No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Multifamily Mortgage Credit Risk: Lessons From Recent History
Abstract
This article uses an innovative default model to explain increases in conventional multifamily mortgage default rates in the 1980s. Factors behind these changes are well known, but quantification of relative influences has not yet been performed. Our theoretical model has investors or borrowers defaulting if the underlying project has both negative equity and negative cash flows, a "double trigger." This leads to modeling default probabilities as a function of primary underwriting ratios, rental market conditions after loan origination, and institutional factors. A binary logit model is estimated with data on over 7,500 conventional multifamily mortgages purchased by Fannie Mae and Freddie Mac from 1983 to 1995. The results are used in simulations to explain why default rates increased in the 1980s and 1990s. We find that the increase was due to lax underwriting, declines in the tax benefits of owning real estate, and declines in rental market conditions. Default rates would have been worse had it not been for declines in interest rates.
... Many others distort prepayment patterns according to the design and level of the penalty. Furthermore, prepayment, like default, appears to be underexercised relative to optimality in an option-theoretic context, prompting further research into transaction prepayment costs such as legal costs, prepayment penalties, and constraints imposed by property value declines (Abraham and Theobald 1997;Bogdon and Follain 1996a;Follain, Huang, and Ondrich 1999;Follain, Ondrich, and Sinha 1997;Fu, LaCour-Little, and Vandell 2000;Goldberg and Capone 1998). ...
... Examples of these recent prepayment studies includeBoyer et al. (1997) andFollain, Ondrich, and Sinha (1997), who make use of FHA multifamily mortgages;Goldberg and Capone (1998), who make use of Freddie Mac and Fannie Mae cash purchases of multifamily mortgages;Elmer and Haidorfer (1997), who use RTC security prepayment data; and Fu, Lacour-Little, and Vandell(2000), who make use of data from Citicorp Mortgage. ...
This article examines the efficiency of multifamily finance in supplying capital to all segments of the multifamily rental market. Four issues of importance are identified: data availability, the efficiency of the small loan market, multifamily prepayment and default risk, and the potential role of the mortgage real estate investment trusts in multifamily finance. The findings suggest that information inadequacies in multifamily finance create possible failures in the operation of the multifamily market, providing a rationale for government involvement. Government- encouraged and supported data provision is necessary to understand, monitor, and guide the flow of multifamily credit so the multifamily market can fulfill its role of primary housing supplier for lower- income households. Four omissions from the current policy discussion also are identified: understand- ing of the capital expenditures (repositioning) loan market, consideration of equity as well as debt financing, the special nature of the low-end market, and explicit evaluation of the benefits versus the costs from securitization and governmental intervention in the allocation of multifamily credit.
... Early studies rely primarily on the foreclosure experience of major life insurance companies. However, although this experience is important, insurance companies represent less than one-fifth of the multifamily investor market and the analysis focuses on only one form of commercial mortgage default (foreclosure). 1 Recent work examines the experience of other investors in the same activity class as life insurance companies, such as Freddie Mac and Fannie Mae (Goldberg and Capone [1998]) and FHA (Follain et al. [1999]), and also focuses on restricted definitions of default. This paper examines the determinants of multifamily mortgage default for the largest class of multifamily originators and investors, banks and savings institutions, using a broad definition of default that recognizes all types of default-related events, such as loan modifications and workouts as 4 Fitch (1996) provides a practitioner analysis of multifamily and commercial loans that includes portions of the data covered by our analysis. ...
... The most recent studies of commercial mortgage default include papers by Goldberg and Capone (1998) and by Follain, Huang, and Ondrich (1999). These two papers have in common that they estimate proportional hazard models that focus on post-origination changes in metropolitan rental market conditions to explain default. ...
Option-based models of mortgage default posit that the central measure of default risk is the loan-to-value (LTV) ratio. We argue, however, that an unrecognized problem with extending the basic option model to existing multifamily and commercial mortgages is that key variables in the option model are endogenous to the loan origination and property sale process. This endogeneity implies, among other things, that no empirical relationship may be observed between default and LTV. Since lenders may require lower LTVs in order to mitigate risk, mortgages with low and moderate LTVs may be as likely to default as those with high LTVs. Mindful of this risk endogeneity and its empirical implications, we examine the default experience of 495 fixed-rate multifamily mortgage loans securitized by the Resolution Trust Corporation (RTC) and the Federal Deposit Insurance Corporation (FDIC) during the period 1991-1996. The extensive nature of the data supports multivariate analysis of default incidence in a number of respects not possible in previous studies. Consistent with our expectations, we find that LTV evidences no relationship to default incidence, while the strongest predictors of default are property characteristics, including three-digit ZIP code location and initial cash flow as reflected in the debt coverage ratio. The latter results are particularly interesting in that they dominated the influence of postorigination changes in the local economy. Copyright 2002 by the American Real Estate and Urban Economics Association..
