The most surprising aspect of the current crisis is that the estimate of bank losses has been subject to continuous and macroscopic revisions. When the first problems arose in the subprime loans sector in 2006, the problem seemed to concern a quite modest segment of the U.S. credit market. But already in December 2007, the Economist estimated that losses stemming from mortgage loan insolvencies would sit between $200 and $300 bn. In April 2008 the IMF predicted losses of $565 bn on mortgage loans and on related securities, and $945 bn including loans and securities related to commercial real estate, consumer credit and corporate loans. Now the IMF has revised its estimate further to $1400 bn. How could it happen that bankers, central banks, international institutions and economic experts made such macroscopic mistakes in insolvency estimates? And how can it be that they are still so uncertain as to their real extent?
The Origins of Uncertainty
The uncertainty originates from the same roots of this crisis, that is from the opacity of the securitization with which banks “packaged” and then sold their credits in structured bonds, often after slicing them in different risk tranches. In this process, only roughly synthesized information was transmitted to the market concerning the underlying loan portfolio or its tranches. So there was a great loss of information that would have helped to evaluate the credit risk of those portfolios.
Since structured bonds and the derivatives written on them were massively bought by banks, insurance companies and trust funds, the uncertainty concerning their value turned into uncertainty regarding the amount of losses and toxic assets hidden in bank balances, and made it difficult or impossible for them to obtain liquidity or raise fresh capital. Indeed, extreme uncertainty generates fear, and fear generates paralysis. This is best illustrated by the case of Lehman Brothers, the large investment bank that played a central role in the securitization process. When Lehman entered distress, the primary U.K. bank Barclays was the only institution that showed interest in buying up Lehman, but for fear that its balance sheet hid more losses and toxic assets than those declared, they asked for a guarantee from the U.S. Treasury against this risk. As the Treasury refused to offer a guarantee, Barclays held back and Lehman failed. One might say that this bankruptcy, the largest in U.S. history, is the outcome of uncertainty. It cannot be ruled out that Lehman would have been solvent if only their assets and liabilities could have been properly evaluated.
The uncertainty generated by lack of transparency is also at the root of market illiquidity. Since June 2007, the market of structured bonds basically froze and even the liquidity on the money markets rarefied. The reason behind this event too is the fear generated by uncertainty: investors were afraid of buying securities that could hide more insolvent loans than expected, so if they had liquidity they preferred to hoard it. This market paralysis in turn worsened the situation of banks, making their assets illiquid and forcing them to curtail credit.
Uncertainty can even explain the swinging and ill-timed behaviour of U.S. policy makers: on September 8, the Treasury nationalized agencies such as Fannie Mae and Freddie Mac, who guarantee most of the U.S. mortage loans. The Treasury had already obtained Congress authorization in July and at that time had insisted there would be no need for intervention. On September 15, the Treasury let Lehman fail. On September 18, the Fed saved AIG, the world’s largest insurance company, by giving them an enormous loan with the option to buy 80 per cent of its shares, replace its executives and nearly eliminate its preexisting shareholders. Finally, on September 20 Henry Paulson, Secretary of the U.S. Treasury, asked Congress to allocate $700 bn (5 per cent of U.S. GDP) to the purchase of the banks’ bad assets, hopefully with adequate haircuts. Yet this policy move had already been proposed as early as April 2008 on the Financial Times by Luigi Spaventa, who had observed that there would be no way out from the crisis unless the authorities intervened to reestablish prices of structured bonds, which markets can no longer establish because of uncertainty.1
Maybe the dimensions of the crisis could have been contained if this suggestion had been put into action earlier. But even this delay was probably caused by the uncertainty as to the real proportion of the problem.
A socially harmful choice
But what can explain the behaviour that is at the root of this catastropic uncertainty, that is, the destruction of a large amount of price relevant information in the process of securitization and rating of structured bonds? The answer is that by simplifying the information transmitted to the market, banks managed to expand the market for the structured bonds that they issued: providing detailed and complex information would have kept away from the market many unsophisticated investors, who would have been at a disadvantage compared to those capable of processing this information.
Therefore, greater transparency would have forced issuers to reduce their security issuance or to accept a less liquid primary market, and this would have reduced their revenues, as well as those of rating agencies. Instead, they preferred to expand the primary market as much as possible, even at the cost of endangering the stability and liquidity of the secondary market.2
Now we know that this choice by issuers and rating companies was socially harmful: market liquidity and credit market stability have a social value that exceeds the private one, to the point that that today the U.S. is willing to sacrifice 5 per cent of its GDP to restore them. But this also indicates that the choice of opacity by issuers and rating companies should have met with far more solid and stringent regulatory constraints. We all knew that transparency is important for the operation of financial markets, but to this time few thought that it could be worth 5 per cent of the U.S. GDP and possibly more. Now that we know this, financial market regulators will have to keep it into account for the future.
1 Luigi Spaventa, “How a new Brady bond could ease the strain”, Financial Times, 11 April 2008. For a more detailed description and motivation of Spaventa’s proposal, see “Avoiding Disorderly Deleveraging”, CEPR Policy Insight No. 22, May.
2 Volpin in "Securitization, Transparency and Liquidity"