Better late than never, or two cheers for Hank Paulson

The US Secretary of the Treasury, Hank Paulson, has at last pulled the plug on the two giant GSEs (government sponsored enterprises), Fannie Mae and Freddie Mac.

That it was the big man himself who wielded the knife (or should that be the bazooka?) rather than his sidekick Jim Lockhart, Director of the Federal Housing Finance Agency (FHFA, the regulator of the GSEs), is clear from Paulson’s statement on Sunday, September 7, 2008: “I support the Director’s decision as necessary and appropriate and had advised him that conservatorship was the only form in which I would commit taxpayer money to the GSEs.”

What was decided?

(1) Fannie and Freddie have been put into administration by the Director of the FHFA.

This is called ‘conservatorship’ – which sounds better than ‘administration’. Conservatorship is not receivership or insolvency. There has been no default by either institution, and it was made clear that all debt will be serviced in full for the duration of the conservatorship. Conservatorship is more like a Chapter 11 for GSEs, under which a conservator appointed by the Director of the FHFA takes over the running of the GSEs as going concerns, which full powers. In a conservatorship, the powers of the company’s directors, officers, and shareholders are transferred to the conservator. The conservator runs the financially troubled entity until a final determination of its fate is made. This could be, but will not be, the restoration to financial health of the GSEs in their current form. The conservator could also make the recommendation to the Director of the FHFA, that either or both of the GSEs be liquidated. Again, that is most unlikely.

During the conservatorship, the stock of Fannie and Freddie will continue to trade and the stockholders retain all rights in the stock’s financial worth. The common stock and preferred stock dividends will be eliminated, but the subordinated debt interest and principal payments will continue to be made. All powers of the stockholders (including voting rights) are suspended until the conservatorship is terminated.

The CEOs will be leaving but were asked to stay on to help with the transition. Two new CEOs have been appointed.

The conservator cannot decide to liquidate the GSEs, that is, put them into receivership – the statutory process of the liquidation of a regulated entity. Only the Director of FHFA can formally make that decision. According to Paulson’s statement, there are no plans to liquidate either of the two GSEs.

Even if a GSE were to be liquidated, its charter would have to be transferred to a new entity. Only an Act of Congress can dissolve Fannie or Freddie.  Of course, the new entity could have the balance sheet of a pawn shop.

(2) The Treasury and the Fed have established a Government Sponsored Enterprise Credit Facility (GSECF).

This will provide loans to Fannie and Freddie, with a maturity between one week and one month against collateral consisting of RMBS issued by Fannie or Freddie and by advances made by the Federal Home Loan Banks. The interest rate charged will be LIBOR plus 50 bps.

The collateral will be valued and managed on behalf of the Treasury by the Federal Reserve Bank of New York “based on a range of pricing services”.

I rather like this.  First, the rate charged means that the funds are not cheap.  Libor continues to be well above the relevant OIS rate (measuring the expected average (overnight) Federal Funds rate of the relevant horizon).  Fifty basis points is also higher than the 25 basis points surcharge over the Federal Funds target rate the Fed currently charges at its primary discount window for overnight borrowing.  This makes it unlikely that this facility will provide a subsidy to the borrower, in the sense of providing funds at a risk-adjusted cost below the government’s borrowing rate.

I like it even better that this is a Treasury facility rather than a Fed facility.  The New York Fed acts as agent of the Treasury, but the FRBNY only provides its expertise, not its financial resources.  Only its reputational capital is at stake.  The liabilities and assets of the GSECF will not be on the balance sheet of the Federal Reserve System.  With a bit of luck they may even be on the balance sheet of the Treasury, rather than being hived off into some separate off-balance sheet vehicle domiciled on Mars.  Fiscal accountability requires that every exposure and all assets be on budget and on balance sheet.

Ideally, central banks should only be involved with their own resources in lending to illiquid but solvent entities.  That would exclude the two GSEs.  Lending, even against collateral, to entities subject to credit risk creates dangers to the central bank’s balance sheet, even if the terms of the loan correctly reflect the (ex-ante) joint probability of default by the borrower and by the issuer of the collateral.  The only exception would be where the collateral is free of default risk.  If they had such collateral, Fannie and Freddie would not have to go to the GSECF.  If a serious default were to materialise, the balance sheet of the central bank could be depleted to the point that it could only assure its financial viability by issuing base money in quantities that would threaten price stability.  The Treasury would have to recapitalise the central bank.  This would give the Treasury leverage over the central bank that could undermine the Fed’s independence in monetary policy.

