外文翻译

内容需要下载文档才能查看

Recovery Ratios in the Savings and Loan Crisis: Evidence From the Resolution Trust Corporation’s Sale of Bank-Owned Real Estate

Daniel BergstresserBrandeis University

Richard Peiser

Harvard Graduate School of Design

Cityscape: A Journal of Policy Development and Research Volume 16, Number 1 2014

U.S. Department of Housing and Urban Development Office of Policy Development and Research

Cityscape319

Bergstresser and Peiser

Introduction

A persistent question during the recent economic crisis has been the appropriate pace at which to liquidate bank-owned real estate. Risks exist on both sides: rapid liquidation can force down real estate prices, but an overhang of unresolved properties can also hold down prices as potential buyers anticipate further increases in inventory coming onto the market.

This article uses data from the earlier financial crisis and examines sales of bank-owned real estate (REO, or Real Estate Owned) by the Resolution Trust Corporation (RTC) after the savings and loan (S&L) crisis of the 1980s.1 Our data cover all sales of bank-owned real estate by the RTC, which assumed control of failed institutions and liquidated assets during the 1989-through-2005 period.These rich data enable us to examine sales counts and recovery ratios by year, by property type, and by U.S. Census Bureau divisions of the country.2 The data show that recovery ratios—which we define as the ratio of the sales price to the gross loan balance amount of the asset—reached a nadir of 46 percent in 1990 and 1991. Average recovery ratios quickly stabilized to the mid-70-percent range.Single-family residential, industrial, and retail property sales enjoyed higher recovery ratios than sales of undeveloped land and sales of office buildings. Nearly one-half of the sales were in the West South Central division of the United States. Although the absence of an appropriate counterfactual leaves us careful about drawing a strong conclusion, this case study does suggest that it is possible for recovery ratios to increase rapidly after a rapid liquidation of bank-owned real estate portfolios.The RTC sales experience generally is regarded as a successful response to dealing with the fallout from the S&L crisis (Wang and Peiser, 2007). William Seidman, chairman of the RTC, made early projections of losses on bad loans taken over by the RTC that were estimated at more than $200 billion, not including interest, which could run the bill up to $500 billion (Cope, 1990). The ultimate loss to the U.S. Treasury was $161 billion. Although the RTC in its early days received considerable criticism—with particular focus on charges that the RTC was selling assets too

cheaply and too quickly—the evidence in this article suggests that recovery ratios on sales of REO properties recovered rapidly from the 1990-through-1991 nadir. The absence of an appropriately compelling counterfactual forces us to be somewhat humble concerning the strength of our con-clusions about policy, but we think it is important to rigorously document the facts about this case study: rapid sales of real estate by the RTC were followed by an initial drop in sales prices in 1990 through 1991 and a rapid recovery as more capital came to the market.

Although our sample is comprehensive in the sense that it covers every sale of REO properties by the RTC, we unfortunately lack critical information about the quality of the assets being sold by the RTC. We do know the state in which the asset is located, its property type, and the method of disposition, but we do not have detailed location data or any information about the physical

Throughout the article, we refer to this real estate as “REO.” This convention is based on the term “Other Real Estate Owned,” which the Office of the Comptroller of Currency uses for real estate that a bank has come to own by foreclosure on a loan or in satisfaction of debts owed to the bank.

1

This article follows the Census Bureau’s nine-division categorization. The West South Central division includes Arkansas, Louisiana, Oklahoma, and Texas. These states were significantly affected by the S&L crisis.

2

320Policy Briefs

Recovery Ratios in the Savings and Loan Crisis:

Evidence From the Resolution Trust Corporation’s Sale of Bank-Owned Real Estate

condition of the asset. We believe the analysis of recovery ratios presented in this article provides useful insight into the RTC’s experience and sheds light on deficiencies in how banks handled nonperforming loans (NPLs) and REO property sales during the Great Recession of December 2007 through June 2009. The implications are important because they bear directly on the speed of the recovery. The extent to which banks sit on bad real estate assets may be slowing down the speed of recovery because property sales prices may remain lower than they would otherwise be.