... Restricting the sample to mortgaged properties permits inclusion of two widely used additional risk factors: the loan-to-value ratio (LTV), calculated as the unpaid principal balance of the first mortgage loan at the time of the survey divided by estimated property value, and the debt service coverage ratio (DCR), measured as the ratio of net operating income to annual principal and interest payments. 21 Goldberg and Capone (1998) argued that DCR and LTV are two of the primary determinants of multifamily default risk. Because LTV and DCR are calculated based on data collected at the time of the RFS, and they may not be applicable to mortgages originated many years previously, recent LTV and recent DCR are defined as the product of an indicator for mortgages originated during 1985 to 1991 with LTV and DCR, respectively. ...
Small multifamily properties with fewer than 50 dwelling units are important from a policy standpoint because they are more affordable and often are located in underserved areas. Small properties appear to experience greater difficulty than larger properties in securing mortgage financing. It is estimated that 41.5 percent of units in the multifamily housing stock are from small properties. Data from the 1991 Residential Finance Survey support the hypothesis that the mortgage market for multifamily properties is affected by segmentation. Information asymmetries may have contributed to domination of the small property mortgage market by portfolio lenders. Differences exist between small and large multifamily markets in the percentage of properties financed and the incidence of adjustable-rate financing. Regression analysis confirms a lower incidence of mortgage financing and higher ex ante likelihood of relational and adjustable-rate financing among small properties, after controlling for risk, location, and owner characteristics.
... But knowing which indicators are superior predictors (individually or in tandem) of default , prepayment, undermaintenance, and other behaviors of importance requires further empirical testing. The work of Vandell et al. (1993); Goldberg (1994); Mills and Lubuele (1994) ; Follain, Ondrich , and Sinha (1997); Goldberg and Capone (1998); and Abraham (1999) provide pioneering steps in this direction. ...
Indicators of the financial condition of the multifamily housing stock can potentially inform several policy issues, such as the loss of affordable rental units, multifamily developers’ access to capital, and the emerging secondary mortgage market for multifamily properties. Several rules of thumb exist for assessing financial condition. This article uses the Residential Finance Survey to investigate whether it matters, from a practical standpoint, which one is employed. Specifically, we ascertain how five measures—loan‐to‐value, debt coverage, rent‐to‐value, net operating income—to‐value, and vacancy loss ratios—relate to each other and rank properties.
We found that Pearsonian correlations among the measures varied dramatically. Factor analysis produced two factors, one corresponding to a rent‐flow measure and the other to a debt‐burden measure. Spearman rank‐order correlations revealed that with one exception, measures yielded noticeably dissimilar financial condition rankings. We conclude that single‐dimensional measures of financial condition should not be used in isolation.
... A regression of capitalization rates on mortgage contract rates is estimated using ACLI data. A similar approach is also employed by Goldberg and Capone (1998, 2002). ...
... In general, if the default probability (P(Y)) is a linear function f of a vector of explanatory variables x, where x includes loan-related and non-loan-related variables, under the logistic distribution, the default probability can be specified as: 5 See He and Liu (2002), Chiang et al. (2002) and Chow and Liu (2003). 6 See Campbell and Dietrich (1983), Vandell and Thibodeau (1985), Gardner and Mills (1989), Capozza, et al (1997), Goldberg and Capone (1998) and Archer, et al (2002). Other studies use the logit model to predict mortgage prepayment risk, for instance, LaCour-Little (1999). ...
