We had said a couple of days ago that while the conservatorship of Fannie and Freddie looked to be a credit event for credit default swaps, meaning they would trigger payment, we did not see this being a financial event (presumably there would be no loss, with the Treasury assurance to prevent a negative net worth) but it could be a very ugly operational event given the large number of contracts involved and the largely manual nature of CDS processing.
We spoke too soon about the “no loss” notion. The Financial Times says there will be losses on Freddie and Fannie CDS settlement. Although the payment per contract is comparatively small (5% of face value), the amount of contracts outstanding is so large that the collective hit will be significant. Indeed, the Financial Times article suggests that this settlement process is placing stress on the CDS market, when the whole point of the Fannie and Freddie rescue was to reduce risk and prevent a systemic event.
From the Financial Times:
The default of up to $500bn of Fannie Mae and Freddie Mac credit derivatives contracts triggered by the US government’s seizure of the mortgage groups could result in billions of dollars of losses for insurance companies and banks who offered credit insurance in recent months.
The potential losses, as well as uncertainty about exactly how the derivatives contracts will be settled and unwound, is putting strains on the unregulated $62,000bn credit derivatives market, which has been a target of regulators worried about the hidden risks it could hold for the financial system….
The exact number of credit default swaps – a kind of insurance against debt default – outstanding on Fannie Mae and Freddie Mac are not known, reflecting the private nature of the sector. However, according to the latest estimates from dealers and analysts, there could up to $500bn of contracts outstanding.
Michael Hampden-Turner, credit strategist at Citigroup in London, estimates there are $200bn-$500bn of outstanding CDS and other credit derivatives referencing Fannie and Freddie.
This would make their default the biggest the market has encountered. The previous record was held by Delphi, the US carparts maker that went bankrupt in 2005 and which had about $25bn of CDS.
Currently, the recovery value of the Fannie Mae and Freddie Mac CDS is expected to be about 95 cents in the dollar, leading to a potential 5 per cent loss for insurance companies or banks who offered protection against a default. On CDS worth $200bn-$500bn, losses would come to $10bn-$25bn.
Hedge funds will also take a hit, although for a different reason: many will have to reverse profits previously booked. Reader Marshall sent us this message from Bull’s Eye Research:
While the Fannie/Freddie CDS will be cash settled, there could be some rather nasty P/L surprises for heretofore successful traders of these products. From poster “cds trader” yesterday:
I buy FRE sub CDS, 5y, at 50bps in $100mm, a while back. Nice trade, since it then widens to 250bps, where I sell it in $100mm. What’s my profit? Well, its 200bps a year for the next 5 years, discounted at the risky rate. 200bps = 2%, and lets say 5 years worth of that is worth 8% (not 10%, as we’re discounting those future cashflows).
SO…my P+L is showing up 8% of $100mm = $8mm. Great, nice trade. EXCEPT…along comes todays event, CDS triggers, but bonds are all above par. So the PAR – RECOVERY payout (ie. getting paid 100 in exchange for “defaulted” bonds) is zero, BUT all the CDS contracts stop paying the premiums.
So now I have received no payments, but my CDS trade where I was paying 50bps has gone away, AND the CDS trade where I was receiving 250bps has also gone away, so now my P+L is zero. Unfortunately, I’d already taken my $8mm P+L, so what this means to day is that I’ve just LOST $8mm, and that was from trading well apparently!!
Quite a few people will get surprised by the effects of this today I think.
13 comments
Comments are closed.
What a tangled web we weave…
Surely booking 5 years worth of profit in year 1 is a blatant breach of the accruals concept, one of the fundamental principles in accounting? The accruals concept is that income and costs should be booked in the year that they are earned. Just because the trade was set up in year 1 does not mean that the income is earned in year 1.
To illustrate with a simple example, say I agree to mow your lawn 40 times a year at $10 per time for 5 years. If it costs me $8 per time to perform the contract this gives a profit of $80 for the year, and clearly this is what I would book each year. Now let’s say I contract with someone to do the mowing for $8 per mow. I sign a contract with them to cover all 5 years. I have now “guaranteed” $2 per mow profit. No way would any auditor let me book $400 ($80 x 5) profit in year 1, yet this is exactly what the hedge funds have done.
This looks spookily similar to the teeming and lading fraud. You take out money from the current month and cover it by booking income from future months so no one can see that the money is missing. If you continue to extract money, you need to take more and more out of future months (a sort of doubling up) until the whole thing implodes because it has got to large. So not only do we have hedge funds operating at 20 to 1 leverage through debt, we have them operating at 5 to 1 leverage through the booking of forward profits, so the total leverage is 100 to 1! Look out for some spectacular unwinds ….
