Monolines Disaster Myopia
August 12, 2011 - Monocle Research Department
I recently paged with interest through two Fitch reports on Credit Default Swap (CDS) spreads and subsequent default rates during the financial crisis. , The earlier of these Fitch reports examines CDS spreads (the annual cost of insuring a notional amount of debt) since 2007 for five US sectors, three with direct exposure to the property market (homebuilders, real estate investment trusts, and monolines), and two financial services sectors (banks and insurers).
After the collapse of two Bear Sterns hedge funds with sub-prime mortgage related exposures in mid 2007 the financial crisis began to gather momentum. Data show a sustained surge in CDS spreads from this time. At that time however it was the homebuilders who were perceived as being most risky, evidenced by an annualised CDS spread of 101 basis points (BPS) for the sector as a whole, implying a probability of default (PD) of 1.7percent, whereas the monolines as a group reflected a PD of just 0.5percent. By October 2008 homebuilders and the monolines had swapped their relative risk rankings within the group of five sectors – evidenced by the fact that the cost of insuring against homebuilder default using a CDS had surged to 447 bps, while insuring against monoline failure had warped to 1656 bps, an implied PD of 20.7percent.
The Fitch report brought to mind a July 2007 article in The Economist that played down the likelihood of the monolines running into trouble. The article points out the monolines’ practice of stress-testing each transaction, accepting only those with no losses under any scenarios. The article goes on to state that it would be difficult to imagine a situation in which the monolines’ capital cushions could be wiped out or even lose their AAA ratings. The reason mooted for monolines’ supposed imperviousness to credit losses is that they were only exposed to the senior tranches of collateralised debt obligations (CDOs).
Far from AAA ratings, today the continued existence of the monolines is in doubt. One of the enduring qualities of The Economist is its willingness to adopt a standpoint on issues, notably in areas such as ethics, climate change, international politics, and the dangers of asset price bubbles inter alia. But by bringing attention to monolines’ massive leverage (MBIA’s outstanding guarantees apparently amounted to 150 times capital at the time), and by downplaying this excessive leverage through the argument of sustainable stress-testing, the article displays the extreme myopia so prevalent of market participants on the brink of a financial crisis.
The Economist failed to see that, in spite of the fact that the monolines had no direct exposure to sub-prime mortgages, the fact that they were both so heavily leveraged and exposed to any sudden reversal in market liquidity left them highly vulnerable. Rising spreads on CDS contracts at the time provided a clue that providers of credit risk insurance were about to become caught up in a cycle of downgrades and capital attacks under conditions of depleted market liquidity. The fact that the monolines’ models showed no losses in all scenarios points to an extremely narrow view of what a scenario is, showing no recognition of leverage and liquidity risk.
The paradox of negligible implied default rates derived from CDS spreads on the monolines in mid 2007 is what piqued my interest in the Fitch report and brought me to recall the Economist article. The paradox lies in the role played by credit default swaps, in the waves of subsequent downgrades in the debt and structured credit markets more broadly, and as a direct cause of capital losses at monolines more specifically.
The latter role of credit default swaps in the monolines’ downfall came about as a consequence of these insurers’ strategy not to simply content themselves with relying on their AAA ratings to guarantee securities with a lower rating than their own, but to leverage their returns by insuring tranches of mortgage-backed and other collateralised debt obligations with credit default swaps, against which they were not required to hold capital. In normal states of the world credit default swaps can be seen as similar to out-of-the-money put options, in that both offer low cost and effective protection against downside risk to buyers, while generating high yields for the writer from the steady source of premium income. But the risk associated with this high yield was disguised in the tail of the return distribution. Rather than sustainable stress-testing, failure to recognise adverse scenarios that could result in increased uncertainty or unstable states of the world meant that the monolines were unable to measure and price the risks associated with writing credit default swaps.
It’s not simply a case that the financial institutions writing credit default swaps were unable accurately to price the risk of these instruments paying out. It’s equally remarkable to consider that, as the later of the Fitch reports points out, CDS spreads did not appear to provide a leading indicator for financial institution default risk during the financial crisis; spreads on corporates provided at best mixed results.
Quite aside from the increased uncertainty prevailing at that time, it is possible that the widespread practice of using implied default probabilities from CDS spreads directly contributed to the waves of downgrades experienced in the debt and structured credit markets. As default rates began to soar on sub-prime mortgages, CDS spreads on structured products backed by mortgages, CDOs, and on the monolines themselves began to ratchet up. Low levels of liquidity in the highly leveraged CDS market tend to ratchet up CDS spreads as downgrades create a self-sustaining spiral of deleveraging, forced selling, and even higher prices on CDS contracts.
Market participants today are keenly aware of the sudden and extreme deleterious effects excessive leverage and a lack of liquidity may have on the capital position. Indeed, this danger has been given explicit recognition by the Basel Committee for Banking Supervision’s new Basel III proposals, wherein a leverage ratio and two new liquidity ratios are to be phased in to a pillar 1 type treatment by 2018. However, the danger still remains that if wholly synthetic measures such as CDS spreads become institutionalised as an instrument for estimating default probabilities, episodes of destabilising procyclicality from both internal and external ratings are certain to reoccur.
 Fitch Ratings Macro Credit Research, “CDS Spreads and Default Risk: A Leading Indicator?”, May 12, 2011.
 Fitch Ratings Special Report, “CDS Spreads and Default Risk: Interpreting the Signals”, October 12, 2010.
 The Economist, “A monoline meltdown?”, July 28, 2007.