The sharp fall of property prices after the Asian financial crisis has led many residential mortgage holders in Hong Kong to experience negative equity. Among other factors, this study looks at the impact of negative equity on the probability of default on mortgage loans, which is an important issue in view of the fact that residential mortgage lending represents a significant component of bank assets. The empirical analysis confirms the role of current loan-to-value ratio (CLTV) as a major determinant for mortgage default decisions. It also finds that the default probability is positively correlated with the level of interest rates and the unemployment rate, and negatively correlated with financial market sentiment. Under a hypothetical scenario that the maximum 70% LTV ratio guideline on residential mortgages were relaxed to 90% some time before 1997, the potential amount of default among the negative equity loans are estimated to be significantly higher than otherwise. Given the importance of the CLTV for defaults, this study lends strong support to the prudential policy of encouraging the adoption of a maximum 70% LTV ratio in residential mortgage lending.
... Using much broader data sets, Archer et al. (1999), Follain and Ondrich (1999), and Goldberg and Capone (1999) more fully describe the nature of commercial mortgage defaults over the recent recession, yet are limited to multi-family loans emanating from agency or RTC sources. In particular, Archer et al. (1999) make an initial attempt to describe regional variation in default rates, through the identification of 26 explicit geographic locations as defined by 3 digit zip codes. ...
We study whether tax considerations are an important determinant of commercial mortgage default. We also study whether large
lenders are better informed, or better at interpreting information for lending purposes, and hence have lower foreclosure
rates; whether lenders have more information on larger borrowers than smaller borrowers, and hence have lower foreclosure
rates on larger loans; and whether commercial mortgage defaults are related to debt service coverage and loan-to-values, both
initial and contemporaneous. The paper’s main findings are fourfold. First, holding all else equal, there is evidence that
tax considerations influence investors’ decisions about when to “put” assets to lenders. The results are consistent with the
argument of Constantinides (J Financ Econ 13:65–89, 1984). Second, the evidence suggests that large lenders are especially knowledgeable about commercial mortgage borrowers and commercial
property markets, in that they have lower foreclosure rates than smaller lenders. Third, on the question of whether lenders
have more information on larger borrowers than smaller borrowers, we find that larger loans have, on average, lower default
rates than smaller loans. Fourth, the findings suggest that lower default rates are associated with higher debt service coverage
ratios, both initial and contemporaneous.
KeywordsDefault–Commercial mortgages–Taxes
There are three prevailing theories of mortgage default: strategic default (driven by negative equity), cash flow default (driven by negative life events), and double-trigger default (where both negative triggers are necessary). It has been difficult to test between these theories in part because negative life events are measured with error. We address this measurement error using a comparison group of borrowers with no strategic-default motive. Our central finding is that only 6 percent of underwater defaults are caused exclusively by negative equity, an order of magnitude lower than previously thought. We then analyze the remaining defaults. We find that 70 percent are driven solely by negative life events (i.e., cash flow defaults), while 24 percent are driven by the interaction between negative life events and negative equity (i.e., double-trigger defaults). Together, the results provide a full decomposition of the three theories underlying borrower default and suggest that negative life events play a central role.
In this paper, I use logistic regression to study the relationship between walkability and mortgage default risk in multifamily housing in a pool of nearly 37,000 Fannie Mae loans. Walkability is measured with Walk Score, a widely available metric. Controls were introduced for loan terms, property characteristics, neighborhood conditions, and macroeconomics. Walkability reduced default risk but the relationship was nonlinear with thresholds. Default risk significantly increased where walkability was very low and significantly decreased where it was very high. The implication is that walkability and its possible benefits to health and the environment could be fostered by relaxing lending terms without adding default risk.
This study examines the relationship between transportation-, location-, and affordability-related sustainability features and default risk in multifamily housing. It finds that sustainability features can be used to improve the prediction of mortgage default and reduce default risk. The study uses 37,385 loans in the Fannie Mae multifamily portfolio at the end of 2011:Q3. The results suggest two implications for practice. First, certain aspects of sustainability can be fostered without increasing default risk by adjusting conventional lending standards. Second, lenders could improve their risk management practices by taking stock of sustainability features when loans originate.
