Muni CDS goes ‘bang’
The use of credit default swaps in muniland is poised to take off, a project that’s being called the “U.S. Municipal CDS Bang.” Starting Apr. 3, the terms and conditions of new muni CDS have been standardized with the stated intent of creating a useful risk-hedging product. This project is being driven not by regulators but by Markit, a private market-data vendor, and the International Swaps and Derivatives Association, a global consortium of Wall Street banks. But it’s not so clear that this is what the market needs at this time.
Although the muni CDS market is unregulated, the SEC is set to implement new rules (§§763 and 766) in the second half of 2012, according to the schedule posted at the agency. Even that may be optimistic, as a lot of the SEC’s deadlines for Dodd-Frank rule-writing have slipped. Nevertheless, it’s curious that these changes to muni CDS contracts are being introduced ahead of the SEC’s rules. In most instances, banks and dealers wait until the regulatory framework has been established before rolling out new products, which is what happened with the Volcker Rule.
Municipal CDS do not trade publicly, nor do they have publicly available pre-trade or post-trade information. All available pricing information for muni CDS comes from Markit, the privately held market-data firm in London. When I asked for details about the firm’s ownership, Markit was unable to provide them. Instead, the firm referred me to Companies House, the UK equivalent of the SEC’s Edgar filing service. From their required filings at Companies House, I discovered that the banks that provide the pricing information for muni CDS are mostly the same banks that sit on the Markit Board of Directors and the Determinations Committee of ISDA, the organization that decides if an event that occurs in the underlying reference bond should trigger a payout of the CDS.
In other words, the handful of banks that control the pricing of muni CDS are the same banks that seem to control the firm that collects these quotes (see table below for listing of banks). Moreover, these same banks seem to decide whether these products pay out in a credit event. Coincidentally, the banks – Bank of America-Merrill Lynch, Citibank, Goldman Sachs, JPMorgan and Morgan Stanley – are the same ones that dominate the cash market in municipal bonds.
These financial institutions, operating through Markit, will exert even more influence on the municipal debt markets and could potentially affect the borrowing costs of cities and states. Given the volatility that muniland went through with Meredith Whitney’s unsubstantiated predictions of massive defaults, it’s not clear why regulators will allow this product to launch without more transparency.
Muniland has recently suffered from a few massive bid-rigging schemes. Markit itself was investigated in 2009 by the Department of Justice, which sought to determine if the banks that owned Markit had unfair access to price information. The spokeswoman for Markit said that as far as the firm understands, the DOJ is continuing a broad review of the credit derivatives and related markets and Markit continues to cooperate and assist the DOJ.
Markit’s opaque business practices will roll back the gains in transparency that the municipal bond market has achieved over the last five years with the expansion of EMMA and the public availability of trade data. Everything about the expansion of municipal CDS goes in the opposite direction of these trends and obscures yet another part of the securities markets.
There is no municipal CDS market
California Treasurer Bill Lockyer, who is responsible for issuing new debt for the state, takes a lot of interest in the price of California’s municipal credit default swaps. The price of muni CDS can affect the cost of issuing new debt, since some investors use muni CDS as a pricing benchmark. But new work published by risk-management firm Kamakura suggests that there is no real market for municipal CDS and prices are generated primarily by broker-dealers posting their best estimates. If a specialty market is just a matter of a few dealers privately sending quotes back and forth, is it a real market? I’m not sure that it is.
There was a big story last June about Treasurer Lockyer rooting out some inconsistencies in muni CDS prices. He noticed that there was a sudden drop in the day-to-day price of the CDS used to insure California’s bonds. Katy Burne of Dow Jones reported in June:
California’s state treasurer is looking into what he believes were erroneous prices reported last month for credit-default swaps tied to the state’s debt.
The annual cost of protecting $10 million of California debt against default over the next five years fell by $45,000 from one day to the next last month, an extraordinarily big overnight move. Tom Dresslar, a spokesman for Treasurer Bill Lockyer, said that CMA Datavision provided the prices to Bloomberg’s fixed-income data service.