Should a central bank, in its liquidity enhancing operations, acquire assets that are subject to credit risk, they would ideally be transferred straight to the balance sheet of the Treasury in exchange for, say, Treasury Bills or Treasury Bonds.  Alternatively, the Treasury would commit itself to indemnify the central bank promptly against all losses incurred as a result of its lender of last resort and market maker of last resort operations (this would unfortunately be an off-balance sheet commitment).  A problem is that for an effort-minimising central bank, this would not provide incentives for accurate valuation, since the assets would not stay on its balance sheet.  The Treasury would have to be able to monitor the valuation practices of the central bank, or even to participate in these valuations.

I don’t know how well equipped the New York Fed is to value RMBS if there is no market benchmark.  With house prices continuing to fall and unemployment now rising rapidly, more mortgages are bound to become impaired and previously liquid RMBS may become illiquid.  In the Primary Dealer Credit Facility and the Term Securities Lending Facility managed by the Fed of New York for the Federal Reserve System, the valuation of illiquid collateral is done by the clearing bank acting as agents for the primary dealers. That is clearly unacceptable, as it creates incentives for collusion between primary dealers and their clearers for valuing pigs ear collateral as if it were silk purse collateral.  Some combination of model-based pricing and reverse auctions will have to be used by the FRBNY (and a few Treasury observers) to value illiquid collateral offered at the GSECF.

(3) The Treasury has established a GSE Mortgage Backed Securities Purchase Program (GSEMBSPP??).

The Treasury will purchase GSE mortgage backed securities outright in the open market.  This is a radical departure from established practice in a number of ways.  First, unlike everything else done so far in the US, the UK and the euro area, this is the first time that intervention takes the form of an outright purchase rather than a sale and repurchase (repo) operation, or other collateralised loan or swap. This is a mixture of market maker of last resort operations (supporting the RMBS market) and a lender of last resort operation (supporting two specific entities: Fannie and Freddie.  As I understand it, the GSEMBSPP will purchase both MBS backed by ‘old’ mortgages and MBS backed by newly issued mortgages.  This is in sharp contrast to the UK’s Special Liquidity Scheme, which accepts only securities backed by mortgages issued before the end of 2007 (the SLS of course only does swaps of RMBs and similar ABS or covered bonds against TBs, not outright purchases).

The risk to the tax payer of this GSEMBSPP will depend on how aggressively the Treasury will lean on the prices it offers for these MBS.  Again I like the fact that this operation is off the balance sheet of the central bank.

(4) The Treasury and the FHFA have established Preferred Stock Purchase Agreements. 

These are contractual agreements between the Treasury and the ‘conserved entities’ (doesn’t this make the GSEs sound like national parks?) which ensures that both Fannie and Freddie will maintains a positive net worth.  This is a neat technical trick to ensure that Fannie and Freddie will not fall into receivership because they are balance-sheet insolvent (liabilities in excess of assets).  It is another reminder of the fact that both the old common stock holders and the old preferred stock holders will be junior to the Treasury as owner of new preferred stock.

It had to be done and it probably won’t cost the tax payer anything.

Clearly, with Freddie and Fannie about to become the victims of liquidity insolvency (and possibly of balance sheet insolvency as well), radical action was required, indeed overdue.  The two GSEs have a combined exposure, through MBS and guarantees, of over $5.5 trillion.  They account for more than 50% of the stock of residential mortgage finance and during the past year for around 80% of the flow of new residential mortgage finance.  If these two entities were to collapse, the collateral damage would go well beyond the damage to the collateral.  A possible recession would have become a certain depression.

The true fiscal-financial position of the US Treasury ought not to be affected negatively by these actions.  In addition to the RMBS liabilities and the contingent liabilities associated with the guarantees, the Treasury acquires a mass of mortgages and mortgage-backed assets.  It can afford to hold these assets to maturity.  Unless the economic downturn becomes truly ugly, the Treasury could end up making money for the tax payer.

The announcement simply makes explicit and formal an understanding that was implicit and informal: the US federal government stands behind the debt of the GSEs. It is now clear that this means backing with the full faith and credit of the US federal government, for the holders of MBS, senior debt and subordinated debt.  That, I think, is a bridge too far.