We study the impact of green building on loans in the CMBS market. A hazard model shows green buildings carry 34 percent less default risk, all else equal. A matched-sample analysis gives similar results. We attribute the effect to a loan-to-value channel, where risk is lowered by a green price premium. The benefit comes at least partly from the level of green achievement, not only the label itself. Loans on buildings that were green at loan origination have slightly better terms than loans on non-green buildings. That difference is growing over time, but the effect is economically small compared to default risk. This article is protected by copyright. All rights reserved
The sharp fall in property prices following the Asian financial crisis has led many residential mortgage holders in Hong Kong to experience negative equity. At the end of September 2004, there were about 25,400 loans with a market value lower than the outstanding loan amount. The total value of these loans was HK$43 billion. The mortgage delinquency ratio reached a peak of 1.43 per cent in April 2001. While it has improved since the second half of 2001, the delinquency rate in September 2004, at 0.47 per cent, is still higher than 0.29 per cent in June 1998 when data were first collected.1 Given that residential mortgage lending represents a significant component of bank assets, how borrowers’ decisions to default are affected by the negative equity position of their mortgages is of interest to policy makers.2
This study introduces the cap rate spread as a novel metric for underwriting commercial mortgages. Cap rate spread is the difference between the cap rate and the fixed coupon rate. The spread predicts performance risk in a sample of 24,951 commercial mortgage-backed securities loans during 1993–2011. We demonstrate that the cap rate spread includes crucial information about performance risk. The results arise from the role of the cap rate spread in generating positive or negative leveraged returns to equity in situations where additional equity is required. Incorporating simplistic cap rate spread requirements in commercial underwriting is expected to reduce loan performance risk.
This article examines the factors driving the borrower's decision to terminate commercial mortgage contracts with the lender through either prepayment or default. Using loan-level data, we estimate prepayment and default functions in a proportional hazard framework with competing risks, allowing us to account for unobserved heterogeneity. Under a strict definition of mortgage default, we do not find evidence to support the existence of unobserved heterogeneity. However, when the definition of mortgage default is relaxed, we do find some evidence of two distinctive borrower groups. Our results suggest that the values of implicit put and call options drive default and prepayment actions in a nonlinear and interactive fashion. Prepayment and default risks are found to be convex in the intrinsic value of call and put options, respectively. Consistent with the joint nature of the two underlying options, high value of the put/call option is found to significantly reduce the call/put risk since the borrower forfeits both options by exercising one. Variables that proxy for cash flow and credit conditions as well as "ex post" bargaining powers are also found to have significant influence upon the borrower's mortgage termination decision. Copyright 2002 American Real Estate and Urban Economics Association.
In this paper we exploit loan level data combining foreclosure histories with information about the revenues and expenses associated with the ongoing management and eventual sale of financially distressed loans to estimate the magnitude of realized excess returns on commercial mortgages. Our findings are striking. We find that average realized excess returns on commercial mortgages are the lowest at the best times á la Stiglitz and Weiss (Am. Econ. Rev., 71:393–409, 1981). We also find that excess realized returns on commercial mortgages are low when lenders are swamped with funds (which we measure by the volume of commercial mortgage commitments) and when promised spreads are low. Copyright Springer Science+Business Media, LLC 2007
Loan-to-value ratio and debt service coverage ratios have long been viewed as the two most important quantitative measures of the default risk of commercial mortgages. Option-based models of default provide strong theoretic support for the importance of original loan-to-value ratio. The same theoretical predictions have found strong empirical support in residential single-family mortgage analyses. However, recent empirical studies of commercial mortgage default have raised questions about the role of loan-to-value ratio in assessing the riskiness of commercial mortgages. These studies generally either find no relationship or a puzzling negative relationship between loan-to-value ratio and default. This paper uses a very large database of commercial loan histories to thoroughly investigate this issue. It finds strong evidence that loan-to-value and debt service coverage ratios are endogenous to the underwriting process. Lenders react to other—unmeasured—risk factors with credit rationing and pricing. As a result, unusually low loan-to-value ratio loans appear to have above average risk in other dimensions and their default probabilities are equal to or higher than average. The results show that the pricing spread that lenders establish as part of the underwriting process serves as an excellent summary measure of the riskiness of the loan. A test of lenders’ ability to appropriately price loan-to-value risk finds that, while there is some unpriced effect of loan-to-value ratio after controlling for the lender’s pricing, introducing lender pricing into the model removes the otherwise puzzling negative loan-to-value and default relationship previously observed in the literature.