Jeremy
“Currently, the recovery value of the Fannie Mae and Freddie Mac CDS is expected to be about 95 cents in the dollar, leading to a potential 5 per cent loss for insurance companies or banks who offered protection against a default. On CDS worth $200bn-$500bn, losses would come to $10bn-$25bn.”
the FT is off-base. recovery is expected to be 100+ as all deliverable bonds trade above par.
you have no basis to claim that the CDS will be settled in cash, that’s for the parties to decide and they will only do that if cash is econoically equal to physical settlement. The buyer of protection is obligated to deliver a bond to get par from the seller, how much he pays for it is irrelevant to the contract and given the current state of the universe, he will pay par+, not 95
Jeremy – that’s mark to market. The two CDSs have a market value based on the rates, and you’re supposed to mark them to market, which produces the profit. If the trader took all the profit between the two rates upfront, I imagine (but I have a good imagination and these are not my specialty so I could be making a blatant error) he was being unrealistic about the mark to market, because the possibility of a credit event was greater than 0, and this would have diminished the value of one and increased the value of the other.
The way traders calculate their pnl is by taking net present value of their position and comparing it to yesterday’s npv. So there’s nothing wrong with accounting for all future flows from this trade.
I just don’t see why he is so irritated – surely he must have been aware of this risk on this position. Or did he think he was making risk-free profit?
Does this low payout high number stress illustrate the bigger problems that will occur with a larger payout issue – or if the government had not stepped in this could have caused a major wipeout
whoa. that hedgie's p&l computation is mind boggling! no wonder the financial system is in trouble.
tell me, if he were to buy, say 5yr FRE BONDS rather than the CDS's. Is he booking all the coupon payments for the next 5 years, upfront– but to his slight credit–present valued?!
Ditto to Jeremy's thoughts. It just sounds like the foundation for a huge ponzi scheme! why wouldn't he do this a million times over in a year and why even bother waiting for a spread margin of 2%? surely if he is manufacturing profits like that, it's in his best interest to do it as quickly as possible before anyone catches on to the scheme.
The FT story does seem to be off base. I’ve got a research report published yestereday by Michael Hampden-Turner in front of me, and it says that index-implied recovery values are 97%. Furthermore, he mentions that the cheapest to deliver bonds have been rallying for obvious reasons, which will surely push recoveries up. I’ve no idea where the FT is getting its figure from.
I just want to understand the actual mechanics of this process and whether it has any bearing on money supply or capital calculations.
Ok, say the purchaser of a $100mm CDS was Joe Pension fund who held $100mm in FRE bonds, matched maturity.
Joe Pension now takes his FRE bonds out of hock, repoed somewhere, and presents it and that more valuable piece of paper to Dick iBank who wrote the contract. Dick iBank takes the bonds, repos them at the Fed for cash to give that plus 5% to Joe Pension? Is that what is going to happen?
And if so, then what does Joe Pension do with the cash? Buys Treasuries? Buys FRE/FNM debt? And if so, what does this mean for Dick iBank’s capital calculation? Does he suddenly need more capital as a result of the haircutted repo and newly acquired bond position?
cdstrader is correct … but the credit hedge funds are well aware of this risk & don't trade that way (unless they have to). Maybe the market makers will get stuck with it, but they're more likely to be flat than directional trading.
Rather than entering into a new offsetting contract to produce a net MTM, you'd unwind or novate the original contract.
This means you'd get your MTM paid out in cash on unwind date, and you wouldn't face the jump to default risk on your P&L.
Leave it to the FT! It is SUCH a better paper than the WSJ and NYT it isn’t even funny.
I imagine this is the incredibly attractive and extraordinarily comprehensive Ms. Van Duyn who authored this?
My, my, she is really on top of things. She’s been on this since the beginning. Her videos are regularly at ft.com. sharp dresser too!
I’m reminded of the quote a few years back by the editor in chief of Asahi Shinbun, the most important financial newspaper in Tokyo.
He said THE major problem facing capitalism is that the derivatives market was operating in a world completely separate from the “real” economy.
When you think of 480 TRILLION in notional derivatives value and the Clinton administration SCROUNGING around for a pathetic 200 million for childhood vaccinations and this begins to come into focus.
Looks like the derivatives world is starting to “RECOUPLE” with the real economic world.
Matt Dubuque
FRE/FNM recovery is priced near par based on CTD. Think FT is confusing what dealers charging as fee to unwind trades before the auction / settlement. Dealers were charging about 5bps at first but now around 2bps.
The trader that has to reverse he gains is not a common occurrence on desks. Any reasonable credit trader for the last year at least has been closing trades via tear-ups/unwinds and not offsetting trades due to c/p risk.
In the end, aside from docs/back office will be no big deal unless LEH files.
to clarify cds trader’s comment on his pnl: say you are flat as he were- the pv of your trades are not only discounted at the risky rate, but also the likelihood of the cash streams continuing into the future (the risky curve). as cds spreads widen, the probably of default increases, the possibility of future cash flows decreases and your MTM, in his case, falls appropriately.