borrowers actually experiencing foreclosure during 20062008. We estimate property values as of the foreclosure sale date with an automated valuation model and the path of values since purchase with a zip code level house price index. Results show that about half of all borrowers had taken large amounts of equity out of the property through refinancing and/or junior lien borrowing with total cash extracted of approximately $105,000 per property, substantially exceeding the average decline in property value of $40,000 over their ownership period.

Prepayment penalties are ubiquitous in the commercial mortgage market yet reviled and highly restricted by law and regulation in the residential mortgage market. Considering the perspectives of both the borrower and the lender, we attempt a balanced cost-benefit analysis of this controversial contract feature for residential mortgage loans.We will address the following questions: What is the economic value of the prepayment penalty feature? Why is It more prevalent in the subprime than the prime market segment? Do borrowers obtain an offsetting economic benefit when they contract for a loan containing a prepayment penalty? What is the average cost of a prepayment penalty to borrowers, and how often is this cost actually incurred? In general, although we find a significant reduction in interest rates for loans containing a prepayment penalty, the expected costs outweigh the benefits by a considerable margin.

We combine loan data from distinct sources to compare and contrast multifamily mortgage lending in Canada and the U.S. After a general comparison of the multifamily housing markets in the two countries, we focus on loan pricing and non-price contract terms in the two environments. We find longer loan terms in the U.S. compared to Canada and attribute this to the greater liquidity available from a more established secondary mortgage market. We also find that while nominal rates are higher in Canada, mortgage spreads are actually lower, a result likely due to contract features that raise the cost of default for borrowers and restrict prepayments". In terms of loan performance, we found greater prepayment risk in U.S. mortgages and greater default risk in Canadian mortgages, although findings regarding default are limited by small sample size.

We combine loan data from distinct sources to compare and contrast multifamily mortgage lending in Canada and the U.S. After a general comparison of the multifamily housing markets in the two countries, we focus on loan pricing and non-price contract terms in the two environments. We find longer loan terms in the U.S. compared to Canada and attribute this to the greater liquidity available from a more established secondary mortgage market. We also find that while nominal rates are higher in Canada, mortgage spreads are actually lower, a result likely due to contract features that raise the cost of default for borrowers and restrict prepayments". In terms of loan performance, we found greater prepayment risk in U.S. mortgages and greater default risk in Canadian mortgages, although findings regarding default are limited by small sample size.

Commercial mortgage default is modeled in two stages. First, a mortgage becomes delinquent, when the borrower stops making payments. Second, the delinquency is either reinstated (payments are resumed) or the lender forecloses. The results of the empirical estimations have implications for lenders’ monitoring functions. Lenders should use the critical variables of loan-to-value ratio, debt coverage ratio and guarantee to identify expected delinquent loans. Within this pool of delinquent loans, these same characteristics can be used to predict the outcome, which could be reinstatement or foreclosure. An important contribution of this paper is to demonstrate that the loan-to-value ratio, debt coverage ratio and guarantee differ in a statistically significant way across these outcomes. Thank you to my doctoral committee members Tsur Somerville, Adlai Fisher and Stan Hamilton. Enormous thanks are also due to the anonymous lender who provided the database. Generous funding from the Real Estate Research Institute is gratefully acknowledged. Any errors are mine.