The current financial mess is fixable, and even fixable quickly; but in order to gather political support for implementation of the right fix it is important to understand, and explain to the public, why the fixes applied since Lehman went bankrupt in September have not worked. The essential point to understand is that the 'modern' securities-based system of Wall St. finance was fatally wounded when Lehman Brothers went under, although this demise has not been accepted by many academics or Wall St. bankers. Like Humpty Dumpty, this system can not be put back together again. Geithner & Co.'s attempts to resurrect it, the latest misguided effort being TALF, will fail, and cause further delay in restoration of normally functioning private credit markets. We need to recapitalize the banks with massive amounts of new private capital, so they have the capacity to hold many more of the loans they make on their own balance sheets, and bank lending can supply the funding gap opened up by the demise of the securities-based credit system..
Flaws in the System
To begin, it is important to recognize how Wall St. has transformed the bank-based credit system, which existed in the 1930's and prevailed until the mid-1990's, into the 'modern' securities-based credit system we have today. Non-bank sources currently supply more than half the credit needs of businesses and consumers. This transformation in the way credit is supplied has made it difficult for the Federal Reserve to reignite credit growth through massive expansion of the Federal Reserve balance sheet, which was the supposed 1930's style antidote. The old-style banking system, in which banks kept the loans they made on their balance sheets, would have responded quickly to Bernancke's interest rate cuts and aggressive injections of excess reserves. But banks today no longer keep most of the credits they underwrite on their own balance sheets, nor do they keep them in the form of individual loans. Instead, banks gather credits together to form asset-based or mortgage-based bonds which they then distribute or sell to pension funds, insurance companies, banks, hedge funds, and other investors worldwide. (Banks do keep some of these newly created asset-backed bonds on their own balance sheets.)
It is also important to recognize the critical and pivotal role that credit rating agencies and the writers of credit default insurance played in the functioning of this 'modern' credit system. Many of the asset- backed bonds kept by the banks were rated AAA by Moody's, S&P, and Fitch. Importantly, the AAA rating was granted by the agencies because it was supported by the added protection provided by credit default insurance ("credit default swaps", or "CDS"), the largest provider being American International Group (AIG), which was the largest insurance company in the world. Based on the impressive strength of AIG's AAA rated balance sheet and the AAA rating granted to the bonds by the rating agencies, it was widely believed that there was almost no risk that the AAA rated bonds would default. Under Basel II bank regulations, banks were required to set aside little, if any, capital reserves against holdings of AAA credits. Thus, under the umbrella of the AAA rating, banks could add earning assets and leverage to their balance sheets without the added burden of allocating more regulatory capital. This same dynamic made the AAA asset-backed bonds attractive to many insurance companies and other regulated entities, including foreign banks. This ability to hold AAA rated bonds without setting aside additional capital led to a highly levered global financial system.
This 'modern' model of writing loans and distributing them has a serious, fatal flaw. It undermines the integrity of promises. It is the classic 'agency problem' identified and studied by countless economists, where the self-centered motivations of the agents conflict with the security interests of the principals. When lenders are separated from borrowers, intermediaries (such as mortgage brokers, home appraisers, loan packagers, rating agencies, or bond salesmen, as well as the top leadership of major banks), can be enticed by lucrative commission revenues to pursue sloppy or even fraudulent lending practices that, in the past, had been constrained by the old-fashioned model, where banks kept the loans they wrote on their balance sheets, and where banks were kept in check by close monitoring by the FDIC.
Without confidence in the writers of credit default insurance and the rating agencies, the model of 'modern finance' can not function. In September, when Lehman Brothers went under and AIG was back-stopped by the government, confidence in the rating agencies and their AAA ratings collapsed. Credit default insurance prices went through the roof and the prices of the heretofore AAA bonds dropped by a corresponding amount. Hardly any sizable organizations were willing to continue writing credit default insurance.
The sharp decline in ABS bond prices and the jump in quotes for CDS immediately cut the value of bank assets and shrank, by the same amount, bank capital positions as calculated for regulatory purposes. This happened despite the fact that the banks had sold the bonds only in rare cases.
Suddenly, many banks become undercapitalized, which froze the credit markets. Writers of credit default swaps had to post massive additional amounts of 'good' collateral, i.e. government bonds or cash. In order to raise this cash, writers of CDS sold anything and everything. The result was a stock, bond, and commodity market crash.
Under current accounting practices, called 'mark to market', accountants and bank regulators insist on using the observable prices available in public and private markets to determine bank capital positions. But these 'marks' are only available in what are now very thinly-traded markets for asset backed bonds and credit default swaps.
In hindsight, it is clear that writers of credit default insurance on individual securities under-reserved and under-priced their product. They failed to take into account systemic risk--that in financial markets, the risk of default of any one bond or counterparty is increased by the defaults of others. Simply put, in financial markets risks are correlated, not uncorrelated. It is also clear that only the government is big enough to underwrite systemic risk. The government can do this either as the 'lender of last resort', which it fully understands, or, given today's 'modern' securities-based credit system, it must act as the 'credit insurer of last resort.' This is, in effect, what the government is doing when it offers 'guarantees' to various capital markets, such as commercial paper and mortgages.
