In his article Clearing the Mortgage Market Through Principal Reduction:  A bad Bank for Housing (RTC 2.0), Adam J. Levitin describes the real estate/mortgage market as it exists today and the need for clearing it before the economy can return to normal.  MND summarized the first part of the paper last week in 'Housing Market Must Swallow Bitter Pill To Address Negative Equity'. 

Levitin said that the present policy toward housing market problems is a combination of "do nothing, affordability, and voluntary principal reduction" strategies.  While he presents a strong case for involuntary principal reduction of underwater mortgages he deems it unlikely in the current political climate.  Thus the only likely change would be to negotiate a quasi-voluntary program in which lenders would reduce principal after strong regulatory or litigation pressure. 

In the second portion of his paper Levitin describes the transactional framework under which he sees this quasi-voluntary program established.  He recognizes the formidable political obstacles it would encounter but says that small bore solutions will do little to fix the housing market.

It must be acknowledged that clearing the housing market means recognizing losses on underwater mortgages and that there are only three places to put them; the government, financial institutions, or investors in mortgages-backed securities (MBS).  There is significant overlap in these categories; the government guarantees obligations of financial institutions and MBS investors and many financial institutions that originated, securitized, and service mortgages are portfolio lenders and/or MBS investors.

Levitin lays out parameters for principal reduction to be effective including that it have scope sufficient to achieve a macro-economic impact, that borrowers be treated uniformly (he suggests abandoning the NPV test), the program should apply to all underwater loans on owner-occupied houses regardless of default status to eliminate moral hazard arguments, and principal reduction should be enhanced with other features such as a cash-for-keys, deed-for-lease, and/or shared appreciation options.

A "bad bank" is an entity created specifically for acquiring and restructuring troubled assets, leaving the performing assets in the "good" bank.  Large banks usually have a separate accounting system that functions as a bad bank and the Home Owners Loan Corporation (HOLC) and the Resolution Trust Corporation were bad banks created by the government to deal with problem loans in the Great Depression and the savings and loan crisis respectively.  Something similar to a bad bank is also often used in bankruptcy reorganizations and was present in the Asian financial crisis and the 1990s European economic downturn.

Levitin suggests that the basic mechanism for market clearing through a quasi-voluntary program would involve pooling underwater mortgages in a bad bank he nicknames RTC2.  It would restructure and resecuritize mortgages via a transparent, standardized restructuring formula (e.g. write all mortgages down to a specified LTV and restructure them to 15, 20, or 30 year fully amortized, fully prepayable, fixed-rate obligations, ideally coupled with intensive borrower outreach.) 

Transferring troubled mortgages to a single entity would address several problems simultaneously. 

• Reunite fractured ownership and eliminate second lien issues.
• Remove contractual limitations on modifications.
• Enable consistent standards for the treatment of homeowners and modifications.
• Allow for a standard and more liquid resecuritization of the restructured mortgages with clearer risks for investors.
• Relieves banks of legacy issues and liability overhang in terms of unrecognized credit losses, the hassle and financial and reputational costs of managing the loans, and litigation risk.
• Gives U.S. financial institutions a fresh start post-crisis.

How the loans would actually be transferred to the RTC2.0 would depend on their ownership. For those held in banks or by the GSEs the transfer would be a simple sale; the only obstacle being whether the financial institutions were willing to recognize the loss.  Second liens, few of which are securitized, present some complications, matching and valuation problems which can be overcome, and the fact that many second liens are held by smaller institutions more immune to litigation or regulatory leverage.  Whether these could be included in a quasi voluntary program is questionable.

The big problem with portfolio loans is that large-scale principal write-downs will significantly decapitalize many major institutions and the throw the bill for the GSE portfolios on the taxpayers.  However, the largest financial institutions have a book equity that greatly exceeds their market capitalization indicating the market believes that banks are carrying assets at inflated values or failing to recognize liabilities.  If principal is reduced to resemble market values then it will also help narrow the book-market gap by both reducing inflated book values and increasing market value.  Indeed wide-scale principal reductions would be a rising tide that could lift all housing prices and increase the value of the written down mortgages.

Securitized loans present different challenges.  The GSE loans can only be removed from pools under specified circumstances, all of which require the GSE to buy the loan at outstanding face value.  Partial prepayment of the loans by the GSEs would get around many of the legal restrictions without saddling them with the liquidity burden of repurchasing underwater mortgages at face value from securitization pools.  Such prepayment would address negative equity but the GSEs still could not restructure the loan nor reunite first and second lies.  These could, however, be undertaken through other means.

Private-label MBS (PLS) also cannot generally be removed from their securitization trusts.   Settlements provide an exception so underwater mortgages could be extracted from PLS as part of a litigation settlement with PLS trustees.

The central negotiation point will be transfer pricing.  At what price will the bad bank acquire the bad loans?  Levitin notes issues that would likely arise in determining price.  First, the transfers would have to be negotiated in bulk, raising some valuation issues but ultimately the question will be how much loss recognition the institutions can afford.

Second, a negotiated solution should cover entities that were part of the origination and securitization process even if they no longer own or service loans.  Those that do not contribute mortgages to RTC2 can still contribute cash.

Third, price may depend on ownership.  For whole loans it is simply a matter of a "haircut" for the bank, albeit complicated by loss recognition especially in regard to second liens.  For securitized loans investor rights must also be considered.  One can imagine a basic transfer pricing schedule like that used for modifications in the federal-state servicing settlement.

There is also the related issue of financing RTC2 which would acquire mortgages for a combination of (1) litigation releases/permission to do future business, (2) cash, and (3) its debt and equity.  RTC2 would require tremendous liquidity to acquire the underwater mortgages and the duration of its liquidity needs would depend on the time needed to restructure and resecuritize the mortgages.  Doing this on a rolling basis would reduce liquidity needs, but they would still be enormous.

There are two realistic sources of liquidity - financial institutions or the government and it is likely that RTC2's needs would outstrip anything that financial institutions could provide.  Ideally, however, a two-tiered liquidity structure could be employed with the Federal Reserve providing a senior liquidity tranche and financial institutions providing a junior one. 

The equity ownership of the RTC2 is the first loss position on the mortgage restructuring.  If that restructuring is convincing and value enhancing then it could be a call option on the U.S. housing market and could be given to financial institutions as part of the transfer pricing.

Once loans are restructured the RTC2 could retain them and collect the mortgage payments but it might be preferable to sell the loans to provide liquidity to the financial institutions.  This could be done by securitizing the portfolio as the original RTC did but this presents some challenges.  The RMBS market is "moribund" and investors might be especially caution of RTC2 loans, as a new product and one that looks quite different from a pool of conventional mortgages.  The product might have to be issued at a discounted price or with some type of government or third-party guarantee.

Levitin concludes by admitting that technical questions could bedevil the design of a bad bank for housing, but the feasibility of his proposal is ultimately not a matter of technical design but rather a question of will.