Using the 1996 Property Owners and Managers Survey (POMS), the authors (1) examine the mortgage debt characteristics of multifamily properties; (2) estimate the size and potential for growth of the multifamily mortgage debt market; and (3) examine the financial characteristics of multifamily properties, each with particular attention to differences by property size. While the authors find that the POMS suffers from severe missing value problems, their analysis indicates large differences in the likelihood that a property has a mortgage, the interest rate conditional on having a mortgage, and indicators of financial distress among properties of different size. They also find that smaller properties, with large potential for growth in multifamily mortgage lending, have much higher concentrations of low- to moderate-income tenants but that larger properties provide the bulk of units that serve these renters.
Renter mobility is a major concern for the performance of multifamily mortgages. If enough new renters are not found to replace those that move, vacancy rates can quickly escalate to where cash flows are negative and property mortgages are in jeopardy. In this study we examine how differences in renter mobility patterns by property type can affect mortgage credit risk within submarkets of an MSA. We expand the default model developed by Goldberg and Capone (2000) to use unique distributions of rental unit turnover and vacancy durations for large and small multifamily rental properties. Monte Carlo simulations then show how credit risk on multifamily mortgages is affected if owners of small properties are able to keep tenants longer than owners of larger properties can. The model can be used to explore other potential intra-MSA differences in property market dynamics.
As a consequence of the rapid growth of temporary agency employment in Germany, the debate on the remuneration of temporary agency workers has intensified recently. The study finds that the earnings gap of temporary help workers in Germany is indeed large and increased during the past decade. Decomposition reveals that the widening gap mainly is driven by changes in relative skill prices and less by differences in the workforce composition. Temps already have to accept a marked earnings decline before entering the temporary help sector. Nevertheless, after leaving the sector temporary help workers no longer have to accept a pay penalty. A recent reform set a high incentive for temporary help agencies to pay their workers according to a collective agreement. Surprisingly, the unionization of the sector could not bring thewidening earnings gap to a halt.
This paper extends existing equilibrium commercial mortgage pricing models by endogenizing negotiated workout into the usual noncooperative lending game. Workout is a feasible subgame strategy for the lender to play whenever foreclosure transaction costs exist for either party to a loan transaction. In particular, negotiated workout solutions Pareto dominate the foreclosure alternative when default occurs. To obtain our results, we embed a cooperative bargaining game within a noncooperative mortgage loan/default game. We also address the valuation "wedge" problem that occurs when foreclosure transaction costs are introduced. Through the notion of replacement game equilibrium, we find symmetric mortgage pricing solutions that eliminate the valuation wedge and thus suggest that lending will occur in commercial real estate mortgage markets even when foreclosure transaction costs exist. Copyright 1994 by Kluwer Academic Publishers
A broad consensus exists that there are significant barriers to accessing capital markets for multifamily rental housing, particularly for housing targeted at low‐and moderate‐income households. While there is a growing secondary market for multifamily mortgages, the market is in the early stages of development and remains quite small. Expanding this market requires increasing standardization of mortgages, increasing credit quality with better underwriting and credit enhancement, and educating investors as to the risks and returns of multifamily investments. In addition, dramatically improved data collection on the performance of multifamily investments and analysis on the determinants of success are needed.
A method is developed for inferring, from the consequences or outcomes of organizational decisions, an implicit choice criterion such that the organization behaves as if it were following this decision rule. The method is quantified for the case of a public bureaucracy facing discrete alternatives, and is applied in a study of the decision rules underlying freeway route selection by a state division of highways. Tests are carried out on the form of the benefit-cost calculus utilized by the bureaucracy, on the implicit evaluation of indirect benefits and costs, and on the influence of political factors on routing decisions.
This paper examines the theory of commercial mortgage default and tests it using a data set of 2,899 loan histories provided by a major multi-line insurance company. A default model is estimated which relates subsequent default incidence and timing to contemporaneous loan term, borrower, property and economic/market conditions. Maximum likelihood estimation is used to estimate a hazard function predicting conditional probability of default over time. Results confirm many expected default relationships, in particular the dominance of loan terms and property value trends over time in affecting default. The effectiveness of the model in discriminating between "good" and "bad" loans is explored. Implications for underwriting practice and credit risk diversification are noted. Finally, suggestions are made for extending these results in pricing applications. Copyright American Real Estate and Urban Economics Association.