A haircut for the creditors.

The earlier implicit US federal government guarantee of the GSEs liabilities was close to, but not quite as good as, a formal, explicit guarantee.  This is reflected in the positive historical spreads of even GSE senior debt over Treasury Bills and Bonds of the same maturity.  The holders of that debt earned a positive risk premium.  They should not be surprised if some of that risk actually materialised.  At the very least, I would make a correction to any interest and principal guarantee based on the historical difference between the Fannie and Freddie yields and the Treasury yields on comparable maturities.

I understand that imposing a haircut on the creditors now is not wise, as it would effectively end the conservatorship and move the two GSEs into receivership instead.  But the conservator can try to sell off the assets of the two GSEs as soon as the markets permit.  Indeed, it could sell the assets to the Treasury through the newly created GSEMBSPP.  Once the narrowly defined social objectives of the GSEs are transferred to government-owned entities (such as Ginnie Mae), the conservator could pull the plug on the entities and distribute the receipts from the asset sales among the creditors and shareholders, in order of seniority.

Fears that foreign official institutions (holding a large chunk of the GSEs’ subordinated debt) would dump everything American should some of the GSE debt be defaulted on and thus start a run on the dollar, seem exaggerated.  Where would they go?  Even if they went, the further sharp depreciation of the US dollar would boost the US trade balance and current account, thus reducing the need for additional external financing.  The reduction in the real value of those US dollar-denominated debt instruments that continue to perform, would add insult to the injury of the formal debt default.  The additional inflation would undoubtedly be non-core inflation, so the Fed wouldn’t care.  A clear case of win-win.

Shrink them, don’t feed them

In his statement, the FHFA Director said that ” … the Enterprises will be allowed to grow their guarantee MBS books without limits and continue to purchase replacement securities for their portfolios, about $20 billion per month without capital constraints.”.  That is a huge mistake.  There should be immediate limits on the amount of new business that can be done by the GSEs.  Preferably no net new MBS business at all, starting immediately, and a gradual shrinkage of the balance sheet after a short period, either by a restricting gross RMBS issuance or by sale of the assets to private financial institutions.  The subsidised guarantees should be restricted to a limited category of mortgages for owner-occupied social housing (low income and net worth).  There is no economic or social justification for providing subsidised mortgage guarantees for middle America.  With the GSE’s no longer crowding them out of the middle-mortgage market, private financial institutions can take up the slack.  The fact that volumes will be modest for a number of years to come is the good news, from the point of view of future systemic stability.

In the longer run Fannie and Freddie can be merged and slimmed down radically.  Indeed, they could ultimately be nationalised formally and merged with Ginnie Mae, the US federal government-owned corporation within the Department of Housing and Urban Development that guarantees MBS backed by federally insured or guaranteed loans. Fannie and Freddie should get out of the securitization business completely.  There is no earthly reason why public sector entities (or private but public sector-sponsored entities) should securitise mortgages (or anything else).  The only value that the public sector can add to mortgages or mortgage-backed products is the federal government’s guarantee that interest and principal will be paid.

Fannie Mae and Freddie Mac were institutionally corrupt.  This institutional corruption was a direct consequence of their design as federally chartered but private shareholder-owned companies.  This is not a statement about the ethics, morals or actions of any individual associated with these entities.  It is a statement about their legal structure, their systemic economic role, their political function and their social perception, at home and abroad. Corruption is the abuse of public office for private gain.  The design of Fannie and Freddie deliberately obfuscated the boundaries between the private domain and the public domain, between social mission and the pursuit of private profit and shareholder value.  We ended up with companies that grew to enormous size without effective governance and with very limited accountability.  They borrowed, because of the implicit federal guarantee, on terms that were much more favourable than those implied by the riskiness of their securitisation activities, their issuance of guarantees and their exposure to a range of other complex derivative structures.  They distributed the rents generated by the implicit federal guarantee among their original shareholders, the executive leadership of the entities and the mortgage borrowers.  Their enormous wealth turned them into the most effective lobbying entities in all of DC.  They bought favours and influence in a truly bipartisan spirit.  They became an abomination. Director Lockhart’s announcement that “… all political activities — including all lobbying — will be halted immediately. We will review the charitable activities. “ is indicative of the depth of the rot in the GSEs structures.

Should housing be subsidized, and if so, how?

