Treasury Secretary Paulson has asked Congress to authorize the government to buy up to $700 billion worth of Mortgage Backed Securities (MBS) in order to stabilize their market price, which reflects the underlying value of the mortgages that back them, and to improve the financial health of the banks and other financial institutions that own them. The proposal has raised more questions than it has answered. After a false start of forbearance in which insolvent Savings and Loan associations were artificially kept alive in the 1980s, the elimination of 747 S&Ls with efficient bank resolution tools left the American banking section in much stronger condition in the 1990s, which helped drive the rapid economy growth of the last decade. Financial institutions that have made too many mistakes should be resolved (efficiently liquidated), not bailed out. Until last week the government's measures in this area (Bear Stearns, Lehman Brothers, Merrill Lynch, Fannie Mae, and Freddie Mac) were broadly appropriate. The as-yet-to-be-fully-defined massive bailout of the MBS market is a much more questionable undertaking. I discussed the background to the current situation in "The ABCs of the Housing Crisis" and more extensively in "The U.S. Mortgage Market: the Good, the Bad and the Ugly".
The background to our current financial problems is that the United States as a whole is over-leveraged. There is too much debt and too little saving. Efficient borrowing/lending can be very beneficial, but too little saving for the whole country and for individuals during their working years reduces productive investment and the income growth it produces and increases the economy's vulnerability to shocks. As a nation our net savings rate became negative in 2005 and remained near zero until this year. This has been possible because of high savings rates in the rest of the world and the ability of foreign governments, firms and individuals to invest large amounts of these savings in the United States. Foreigners were willing to invest in the United States (largely in government and private sector debt and to a much lesser extent in equity) for only modest yields because of their confidence in the safety of U.S. investments and they were able to do so because our (previously) overvalued currency (the so called "strong dollar") created a large current account deficit (excess of imports over exports) that foreign investments in the U.S. helped finance.1
The inflow of foreign savings has kept interest rates low in the U.S. in recent years. This is a so-called "real" rather than a "monetary" phenomenon. During the period from December of 2001 until November of 2004 the Federal Reserve's overnight interest rate (fed funds rate) target was below 2% and the unusually low interest rates of this period are thought to have fed the housing price bubble. Many have claimed that these low rates were the result of excessively easy monetary policy, but this claim is somewhat contradicted by the evidence. During this period of low interest rates, the broad measure of the money supply (currency plus the public's deposits with banks-M2) grew on average at 6.2%, exactly the same average rate as between 1980 and now.2 During the shorter period between December 2002 and June 2004, when the fed funds rate was below 1.25%, M2 grew on average at 6.1% per year. However, when the fed funds rate was above 5% from November 1994 through March of 2001 the average growth rate of M2 was 6.4%.3 The Federal Reserve's adjustments in its interest rate target were largely stabilizing the growth in the money supply. A significantly higher fed funds rate during the 2002 to 2004 period would have resulted in much lower if not negative M2 growth. Interest rates were determined largely by the inflow of foreign saving. Nonetheless, interest rates were negative in real terms (less than the inflation rate) and contributed to market exuberance. A slower than trend M2 growth (and thus higher fed funds rate) might have been more appropriate. Furthermore, M2 began growing in March of 2004, and the Fed should have begun increasing its Fed funds target rate sooner than it did.
During this period of low interest rates, housing prices rose dramatically and financial market innovations exploded. Among the most important innovations were the growth and heavy reliance on securitization of debt (reselling it) and the reliance on collateral rather than expected cash flow when approving loans. In the case of mortgages, mortgage originators often sold the mortgages they issued to Fannie Mae or Freddie Mac or to private sector consolidators who packaged them together as collateral for a type of bond called a mortgage backed security (see "The U.S. Mortgage Market: the Good, the Bad and the Ugly"). These MBSs were then sold far and wide around the world to investors whose funds financed the growth in home ownership and rising home prices.
With the inevitable bursting of the real estate bubble, the estimated mortgage delinquency and foreclosure rates priced into mortgage lending rates and the privately issued MBSs proved optimistic and the value of these MBSs began to decline.4 This, and uncertainty over how far housing prices will fall and foreclosures will rise, are core features of the financial crisis we are now in. The losses in value to investors in mortgages and mortgage backed securities, though very large, should be absorbable by the widely dispersed holders of such securities without too much difficulty. However, the considerable uncertainty over the ultimate magnitude of these losses and increased concern over the status of other complex financial assets as economic activity slows have reduced the ability to trade these instruments (reduced their liquidity) resulting in a scramble among commercial and investment banks to replace their lost liquidity and to replenish their reduced capital. Not knowing who held these risky MBSs, banks also became reluctant to lend to each other. The continued flow of normal bank lending has been tightened and a credit crunch threatened.