Lockyer suspects the cost of credit-default swaps, which are expressed as a percentage of the amount of debt covered, was artificially high before the adjustment.
“To the extent our prices are wrong, particularly if they are on the high side, that presents an inaccurate picture of our creditworthiness, at least in some corners,” said Dresslar in an interview Tuesday.
The sharp-eyed treasurer noticed the pricing blip, which could have been either an error — a trader’s attempt to recalibrate the CDS in one shot — or the result of a real sale rather than just best-estimate price quotes. I noticed the same thing for eight states whose muni CDS blew out in December. During that time the muni CDS for New York State leaped 17 percent in one day. This suggests that the market was very thinly traded and when a seller showed up to unload securities, not many were there to buy them. In this case the buyer has a very strong advantage and can force a deep discount to “help out” a panicked seller.
The newly published work of Kamakura confirms that muni CDS markets are very, very thinly traded. In their groundbreaking work Kamakura showed that almost no real trades happen within the municipal CDS markets, and the few trades that do happen are between broker-dealers. Kamakura found:
- The median number of non-dealer trades per day over the 77-week period among the 9 municipal reference names was 0.06.
- The highest average number of non-dealer trades per day for the full 77-week period among the 9 municipal reference names was 0.36.
“We find, unfortunately, that (in the words of Gertrude Stein) ‘there is no there there,’” Kamakura says. With a median of only 0.06 of trades per day being done by non-dealers, I would definitely say there is nothing there.
Reading the muni CDS tea leaves
I saw a strange tweet this morning that said “State CDS blew out yesterday per Bloomberg. Not sure what I missed here.” The anonymous tweeter attached the image above of graphs of credit-default swaps for 9 big states. Notice the very sharp one-day spike for every state except Ohio. Those spikes mean that those who trade muni CDS suddenly thought U.S. states were riskier, by anywhere from 2.09 percent to 17.02 percent, in one day. That is a big gap up.
Municipal CDS reference the equivalent cash bonds of the obligor. So a NY10Yr CDS references New York State general obligation bonds that mature in 10 years. CDS and cash bonds use different units of measurement but generally move proportionally to each other. So if investors no longer want New York State general obligation bonds and their price declines, one would usually see the CDS sell off too.
But municipal cash bond markets didn’t sell off yesterday. You can see in the Thomson Reuters Municipal Market Data chart below that New York State general obligations have been trading pretty steady recently. There certainly wasn’t a 17 percent drop yesterday like there was in the NY10Yr muni CDS. What’s going here?
Well-known blogger ZeroHedge published a Muni CDS Market Primer last January that provides good background on the product:
Theoretically, MCDS [municipal CDS] is equivalent to a financed purchase of a muni bond with an interest rate hedge and, therefore, the MCDS spread should track the muni bond spread. However, like in the corporate market, there could be a basis between the two spreads based on demand/supply and transaction cost considerations.
Markit US and Municipal CDS Report: Politics in the New Usual – December 2, 2011
This week saw uncharacteristic volatility in the municipal CDS market, both in single names and the MCDX, widening out dramatically on Tuesday, by 18bp, only to tighten back on Wednesday by the same amount. The tightening Wednesday was easily explained based on the prior day’s move and central bank interventions to provide dollar liquidity to the European financial system helped risky assets rally broadly. However, the widening on Tuesday caught some market participants a bit by surprise. The move appears largely technical and could be due to large trade unwinds hitting the market. More than $1.3bln traded in the Markit MCDX in the prior week according to figures from DTCC, which is significant. Movements in the municipal cash bond market were not nearly as pronounced in comparison lending further credence to the idea that the spike was related to unwinds.
Crawling in the dark through the muni CDS market
I’m beginning to think that Europe’s sovereign debt crisis might kill more than municipal credit default swaps. As the financial system trembles alongside the deliberations of the Greek government, areas of the markets that have quietly lumbered along in the dark are getting more and more attention.
