Special thanks to the leading analyst that pointed out this example. The excerpt that follows is from Morgan Stanley's 2009 10-k.
Monoline Insurers. Monoline insurers (“Monolines”) provide credit enhancement to capital markets transactions. 2009 included losses of $231 million related to Monoline credit exposures as compared with losses of $1.7 billion in fiscal 2008 and losses of $203 million in the one month ended December 31, 2008. The current credit environment continued to affect the capacity of such financial guarantors. The Company’s direct exposure to Monolines is limited to bonds that are insured by Monolines and to derivative contracts with a Monoline as counterparty (principally MBIA Inc.). The Company’s exposure to Monolines as of December 31, 2009 consisted primarily of asset-backed securities bonds of approximately $458 million in the portfolio of the Company’s
Subsidiary Banks that are collateralized primarily by first and second lien subprime mortgages enhanced by financial guarantees, approximately $2.0 billion in insured municipal bond securities and approximately $651 million in net counterparty exposure (gross exposure of approximately $5.4 billion net of cumulative credit valuation adjustments of approximately $2.8 billion and net of hedges). Net counterparty exposure is defined as potential loss to the Company over a period of time in an event of 100% default of a Monoline, assuming zero recovery. The Company’s hedging program for Monoline risk includes the use of transactions that effectively mitigate certain market risk components of existing underlying transactions with the Monolines.-
The disclosure states that Morgan Stanley's exposure is "principally MBIA Inc." Morgan Stanley does not reveal how exposure is measured (i.e. by total loss or total par insured). It does go on to reveal $2.8B in credit reserves and nearly $2B in (CDS) hedges by way of inference. If one only considers the credit reserves and assumes that MBIA makes up $1B of that, well then either Morgan Stanley would shockingly accounts for half of MBIA's expected repurchase reserves OR - and perhaps more shockingly to most people - Morgan Stanley's MBIA credit-loss reserves would exceed MBIA's Morgan Stanley recovery bookings.
And that gives no value to the proceeds Morgan Stanley would realize from unwinding hedges against monoline exposure, which in MBIA's case would most likely exceed a quarter of the notional value hedged. However, the excerpt suggests that Morgan Stanley actually hedged its monoline exposure using derivatives on the underlying exposure, not on the insurer. In that case, benefits from unwinding hedges would not act as an addition to loss/repurchase/litigation reserve metric. This examination has also given no consideration to any actual litigation and repurchase reserves; the beauty is in it not having to do so.
This is the math that (while still undisclosed in most cases) led JP Morgan and the Canadian Banks to commutate their MBIA exposure. Both sides probably gained. We have just gone through a hypothetical scenario that, while full of assumptions, nonetheless underpins strong incentives for Morgan Stanley to come to the table with its financial guaranty counterparties.
Perhaps they already have.