What is it about housing that the government should subsidies?  What are the social problems and/or market failures associated with housing?

Homelessness.  Not having a roof over your head makes you a non-person.  You cannot sleep rough and be a citizen in any meaningful sense.  Homelessness is a serious social problem.  In an affluent country like the US, it is also a national disgrace.  The key determinants of homelessness are poverty, mental illness and the breakdown of co-residential living arrangements.  The first of these is often a consequence of the last two.  Clearly, neither the deductibility of mortgage interest, nor issuance of RMBS by the two GSEs, nor the mortgage guarantees and other enhancements to home loans provided by the government and other official and quasi-official entities address the problem of homelessness.  Targeted programmes that directly focus on homelessness and its causes are bound to be both more effective and cheaper.

Foreclosures and repossessions are not a major contributor to homelessness.  Most of the time those who lose their homes because they are unable to pay their mortgages move to smaller accommodation, either rented or owner-occupied.  While this can be disruptive, painful and even traumatic, it does not make for a case for public intervention, except possibly through the provision of information aiming to reduce the incidence of people taking out mortgages that may be too large for them.

Under-spending on housing.  Do people spend too small a fraction of their incomes on residential accommodation, rented or owner-occupied?  If so, is this because of some market failure – perhaps caused by some cognitive quirks or commitment problems afflicting (part of) humanity?  If so is there a case for government intervention, through taxes, subsidies or libertarian-paternalistic ‘nudges’?

I don’t see any obvious market failures that make a strong case for either regulatory, fiscal or informational intervention here.  If the poorest members of society live in sub-standard housing, this could be either because they undervalue the benefits conveyed by their spending on accommodation on themselves, their dependents or the neighborhood.  Alternatively and, I would think, more plausibly, even though they are fully aware of all the good things brought to one and all by higher spending on housing, they are too poor to spend more on housing rather than on food, clothing, transportation or whatever.  If poverty is the problem, address poverty, don’t subsidize housing, let alone housing finance.

External benefits of home ownership.  Homeownership is no more or less meritorious than ownership of any kind of real or financial asset.  When the Republican candidate for the US Presidency has eight (or God knows how many) homes, I say: good luck to him and to Mrs. McCain, but I don’t believe this activity needs to be subsidized, regardless of whether these properties are rented out or owner-occupied 1/8th of the year and empty for the remaining 7/8th (on average).

External benefits of owner-occupancy.  Owner-occupancy, unlike buy-to-let ownership, is reputed to have serious positive externalities.  There clearly is an asymmetric information/costly monitoring problem that ensures that tenants tend to look less well after the residences they live in than do owner-occupiers.  The upkeep and appearance of an apartment or house have a direct positive amenity effect on the value of the property of the neighbors.   This provides an argument for owner-occupancy as opposed to living in rented accommodation.  Depending on the nature of the externality, a fixed subsidy (per residential unit) or one that increases with the size or the value of the residential unit could be constructed.

The owner-occupier externality does not call for making borrowing using residential property as collateral cheaper.  It argues for encouraging owner occupation regardless of whether the owner-occupier buys the property outright, inherits it, steals it or borrows against the value of the property to purchase it.  So again, mortgage guarantees, MBS securitization by GSEs, other subsidized enhancements to mortgages or MBS and indeed residential mortgage interest deductibility are extremely round-about and inefficient ways of encouraging owner occupancy.

A good case can be made that the subsidization of household borrowing collateralised against residential real estate, through public guarantees and through the deductibility of mortgage interest, has discouraged household saving and led to overinvestment in residential housing.  Both the extremely low gross rates of national, private and personal saving in the US and the common sight of great houses served by crummy infrastructure (roads, water, sewage, natural gas, power, fixed-line telephone etc.) suggest that it is time to stop subsidizing borrowing secured against residential real estate.

Let the death of Fannie and Freddie be the beginning of the end for the mortgage boondoggle.  I am hopeful, because Paulson’s statement was all but dripping with hostility to the flawed business model and the inherently conflicted nature of these two chimerae.

Maverecon: Willem Buiter

Willem Buiter's blog ran until December 2009. This blog is no longer active but it remains open as an archive.

Professor of European Political Economy, London School of Economics and Political Science; former chief economist of the EBRD, former external member of the MPC; adviser to international organisations, governments, central banks and private financial institutions.

Willem Buiter's website