ABS Pricing and the Debate Over Bank Solvency
There is no doubt that banks have many bad loans and securities on their balance sheets. But the size of the losses is in dispute because the structure of the asset backed bonds is complex, and because the value can reasonably be assessed in more than one way. Markets for the heretofore AAA bonds and the corresponding credit default swaps are thin, with wide spreads between bid and asking prices. Nevertheless, transactions occur, and they occur at levels that banks contend are unrealistically low. Based on the 'marks' currently extant, many banks are grossly undercapitalized, and in several cases, if banks were required to sell their bonds at current prices, they would be insolvent.
The banks claim that reasonable estimates of discounted cash flow make it ridiculous to sell bonds at such low recent prices or to use these prices to determine bank capital requirements. They further contend that the potential returns on these asset-backed bonds are greater than the potential returns on new loans, so why sell them?
Accountants and many bank regulators insist that the quotes are the best available and should be used to calculate bank capital levels and solvency. Others make the valid point that it is in the interest of banks to lie about the true value of their assets.
But the banks have a valid counter argument when they claim that recent low price levels reflect a sizable 'uncertainty discount'. Because the bonds (or the various bond tranches) are comprised of a multitude of individual credits or mortgages, a potential buyer of the bonds has no way of assessing value without drilling down into the details. This research is intensive, expensive and time consuming. Only a few investors do it. And even then, value must be judged in light of assumed economic conditions, by region or generally, now and over the next few years. Because of this extraordinary complexity it is not unreasonable to accept that the banks, which have been holding these assets for some time and have large staffs, may have a much better idea of the discounted present value than nearly all potential bond buyers or CDS writers.
Another factor to consider when attempting to price these bonds is that some bond 'marks' reflect forced liquidation because of margin calls or regulatory pressures, or desperate attempts to obtain scarce and prohibitively costly credit default insurance.
Nobody disputes that banks will suffer large losses. Loss estimates range from $1 trillion to more than $3 trillion depending on whether one uses estimates of discounted cash flow or relies on the 'marks' of the thinly traded markets.
The essential point to come to grips with is that, because of their complexity, it is virtually impossible to establish an indisputable value of these asset- backed bonds anytime soon, and possibly not before the contractual maturity of the instruments in question. It simply cannot be done.
Nevertheless, it is widely recognized that, if the credit markets are to regain their normal working condition, the so-called 'toxic', hard-to-value asset-backed bonds must be removed from bank balance sheets, or 'ring fenced' by some form of government guarantee. With economic recovery dependent on the reemergence of normalized credit conditions, there is an obvious urgency to making this happen.
How Not to Fix the Credit System
Bank nationalization, the Swedish model of the early 1990's, is not the answer. There are three reasons why nationalization should be avoided. First, Sweden had only 5 banks and there were few if any derivatives and credit default swaps outstanding. We have 10,000 banks and lots of credit default swaps outstanding. Second, because the value of the 'toxic assets' is disputable, it is not clear that the banks are at present insolvent and must be taken over by the government. Other ways of dealing with the problem exist that avoid the well-recognized problems of government ownership. And third, under a government takeover, many outstanding derivative contracts--what Warren Buffett has memorably termed "financial weapons of mass destruction"--might have to be commuted. This would trigger events similar to those that followed the Lehman bankruptcy and have catastrophic implications for the financial and industrial entities around the globe, and could lead to a cascade of bankruptcies.
The public/private scheme, announced February 10th as part of the Capital Assistance Program, is another less-than-optimal option. It is basically TARP 1 warmed over and obfuscated. Like TARP 1, it is designed to enable banks to remove the 'toxic' assets from their balance sheets at prices above those quoted in the current thin markets. Private investors would buy asset-backed bonds from the banks, on a leveraged basis, perhaps 10 to 1, with borrowed money lent to them by the government on a non-recourse basis at near government rates.
Even though this public/private scheme could help recapitalize the banks by paying above recently quoted prices, it is misleading (to say the least) to claim that the elevated values for the 'toxic' assets will be established by the 'private market.' Let's remember that it is government credit that will underpin the leveraged transactions. More important, the scheme is unfair to taxpayers because they would be the ones supplying the credit and would, therefore, be on the hook for the losses. The government could just as well buy the assets, or guarantee them, at the same elevated prices that the leveraged private investors might be comfortably paying. Any prospective profit that the private investors hope to make would then accrue to the taxpayers. Why concoct a subterfuge that subsidizes prospective profits for Wall St. fat cats? Who are the authorities trying to fool?
Finally, any program that attempts to avoid mortgage defaults, or requires a re-write of mortgage payment terms has serious problems, both legal and financial, and is likely to be counterproductive in re-establishing confidence.