A number of factors have been assigned some blame for the unexpectedly high losses in value and liquidity of some MBSs. Lenders were pressured into expanding sub-prime loans by government policy that wished to increase home ownership for lower-income families. However, even the reduced standards for qualifying for a mortgage loan were sometimes exceeded or ignored by lax or negligent underwriters eager to collect commissions. As long as housing prices were rising the risks of default seemed small. Over reliance on collateralized rather than cash flow lending contributed to weakened underwriting standards or the monitoring of compliance with those standards. MBSs issued by Fannie Mae and Freddie Mac are guaranteed by them (ultimately by taxpayers) while private issue MBSs were generally divided into risk tranches and rated by rating agencies.5The rating agencies grossly underestimated the risks of subprime and Alt-A mortgages, which had only a limited history in a rising market on which to base expected default rates.
Unregulated or lightly regulated capital markets have many advantages. Regulations impose restrictions and compliance costs on financial dealings. If these costs exceed the benefits to the market in the form of safer loans, the costs of raising and allocating capital are increased to the detriment of investment and growth. In unregulated markets financial engineers are limited in the instruments they can create only by the willingness of investors to buy them. But this is a serious and important limitation common to most all products and their markets. The contracts by which mortgage backed securities are packaged and sold were carefully developed to reassure potential investors that their risks were carefully controlled and understood. Underwriting standards are clearly stated and their enforcement is generally guaranteed or insured by the issuer or by third party insurers (by Fannie Mae and Freddie Mac in the case of MBSs issued by them). It is hard to imagine that better contracts could be designed by regulators, and they facilitated the flow of large amounts of money into mortgages, most of which are sound and beneficial. Yet things went wrong.
Peter Fisher, former Treasury Undersecretary for Domestic Finance, has argued that over-reliance on collateral (for a mortgage, the house itself is the collateral) rather than cash flow (an assessment of the ability of the borrower to pay) lending has contributed both to lax underwriting standards (e.g. no-doc or liars loans) and lax monitoring of the enforcement of those standards. He anticipates the reduction of securitization (collateralized borrowing) either as a result of regulation or market forces.6
Heavy reliance on collateral leaves borrowers to decide whether they can afford the loan or not. Borrowers have a knowledge advantage in making the assessment. Whether they use that knowledge wisely will depend in part on whether they expect to suffer negative consequences from wrong judgments. With rapidly rising housing prices both borrowers and lenders assigned low risks to default. In fact, many borrowers fully intended to resell homes they knew they could not afford at higher prices before being forced to default (flipping). A critical feature of keeping borrowers honest and thus of relying on borrowers' assessments of their own ability to meet their payment obligations is to insure that they have enough of their own money at risk. Thus a reasonable down payment is critical. The traditional norm was 20% but government-sponsored FHA loans require as little at 3%. When borrowers have little or no financial stake in the loan, stricter regulation and scrutiny of their income and credit worthiness is needed.
Some in Congress have proposed to give courts the authority to force a loan restructuring if, in the judgment of the court, the resulting loan could be serviced by the borrower and the lender would be better off than foreclosing. Lenders are already free to renegotiate loans under these conditions. Giving the courts such authority would add an additional risk to lenders that they would need to reflect in either higher interest rates or higher eligibility standards (as are now applied to second homes where the courts already have such authority).7The proposal would create a moral hazard among borrowers to enter into mortgages they cannot service in the expectation of a court-ordered renegotiation.
Fannie Mae and Freddie Mac ensure the timely servicing of mortgages backing the MBSs they issue. Privately issued MBSs backed by mortgages with low down payments are generally insured by third parties as well. The insurer has its money at risk and thus has a financial incentive to carefully evaluate the borrower's capacity to pay. Much of the mortgage insurance was provided by Fannie and Freddie, who gambled with taxpayers' money and lost more often than they would have if they had been risking their own money. Privately issued MBSs also accepted unrealistically low borrower standards, gambling that home prices would continue to rise. Fraud played a role as well. While carefully targeting regulations to improve transparency and capital charges against contingent liabilities (credit lines and guarantees/insurance) might be helpful, the market losses of careless lenders is the strongest guarantee of more prudent behavior in the future. Thus short-term measures to help borrowers or lenders that reduce the consequences to them of their actions run the risk of promoting such behavior again in the future (moral hazard). Successful market regulation of financial markets depends heavily on the participants bearing the consequences of their actions.