Background and Literature Review

The S&L crisis of the 1980s and 1990s was responsible for the failure of hundreds of thrift institu-tions that had book-value assets worth hundreds of billions of dollars.3 The Financial Institutions Reform, Recovery, and Enforcement Act (FIRREA) of 1989 created the RTC to acquire, manage, and dispose of the assets of failed institutions. The RTC existed from August 1989 through December 1995. FIRREA gave the RTC responsibility for managing and resolving all failed S&Ls previously insured by the Federal Savings and Loan Insurance Corporation (FSLIC). Congress established the RTC as a temporary federal agency to clean up the S&L crisis after the FSLIC became insolvent. One of the RTC’s objectives was to maximize the value of the disposition of the failed thrift institutions and their assets while minimizing the effect on local real estate and financial markets. Another objective was to maximize the availability and affordability of residential property for low- and moderate-income families (FDIC, 1998). The Federal Deposit Insurance Corporation (FDIC) typically dealt with ongoing franchises and emphasized the sale of the maximum amount of assets to the acquir-ing institution. The RTC, by contrast, focused on selling the assets directly to purchasers—most of whom specialized in buying pools of performing loans and NPLs and REO properties.

Several papers from the 1990s examine the disposition of assets in the context of the S&L crisis. Ely and Varaiya (1997) examined whether bidders overpaid for thrift institutions purchased from the RTC. They predicted the expected purchase price based on the number of participating bidders and the uncertainty of the thrift’s franchise value. In their sample of sales, they did not find evidence that the RTC underpriced the thrift institutions. Balbirer, Jud, and Lindahl (1992) investigated the monetary returns to stockholders who acquired thrift institutions in federally as-sisted mergers. They found that shareholders of acquiring firms earned significant positive returns, suggesting that—in contrast to the Ely and Varaiya result—some underpricing of the acquired assets may have occurred. Gosnell, Hodgins, and MacDonald (1993) also investigated whether ac-quirers benefited from significant positive returns in federally assisted mergers of thrift institutions. Although they studied sales from a slightly earlier period—1989 through 1991—than Balbirer, Jud, and Lindahl (1992), they did not find evidence of positive abnormal returns. Where they did find wealth transfers, they attributed it to the implicit guarantee of continued operation granted by the regulator to the acquirer.

Nanda, Owers, and Rogers (1997) also investigated whether purchasers of assets from the RTC ex-perienced extraordinary gains. To the contrary, Nanda, Owers, and Rogers (1997) found that most subsets of winning bidders—notably those who acquired former mutual institutions and properties

3

http://www.gao.gov/archive/1996/ai96123.pdf.

Cityscape321

Bergstresser and Peiser

in the RTC’s West category—had experienced persistently negative and abnormal returns. The only subset of transactions in which the acquirers earned significant gains was transfers of insured deposits. Nanda, Owers, and Rogers (1997) focused on the auctions of whole institutions, branch sales, and the transfer of insured deposits. In these transactions, the winning bidder acquired the assets and the liabilities of the institution or branch. Some of the sales included options from the RTC that “materially reduce the risk for the acquirers” (Nanda, Owers, and Rogers, 1997: 286). Their paper provides a good description of the RTC’s resolution process using auctions. Their focus on the sale by auction, merger, and acquisition of whole and partial institutions by the RTC differs from our article, which focuses specifically on the sales prices of REO properties by the RTC. Another difference is that they analyzed only publicly traded acquirers in RTC transactions in 1989 and 1990, while we analyze all sales of REO properties for RTC’s entire existence. In our sample, the final property dispositions occurred in 2005, which was 16 years after the RTC was established.

Nearly from the beginning of the RTC, politicians expressed concern that the RTC was selling assets too cheaply. The Economist (1991) reported on the political difficulties that the RTC faced in its early days to obtain government funding after its initial sales of failed institutions were at cents on the dollar. Lincoln, the savings and loan institution owned by Charles Keating, which had assets of $5 billion at its peak, sold for only $12 million.4 The sale of assets at very low prices caused an outcry for more careful oversight of future RTC sales.