Although re-writing debt contracts is standard practice in bankruptcy, it creates uncertainties when it is done outside standard bankruptcy proceedings.. In the case of mortgage-backed bonds, re-writes of payment terms can trigger rating downgrades and lower mortgage-backed bond prices. This will negatively impact bank balance sheets and necessitate increased collateral obligations for writers of credit default swaps, such as AIG. In other words, an attempt to directly help homeowners by cutting payment terms may increase the losses recorded by AIG, which the government has committed to make good, and make it more costly for banks to recapitalize. It will also make investors more wary of making new mortgages, and necessitate an expanded role for government owned FNM and FRE.
The recent well-intentioned attempts to prevent foreclosures by reducing payments are also patently unfair to taxpayer homeowners who keep current on their own mortgage payments but will be required to pay subsidies to others who are delinquent on their payments.
The Way Forward
The best way to reestablish confidence in the financial system is the following:
Part #1. The 'uncertainty discount' in asset backed bond pricing should be dealt with by fixing a floor under the asset side of bank balance sheets. But this should be done without buying the toxic assets from the banks (the TARP plans). Instead, it can be done by a government guarantee, or re-insurance, that limits the losses any bank can incur from further decreases in asset backed bond prices. In effect, the government would act as 'credit default insurer of last resort', or as the re-insurer of systemic, or aggregate risk. The government, with its unlimited balance sheet, is uniquely positioned to capture 100% of the market for systemic credit-default insurance.
Under this scheme, a bank would be required to 'buy' credit default insurance from the government in exchange for preferred stock which would be convertible into 10% to 15% of common equity after the recapitalization outlined in Part #2 below. The basic model is similar to that used for Citibank last year. In that agreement, CITI takes the first 10% of a $300 billion package of 'toxic' assets, and 10% of any losses below that first $30 billion. The taxpayers foot the bill for all other losses. This was a good idea, and it turned around the stock market in mid-November. But the insurance scheme was not made systemic and it did not incorporate the growing risk of losses from an extensive, severe economic contraction. In addition, it was not accompanied by a recapitalization of the bank. In other words, its scope was too limited, too timid.
Part #2. Once the government underwrites systemic credit default insurance, the asset position of the bank is stabilized. The upper limit of losses becomes firmly established. This makes it possible to raise private capital in large amounts, although it will entail dilution of current common stock shareholders. Bank stock prices have fallen to levels that already assume massive dilution, if not nationalization. In any event, concerns about
the extent of the dilution should not stand in the way of recapitalizing the banks, which is essential to restore confidence. The banks should be required to recapitalize by raising, in the private markets, enough common equity capital to satisfy FDIC regulators.
The positions of the bondholders and preferred stock holders should not be crammed down or diluted, even though, in a strict capitalistic sense, the bond and preferred holders may deserve a haircut. But given the opaque, hard-to-value nature of many of the bank assets, it is not clear that the bond and preferred stock holders deserve it. In any case, cram-downs are not necessary to perform the recapitalization. They can be disruptive to the credit-default insurance market, and disruptive to holders of the preferreds--fiduciaries such as pension funds and insurance companies.
In summary, the Part #1 government systemic reinsurance 'ring fences' the bad assets, or the bad part of the bank, and the new equity capital raised in Part #2 creates a 'new bank.' This all takes place within the legal and physical infrastructure of the legacy bank. There is no need to create and staff a new government bank or 'RTC' to hold and to manage the bad assets. There is no need for large up-front government expenditures to buy the asset-backed bonds from the banks. There is no need to underwrite hedge funds to buy toxic assets. The convertible preferreds issued to pay for the government insurance give taxpayers a stake in the recovery of bank prosperity. Concern about the risk of large losses on the systemic insurance should be calmed by the fact that pricing of the insurance will be bench-marked near current quotes. Taxpayers will be on the hook, in any case, for even greater government expenditures if action to 'ring fence' the toxic assets is not taken, and normal credit conditions are not restored.
Changes in top management and directors are probably required in many or even most cases. The existing leadership mismanaged the banks, and the common equity holders must pay the price via dilution.
One final and important note: despite support from academics and Wall St. bankers, it must be accepted that the era of 'modern' securities-based credit markets is over. (Humpty Dumpty cannot be put back together again.) Confidence in the rating agencies and the credit default swap counterparties was shattered by the Lehman bankruptcy. This destroyed the private market model of 'modern finance', and it cannot be resurrected in its old form. Buyers of asset-backed bonds have been badly burned, and like the cat that sat on the hot stove top, they will not return. The recent attempt by the government to reactivate the asset backed bond market (TALF), which employs government guarantees and government financing,, and relies on the discredited opinions of the rating agencies. is an ill-conceived imitation of the private market model of 'modern finance'.
Normal credit conditions can only be restored by returning to the former, 'old fashion' model where banks keep the loans they make. Bank capital positions must be built up sufficiently to enable them to keep many more loans on their own balance sheets. This implies that the size of the bank recapitalization, outlined above, must be very large.