If banks have over-exposed themselves to risky mortgages and mortgage-related assets to the point of insolvency, the United States has a relatively good and efficient bank insolvency regime. The process of liquidation is orderly and minimizes any secondary disruptions to financial markets. Shareholder owners bear the losses of their mistakes providing a very important market discipline for the future. The same legal tools are not available for investment banks and thus the Federal Reserve was more constrained in its options for resolving the financial problems of Bear Stearns. While Bear Stearns shareholders should probably have been totally wiped out, the arranged sale to JPMorgan Chase, with the help of some Federal Reserve guarantees came about as close to a typical FDIC failed bank resolution as the law permitted. Whether the existing bankruptcy regime would have worked as well is an important question for investigation, but it was a risk the Federal Reserve considered too great to take at the time. The government's takeover of Fannie Mae and Freddie Mac is (hopefully) the necessary first step in correcting the government's folly in creating them in the first place and perpetuating them as semi-private entities with both private profit and social/public mandates. The subsequent hands-off approach to Lehman Brothers and Merrill Lynch were a welcome return to full market disciple. This was followed by the Federal Reserve's takeover of AIG insurance potentially wiping out the existing shareholders' equity.8AIG had extended its traditional insurance business into insuring bank credits through "credit default swaps", upon which market confidence in credit instruments has increasingly depended. No one really knows what the knock-on effects of a failure of an insurer of trillions of dollars in bank credits would be, and the Fed was not willing to risk finding out the hard way.
Secretary Paulson has now urged Congress to authorize the Treasury to buy up to $700 billion of MBSs "from any financial institution having its headquarters in the United States"9under terms and conditions determined by the Treasury. The Treasury would be required to "take into consideration means for...protecting the taxpayer."10The motivation for this staggering program seems to be to replace now-illiquid assets of unknown value on bank balance sheets with cash (or government securities) and to stabilize the secondary market for MBSs, thus restoring their liquidity. As-yet-unanswered questions include which MBSs the Treasury will buy and what they will pay for them. Presumably they wish to buy those MBSs banks have the most trouble valuing or selling. George Will called the scheme "Lemon Socialism."11If the Treasury pays more for MBSs than they are subsequently worth, they will have subsidized the investors' mistakes with taxpayer money with potentially very serious moral hazard consequences. If they pay fair value somehow determined, this will shift any remaining uncertainty and risk from the investor (banks and other financial firms) to taxpayers, which is indeed the intent. This should free up bank liquidity to continue normal lending. It would also lock in a potentially large profit for those sellers who bought the MBSs at a steep discount.
If the intent is really to restore liquidity to MBSs rather than to raise their value and thus the capital of those institutions holding them, the scheme is unnecessary because the Federal Reserve can lend banks any amount against these MBSs as collateral and has already introduced new lending facilities with longer maturities. There are many advantages to the Fed's lending against MBS collateral over the proposed Treasury purchase of them. Their valuation (or misevaluation) is much less important unless the borrowing bank fails, because the MBSs remain the property of the borrowing banks and are returned to them when the Fed loan is repaid. The proposed Treasury scheme would exchange MBSs for treasury securities and thus would not increase base money and the money supply. Fed loans can also be extended without unwanted increases in the money supply by matching such loans to particular banks with open market sales of treasury securities owned by the Fed to other banks.12 If the proposed Treasury scheme is not necessary as a means of increasing liquidity, the Treasury's purpose must really be to raise bank capital by increasing the value of the MBSs in their portfolios. This would be a taxpayer subsidy of investor mistakes with serious moral hazard consequences.
Sebastian Mallaby argues convincingly that the Treasury plan only superficially resembles the highly successful Resolution Trust Corporation (RTC) set up in 1989 to maximize the recoverable value of bad S&L loans. The RTC expedited the liquidation via purchase and assumptions or otherwise of failed S&Ls. It had no real choice over the problem loans it had to sell for which it paid the best estimate of a fair market price.13More importantly, the creation of the RTC was part of a broader strategy of resolving failed S&Ls via takeover and bankruptcy that ended several years of forbearance. It is worth remembering that because of inadequate funds in the reserves of the deposit insurers, insolvent S&Ls were artificially kept alive for several years with disastrous results that greatly increased the ultimate $150 billion or so cost to taxpayers. This weekend's Treasury proposals smell more like the banking forbearance that preceded the RTC than their resolution of which the RTC was a part. The Treasury would be choosing what MBSs to buy from still operating financial institutions. As noted above, these institutions are already able to borrow very large amounts from the Fed using these same MBSs as collateral. From the proposed law and what little the Treasury has said about how it would implement it, it is hard to see how it would contribute much beyond a huge injection of taxpayer money to save those institutions that made bad investment decisions. Market discipline could suffer for many years virtually requiring a significant increase in supervision to take its place. According to Mallaby, "Within hours of the Treasury announcement Friday, economists had proposed preferable alternatives. Their core insight is that it is better to boost the banking system by increasing its capital than by reducing its loans."14
It is appropriate for the government to intervene to restore or preserve stable financial market if doing so saves taxpayers money in the long run. However, such interventions should be well-targeted and should minimize to the extent possible creating moral hazard incentives for financial institutions to take risks with taxpayers' money. We are rapidly reaching the point where the creditworthiness of the United States government itself is coming into question as a result of the growing list of unfunded commitments it has made.