This paper develops a pricing model for commercial real estate mortgage debt that recognizes the influence of default transaction costs on the borrower's default decision. These costs are heterogeneous across borrowers and largely un-observable to the lender/investor at the time of origination or loan purchase. A recognition of these unobservable costs can explain why borrower default decisions may differ from those predicted by "ruthless" mortgage-default pricing models. We address the determinants of default choice and timing by replacing sharp default boundaries found in the ruthless models with "fuzzy" boundaries that account for investor uncertainty with respect to evaluating borrower default decisions. To implement our model, we estimate probabilities of default as a junction of time and net equity in the property. Then, given that default occurs, loss severities are modeled based on expected property value recovery net of foreclosure costs and time until the asset is actually sold. Under reasonable parameter value choices, resulting Monte Carlo simulations produce numerical mortgage price estimates as well as component default frequency and severity levels that realistically reflect default premiums and loss levels observed in the marketplace. Copyright American Real Estate and Urban Economics Association.
In this study, we present a simple analytic framework that divides the real estate market into two markets: the market for real estate space and the market for real estate assets. After describing the size and character of flows and stocks in the U.S. real estate market, we use our framework to demonstrate the important connections between the space and asset markets. We illustrate how these real estate markets are affected by the nation's macroeconomy and financial markets, tracing out the impacts resulting from various exogenous shocks on rents, asset prices, construction and the stock of real estate. Copyright American Real Estate and Urban Economics Association.
This study examines the correlation between loan-to-value ratios and probability of default, and tests alternative cost-based default models, using data from the portfolios of FSLIC and a major regional lender. The theoretical premises for the research are synthesized from a review of the linkages between supply and demand for space, and for capital; and of the appraisal process and its reliability. Two major conclusions are drawn from statistical evidence: (1) there is no apparent correlation between value-based underwriting criteria and between default, and (2) hard cost-based criteria correlate closely to probability of default.
Commercial Mortgage Pricing with Unobservable Borrower Default Costs rMultifamily Mortgage Credit Risk: Lessons From Recent History Cityscape 113 RiddioughStrategic Default, Workout and Commer-cial Mortgage Valuation
- Timothy J Riddiough
- H E Thompson
- S B Wyatt
Riddiough, Timothy J., and H.E. Thompson. 1993. “Commercial Mortgage Pricing with Unobservable Borrower Default Costs,” Journal of the American Real Estate and Urban Economics Association 21(3):265–292. rMultifamily Mortgage Credit Risk: Lessons From Recent History Cityscape 113 Riddiough, Timothy J., and S.B. Wyatt. 1994a. “Strategic Default, Workout and Commer-cial Mortgage Valuation,” Journal of Real Estate Finance and Economics 9(1):5–22 ———. 1994b. “Wimp or Tough Guy: Sequential Default Risk and Signaling with Mortgages,” Journal of Real Estate Finance and Economics 9(3):299–322
Claims Forecasting Models for FHA Multifamily Housing Loans Paper written under contract to the U.S. Department of the Treasury through HCI, Inc
- Goldberg
- Lawrence
Goldberg, Lawrence. 1994. Claims Forecasting Models for FHA Multifamily Housing Loans. Paper written under contract to the U.S. Department of the Treasury through HCI, Inc. Reston, VA: Economic Research Laboratory, Inc
Rental Vacancy Rates and Rental Markets
- Eric S Belsky
Belsky, Eric S. 1992. “Rental Vacancy Rates and Rental Markets,” Housing Economics, February: 9–12
Smart Investments in Multifamily
- Kenneth F Danter
Danter, Kenneth F. 1996. " Smart Investments in Multifamily, " Mortgage Banking,
July: 74–79.
On Developing a Model of Mortgage Default for Multifamily Rental Housing Unpublished manuscript
- Roberto Quercia
Quercia, Roberto. 1995. " On Developing a Model of Mortgage Default for Multifamily
Rental Housing. " Unpublished manuscript. Washington, DC: The Urban Institute,
October.
A Cash Flow Model of Property Performance Unpublished manuscript
- Jesse Abraham
Abraham, Jesse. 1993a. " A Cash Flow Model of Property Performance. " Unpublished
manuscript. McLean, VA: Freddie Mac.
A Model for Assessing Commercial Mortgage Risk-Based Capital Factors
- Walter C Barnes
- S Michael Gilberto
Barnes, Walter C., and S. Michael Gilberto. 1994. " A Model for Assessing Commercial
Mortgage Risk-Based Capital Factors. " Mimeograph.