Rocks, Hard Places and Pricing Models

____________________________________________________
Abbreviation Key

ABCP: Asset-backed commercial paper
BIS: Bank for International Settlements
CDO: Collateralized debt obligations
CDS: Credit default swaps
CLO: Collateralized loan obligations
GFM: Global financial meltdown
ISDA: International Swaps and Derivatives Association
LBO: Leveraged buy-out
SIFMA: Securities Industry and Financial Markets Association
SIV: Structured Investment Vehicle
_________________________________________________________

American monetary policy and financial markets have been caught between a rock (huge debt) and a hard place (fundamental economic dislocations) for a considerable time; the effects have been visible since at least 2000, when the burst of the dotcom bubble necessitated sharp interest rate cuts and the strong dollar policy was abandoned by the Bush administration. The events of 9/11 further accelerated the rate cuts, to protect the US economy from the potential threats of a deflationary spiral.

Many blame Alan Greenspan for the current credit/asset bubble, unjustly in my opinion. The Fed was not responsible for the gutting of the American industrial base and the resulting explosion of the current account deficit, nor the deep neo-conservative tax cuts that ballooned debt. Finally, the Fed was certainly not responsible for the ruinous Iraq war. All the Fed could do was to combat the effects of these deeply misguided political choices with the blunt instrument of short-term interest rates. What was even worse, even this instrument became increasingly ineffective because the explosive rise of credit derivatives (CDSs), short-circuited monetary policy and kept driving long rates lower, even as the Fed was trying to push them higher (this "conundrum" was frequently mentioned by Greenspan). But this post is not an apologia for the past Fed Chairman. He can defend his own record while making a fortune as a best-selling author.

The process of inflating the credit/asset economy started in 2001-02, accelerated in 2004-06, and the final blow-off stage was experienced in late 2006 - early 2007. Massive amounts of structured finance securities were issued (CDOs, CLOs and a dizzying array of hybrids) and the pace of LBOs reached astounding heights.

The two charts below speak volumes about the process of ramping up asset prices and borrowing against them. This was by no means solely a US phenomenon - it happened globally.

Data: SIFMA

Chart: BIS

The issuance of such enormous amounts of structured debt to finance ever-riskier mortgages and LBOs and thus pump asset prices higher would have been impossible without the massive expansion of the CDS market, which kept marking the price of risk lower and lower. The two charts below show outstanding notional amounts for CDSs and risk spreads. Clearly, it was the lowest-rated entities that benefited the most, both in the US and the emerging markets.

Data: ISDA

Note: This is an older chart that shows pre-crisis CDS spreads Chart: BIS

All of these elements resulted in a sort of malignant "virtuous cycle". Less risk signaled by the CDS market meant that more and more risky loans could be pooled together and securitized into bigger tranches of nominally AAA-A CDOs. At some point during 2004-06 the driving forces even reversed: instead of loan demand driving CDO issuance, it was the CDO issuers' demand for "product" to package that drove loan origination. The low interest rate environment created enormous speculative demand for anything that paid a spread over Treasurys or LIBOR (i.e. CDOs, etc.) and the high fees involved in securitizations made the whole business extremely lucrative. Predictably, loans of dubious provenance were originated by a vast network of fly-by-night mortgage boiler rooms. Again, this did not happen in the US only.

The exact same process occurred in the more rarefied (in image only) world of LBOs, where ghosts of junk past were suddenly alive once more. Absurd prices were bid for second tier companies not because the takeover entities could run them more efficiently (which would at least produce some economic productivity benefits), but because they could get them done with cheap money and thus earn large transaction fees. The number of private equity funds multiplied fast, including some in impossibly unlikely places, run by newly-minted "financiers" with zero experience in operating real economy businesses. As soon as cheap financing disappeared the deals immediately started falling apart - because there were no fundamental reasons to undertake them, in the first place.

Which finally brings us to the subject of the pricing models used to issue and mark-to-model all those structured finance securities. It is quite obvious that the primary variable in all pricing models is their sensitivity to credit risk, i.e. the risk of default. During the "virtuous" cycle, when credit risk goes down, such models indicated higher prices, which were used to issue structured finance securities at higher prices and with higher proportions of readily salable AAA-A merchandise. But as the cycle turned "vicious" those models started throwing out lower prices - and when the credit risks signaled by the CDSs jumped suddenly and substantially, as happened in August, those model-calculated prices moved radically down, causing havoc in the balance sheets of existing CDO holders such as banks, SIVs and hedge funds. The trouble was further enhanced by the fact that many holders were highly leveraged, i.e. they had borrowed heavily through ABCPs and prime-bank margin to buy those securities. Judging by market action quite a lot of margin came from the yen carry tactic. Live by the sword, die by the sword...

But why did participants choose to mark-to-model instead of mark-to-market? Two reasons:

(a) Because of the fragmentation in the structured finance business there were thousands of "made to order" issues that had next to zero secondary market liquidity. Usually the issuers pledged they would maintain a secondary market, but for practical purposes this was an empty promise. Even in good times the spreads between quoted bid-offer prices routinely exceeded 5 points ($50 per $1000 face) and in tiny amounts (eg $500k). We call this "trading by appointment only". Therefore, they couldn't truly mark-to-market because there simply was no active market.

(b) During the "virtuous" cycle marking-to-model served to hide the enormous embedded fees paid by real money buyers in the new issue market. For example, pension funds bought large amounts ($50 million and more) of such securities at par, a price that routinely included 5-8% underwriting fees - an atrocious percentage for AAA-A bonds when bought in size. I know of several instances where fees even exceeded 10%. By comparison, highly rated agencies and straight corporates are issued with fees of 0.5-1%.

Which brings us to what could be the "next shoe to drop". As defaults in the mortgage and junk loan sectors rise, as they are already, the models will calculate significantly lower prices, particularly for those issues that were put together with overly optimistic default assumptions. I won't be surprised to see some issues eventually model-priced at 10-20 cents on the dollar, even if their prospects for partial recoveries mean that their true values are double that. In other words, just as the models produced "garbage" prices on the upside, they have the potential to come up with "garbage" prices on the way down.

This mark-to-model snafu must have also been a factor in the Fed's decision to cut more than expected last week. Imagine a banker tearing his hair out as he saw his model inputting junk CDS spreads of 400+ bp and spewing out CLO prices in the 60s-70s, i.e. mark-to-model losses of 30-40% - and the same thing happening to his SIVs and his numerous leveraged hedge fund customers. Those CDS spreads had to come down immediately or else ... GFM was a real possibility.

As we know CDS spreads are most immediately correlated with equity prices, so the best way to have them drop fast and a lot was to engineer a stock market rally, even if artificially. Ergo, the Bernanke Fed "surprised" markets on cue, as needed.

But this is clearly a very short-term band-aid. The reality is that overextended debtors don't have the ability to service their debts and their assets are no longer worth what they once were, from real estate all the way out to LBOs, which are starting to drop like flies. The Fed may have staved off disaster for the time being*, but unfortunately nothing has changed on the ground.

_____________________________________________________________
* I can't really blame Mr. Bernanke too much. I know intimately what monetary/financial crises feel like and as a central banker you can't let them turn into immediate disasters. So you do what you have to do now, hoping to resolve the longer-term issues later.