In an examination of distressed commercial real estate assets that the FSLIC sold in the late 1980s, Curry, Blalock, and Cole (1991) determined that the average rate of recovery was 64 percent. They found that local market conditions, the difficulty of management, and disposition and write-downs before the FSLIC was declared insolvent were the primary determinants of the recovery rate.In a related strand of literature, Lea and Thygerson (1994) and Benveniste et al. (1994) developed models for maximizing asset recovery in the context of RTC-style resolution. Lea and Thygerson created a set of optimal disposition rules based on multiperiod cash flow maximization. They concluded that liquid assets and retail deposit franchises needed to be sold as quickly as possible; performing illiquid assets needed to be securitized with seller financing from the RTC, and non-performing illiquid assets needed to be sold with equity-participating loans from the RTC (Lea and Thygerson, 1994). Benveniste et al. (1994) concluded that the RTC would maximize its returns by retaining full or partial ownership of the assets for risk-sharing purposes while placing managerial control of distressed assets in the private sector.

The most comprehensive study of RTC recovery rates, Managing the Crisis: The FDIC and RTC Ex-perience, 1980–1994 (FDIC, 1998), was published by the FDIC in an inhouse analysis of its experi-ence selling the assets of the institutions it acquired from 1980 through 1994. The study addressed several of the areas we focus on in the present article. In particular, the RTC was concerned about

Charles Keating served 5 years in prison for his mismanagement of the Lincoln Savings and Loan Association. Five senators—Alan Cranston (D-Calif.), Dennis DeConcini (D-Ariz.), John Glenn (D-Ohio), John McCain (R-Ariz.), and Donald Riegle (D-Mich.)—were accused of corruption in 1989 after their intervention into an investigation of Lincoln by the Federal Home Loan Bank Board (FHLBB). The FHLBB subsequently backed off taking action against Lincoln.

4

322Policy Briefs

Recovery Ratios in the Savings and Loan Crisis:

Evidence From the Resolution Trust Corporation’s Sale of Bank-Owned Real Estate

public perception of a fire-sale mentality or “dumping” of assets from the start. As a result, the FIRREA legislation that established the RTC precluded the sale of real estate assets for less than

95 percent of market value, which was defined as appraised value. This requirement caused initial sales to be very slow, but FIRREA was amended in 1991 to lower the bar for sales to be not less than 70 percent of the appraised value (FDIC, 1998). The RTC had to dispose of all of the assets held by the institutions it acquired. These assets included not only real estate but also collateral for loans that included everything from wine cellars to bull sperm.5 In the present article, we focus on REO properties or “owned real estate” (ORE).

Although ORE sales represented a small percentage of total assets for both the FDIC and the RTC, their disposition was highly visible and attracted much public attention. The FDIC and the RTC were criticized for holding properties too long or selling below market value and adversely affecting real estate markets. (FDIC, 1998: 305)

The FDIC’s Managing the Crisis (1998) reports the sales price as a percentage of book value for sealed bid loan sales by the FDIC from 1986 through 1994 (FDIC, 1998).

Exhibit 1 shows recovery ratios ranging from 31.5 percent in 1988 to 79.5 percent in 1992 for loan sales that include both performing loans and NPLs.

The RTC, more so than the FDIC, found itself with an extraordinary volume of assets. As a result, unlike the FDIC, which up to a point was able to take the assets in, manage them for a short period, clean them up, and then sell them, the RTC generally did not have the luxury of time and would market assets without much prior due diligence. For that reason and because the assets held by the RTC were, on the whole, of a lesser quality, the FDIC was generally able to receive a better sales price. (FDIC, 1998: 331)Exhibit 1

19871988198919901991199219931994

Total/average

91,123 71,865 28,284 106,668 143,462 96,529 136,347 63,780 866,837

860,360 875,419 493,132 1,341,397 2,119,000 4,094,093 5,386,787 4,562,358 20,074,529

331,061 315,490 213,597 673,515 1,413,000 3,157,408 3,338,579 2,608,154 12,207,420

303,338 276,061 210,778 645,596 1,452,000 3,253,847 3,332,402 2,654,237 12,306,252

35.3 31.5 42.7 48.1 68.5 79.5 61.9 58.2 61.3

FDIC = Federal Deposit Insurance Corporation.

5

To be fair, it was actually a bull sperm bank. See Gravino (1993).

Cityscape323