Here's a Bloomberg article about how banks made money by doing interest rate swaps with Detroit, and now Detroit is sad, because like a lot of municipalities Detroit swapped its floating rate bonds to fixed to hedge the risk of rates going up, and rates went down, and now the PV of Detroit's swap liabilities is like $350mm, which is big, and that's sort of that. I'm generallyunmoved by the notion that municipalities should be able to get out of swaps that move against them for free, and while I'm sure there's some nefarious record of mis-selling and fee-inflating in here somewhere, which would justify you getting all mad at Detroit's banks, Bloomberg has not dug it up. The evidence so far is "rates went down," so whatever.
Still this a pretty interesting story. The normal posture of these swaps cases is:
- City has floating-rate bonds and swaps to fixed.
- Rates go down but city is still effectively paying high fixed rate.
- City says "WTF, why don't we stop doing this?"
- City goes to bank and says "remember that swap? never mind"
- Bank says "we'd be happy to tear up the swap, just pay us a $400 million termination fee."
- City freaks out, calls press, etc., shouts about windfalls, etc.1
But Detroit is different! Detroit, to its great credit, doesn't want to tear up its swaps. The banks do. But they're not exactly pushing it:
Wall Street firms could end the deals and call for full payment because Moody’s Investors Service last March cut unlimited general-obligation bond ratings to B2, five levels below investment grade, according to the city’s 2012 financial statement. In November, Moody’s cut the rating again, sending it down two levels to Caa1.
The cuts mean there is “significant risk in connection with the city’s ability to meet the cash demands” under the swap, according to Detroit’s financial report.
The city has been talking with holders of its swaps, the report said. [Detroit CFO Jack] Martin said no negotiations are occurring.
“I don’t think we’re going to have any problems with the counterparties wanting to get those dollars any time in the near future,” said Martin. “We believe, and they believe, it would not be in the city’s best interest, and wouldn’t be in their best interest either.”
That is: Detroit's swaps contracts give its banks the right to terminate at fair value - which, at current rates, means a ~$350 million payment from Detroit to the banks - if Detroit's credit ratings fall below Ba3. And they did. And the banks decided that getting paid $45 million a year, with a lot of credit risk but not too much complaining, was a better idea than trying to get their $350 million out now, probably failing, and looking like terrible terrible baddies.2
On a personal note, I gotta say, that significantly increases my sympathy for Detroit. They'll pay off their bets that went the wrong way; they just don't want them to get accelerated.
A thing I have never, ever understood about municipal finance is why municipalities do long-term floating-rate bonds and then swap them to fixed. American corporates essentially never do that (though of course they sometimes swap long-term fixed-rate bonds to floating), which feeds my suspicion that it's less for market-segmentation reasons and more for bamboozled-by-bankers reasons. (Because: [fixed | floating] bond + swap to [floating | fixed] may or may not be cheaper for you, all-in, than just doing a [floating | fixed] bond in the first place, but it always makes more money for your banks.) [Update: I am now persuaded it's for market-segmentation reasons. One reader writes, "The reason this ostensibly worked was what municipal finance people call 'The Muni Puzzle,' which is that the tax-exempt yield curve was always much steeper than the taxable yield curve because there are no natural buyers for tax-exempt debt out past 15 years or so." So long fixed bonds are more expensive than long floating bonds swapped to fixed.] One problem with this structure recently is that the floating rates on the swaps and bonds can mismatch, leading to horrors.
Another problem is that it makes liability management harder. If you're an investment-grade-rated Detroit and you issue $800mm of 6.3% bonds in 2006 - which basically happened, only it issued floating rate bonds and swapped them to fixed at around 6.3% - and in 2012 you're rated Caa1, you're still paying $50+ million a year on those bonds. But, even though rates are down, those bonds aren't trading at 6.3%. They're trading at, I dunno, 7.5-8%, or a dollar price of ~93 cents on the dollar.3 Effectively your "termination payment" on the whole thing is negative (by about $50 million): you can pay off $100 of debt for $93, because your deteriorating credit has made your debt worth less. Ask Greece for details.4
When your bonds float, though, that's harder: they'll trade closer to par, and much of the decline in present value will be in your swap. And you can't just go to your banks and ask them to let you out of your swaps at a discount just because your credit has moved against them.
Wait, why not? I mean, the contract presumably doesn't provide for that discount, but banks are rational creatures and they should be willing to unwind at a price that reflects current actual fair value (plus a healthy fee). The present value of Detroit's liabilities to the banks may be $350 million,5 and the contract may provide for a "fair value" unwind at $350 million, but the value of that contract to the banks sure isn't $350 million. They're owed $350 million, over 20 years, by an entity with a Caa1 rating. I suspect they're sitting on like a $100 million loss.
Mark-to-market loss, of course. Which will be reversed over time if Detroit keeps paying its $45 million a year on those swaps. And would be reversed instantly if Detroit paid the $350mm termination payment right now, but: no. For the banks, trying to terminate the swaps now would precipitate a fight over how much of their mark-to-market losses they'd have to realize immediately. The answer wouldn't be all of them, but it probably wouldn't be zero either. (Plus: they look like baddies while they negotiate.) Better, perhaps, to wait and hope and reduce that mark-to-market loss a little at a time.
For Detroit, of course, same thing: terminating the swaps now would cause them to realize all of their ($350mm) losses on rates immediately. Better to wait and hope and pay off those losses a little ($45mm) at a time. There might be an efficient trade here - Detroit pays the banks $300 million, reducing both Detroit's loss on rates and the banks' loss on credit - but, as Detroit's CFO said, "no negotiations are occurring."
1.One quick easy way to improve financial journalism: never use the word "windfall" again. Banks live in a world of probability distributions; windfalls are rare. Those swaps were hedged; that $400 million termination fee is no windfall.
2.Some facts, including the Ba3 threshold, drawn from Detroit's financials. The $45 million number comes from eyeballing the swaps disclosure on page 118, which shows $800mm of swap notional where banks are paid ~6.3% fixed and pay ~3mL+30bps floating, so call it the banks net 5.7% at today's rates. Incidentally the total negative fair value to Detroit in these financials is $439mm; that's as of June 30 and Bloomberg says it's now down to ~$350mm.
3.I'll never figure out munis but I found a Detroit 4% G.O. bond due 2017 trading at like 86-88 cents on the dollar, or ~7.5-8%. CUSIP 251093ZR. If this is not the right thing I don't care.
4.This is not necessarily an immensely practical strategy because you don't have cash lying around so you gotta refinance it.
5.Munis don't adjust for DVA don'tcha know.