Both private sector and government debt has skyrocketed in recent years. Credit market debt stood at $2.3 trillion in 1990, $6.4 trillion in 2000 and $12.9 trillion at the end of 2007. U.S. government debt (leaving aside state and municipal debt) was $3 trillion in 1990 (51% of total national output that year), $5.8 trillion in 2000 (59% of GDP), and $9.5 trillion (69% of GDP) at the end of June this year. Twenty-eight percent of it ($2.6 trillion) is foreign owned. The Treasury and Federal Reserve have just added a potential additional trillion dollars to this debt with their various rescue actions and proposals.15Annual interest on this debt at the 3.8% currently paid for 10-year government bonds (while below the average interest on the debt) would be almost $400 billion or about 14% of the total federal budget. This seems quite manageable until the true deficit including the unfunded liabilities of Medicare/Medicaid and Social Security is calculated. According to Richard Fisher, President of the Federal Reserve Bank of Dallas, these are ten times the existing debt.16 Interest payments on the full true debt would exceed total federal tax revenue. Existing government spending promises simply cannot be met. Some mix of reductions in existing entitlements and increases in tax revenue will be required. Our goal should be to minimize the negative impact on economic growth of that mix in order to maximize the additional government revenue arising from a higher tax base.17
Warren Coats retired from the International Monetary Fund in 2003 as Assistant Director of the Monetary and Financial Systems Department, where he led technical assistance missions to central banks in over 20 countries. Prior to that he served as visiting economist to the Board of Governors of the Federal Reserve System, and to the World Bank, and was Assistant Professor of Economics at the University of Virginia from 1970 to 1975. He is currently a director of the Cayman Islands Monetary Authority, Senior Monetary Policy Advisor to the Central Bank of Iraq for BearingPoint, an IMF consultant to both the central bank of Afghanistan and the Palestine Monetary Authority, and an Asian Development Bank consultant to the National Bank of Kazakhstan on inflation targeting. In 1989 he coauthored the World Bank's World Development Report on "Financial Systems and Development." His most recent book, One Currency for Bosnia: Creating the Central Bank of Bosnia and Herzegovina, was published in November 2007. He has a BA from the University of California at Berkeley and a PhD from the University of Chicago in Economics.
- This suggests that foreign savings were pulled into the U.S., but an equally convincing case can be made that China's and Japan's desires to prevent or moderate the appreciation of their currencies against the U.S. dollar resulted in huge accumulations of dollars in the reserves of their central banks, which could only be invested in the U.S.
- Within this average, the year-on-year percent change varied considerably from zero to 13%.
- Federal Reserve Bank of St. Louis
- Payment of interest and principal on mortgages underlying MBS issued by Fannie Mae and Freddie Mac are guaranteed by them so the risk of delinquencies or defaults falls on them rather than on the investors in their MBSs.
- The role played by Fannie Mae and Freddie Mac in encouraging excessive mortgage lending are discussed in "Fannie and Freddie, More Good, Bad, and Ugly"
- Peter Fisher, "Lending's Blind Spot" The Washington Post, September 20, 2008, Page A19
- Applying this authority retroactively to established contracts would violate the constitution's ex post facto restrictions in Article I Section 9.
- The Fed has yet to clearly explain the details of this operation.
- Legislative Proposal for Treasury Authority to Purchase Mortgage-Related Assets, Sec. 2.
- George F. Will, "Bailout on Wheels," The Washington Post, September 21, 2008, Page B07
- The Fed has already run out of treasury securities that it could sell but it is already achieving the same effect via the sale of treasury securities by the Treasury (at the request of the Fed) in excess of what the Treasury needs to sell for government financing. The proceeds of these extra sales are deposited in a special frozen treasury account at the Fed. The monetary effect is the same as if the Fed had treasury securities in its portfolio and sold them in the same amount.
- Sebastian Mallaby, "A Bad Bank Rescue," The Washington Post, September 21, 2008, Page B07
- In fact, the taxpayer cost will be much smaller than this. In the latest proposal the Treasury will issue up to $700 billion in treasury securities to exchange for MBSs (an asset swap). Taxpayers lose only to the extent that the MBSs are ultimately resold for less than the Treasury paid for them.
- "Storms on the Horizon," Speech to the Commonwealth Club of California, May 28, 2008
- Bruce Bartlett, "Back to Square One," The Washington Post, September 21, 2008, Page B07