Bloomberg held an excellent state and local finance conference on Wednesday where there was a brilliant session with the state treasurers of North Carolina, Delaware and New Jersey. Bloomberg Editor-in-Chief Matt Winkler grilled the trio on the use of derivatives, mainly interest rate swaps, by public officials. In the most thought-provoking back-and-forth, Winkler asked what the difference was between Alabama’s Jeffereson County and Greece. Both Greece and Jefferson County have extensive ties to financial institutions through the bonds they issued and derivatives they have either entered into or that have been written on them. In the case of Jefferson County this exposure is mainly concentrated with JP Morgan Chase, the lead underwriter for most of the bonds and derivatives written on the county. In the case of Greece it’s fairly well-known who owns its bonds but it’s practically unknown how deep the interconnections are for derivatives. As Bloomberg wrote:
The European nations are linked in a network of debts, as Bill Marsh recently illustrated in the New York Times with a beautiful piece of graphic art. The image is like a complex wiring diagram for a ticking debt bomb. Yet what it shows may be less important than what it leaves out: a largely invisible network of ties among institutions around the world, which could ultimately cause global financial chaos.
This hidden network has been created by institutions that buy and sell unregulated credit-default swaps. These are essentially insurance contracts on bonds; in the event of a default on the bond, the seller of the swap promises to pay the buyer the bond’s value.
This web of interconnections is amplified many times over because of the intense concentration of derivatives on the trading books of a handful of large, global banks. The graph below from the blog Jesse’s Cafe shows the level of concentration for U.S. commercial banks using data from the Office of the Comptroller of the Currency (OCC). Basically five U.S. banks control the whole U.S. market and embody the concentrated risk. A publication from ISDA, the trade association that represents the large, global banks in the derivatives area, said the OCC reported the notional amount of derivatives outstanding at the five largest U.S.-based dealers was $281 trillion as of June 30, 2010. That is a massive spiderweb of risk given that the U.S. annual GDP is about $14 trillion.
Let Europe kill municipal CDS
The solution to Greece’s debt crisis that Europe’s leaders announced on Thursday has market participants and commentators howling. It includes a provision that changes long-established rules for credit-default swaps mid-game. Mike Dolan, Reuters’ Investment Strategy Editor in Europe, said this:
For all the ifs and buts about the latest euro rescue agreement, one of its most profound market legacies may be to sound the death knell for sovereign credit default swaps — at least those covering richer developed economies.
I’d suggest that death knell just rang for U.S. municipal credit-default swaps (CDS), too. They’ve recently been on their last legs amid collapsing volumes, but actions in Europe just might have delivered the deathblow.
Credit-default swaps play an arcane role in financial markets. Firms allegedly buy them for protection against the default of bonds they hold in their portfolios. For example, if XYZ Investment Group owned $10 million worth of Greek government bonds that matured in 10 years (GGGB10YR:IND) and the Greek government couldn’t pay their obligations, then the seller of the CDS would step in and pay the CDS owner. Think of the CDS seller as a guarantor or insurance provider of sorts.
CDS are marketed as protection against the risk of default of the cash bond that they reference. Contrary to standing convention though, when the announcement was made that Greek bondholders would be asked to take a 50 percent haircut (or markdown) on the value of their bonds, the CDS governing group announced they would not be triggered. Their rationale was that the bond swap would not be compulsory and that CDS sellers would not need to make payouts to make up losses. Here is how a twitter user responded:
@amb5160 amb5160 the real story today is that CDS, a multi billion (trillion??) dollar asset class is GONE in one day! no more quotes on bberg. insanity
Why invest in insurance if the insurer says no payment is necessary because we have new conditions? It’s easy to understand the outrage.
I was under the impression that ISDA had not made a ruling yet on whether the haircut would be considered a default event triggering payment. If they do not it, I agree it will be a disaster for the cds market.
The weakest states are stronger than U.S. banks
The weakest states are stronger than US banks
I noticed something very interesting in some research that Markit, a data provider that tracks the credit-default swap market, released yesterday: the worst U.S. municipal credits (California, Illinois and New Jersey) are considered much stronger than all the major U.S. banks save JP Morgan. New York state is considered stronger than Mr. Dimon’s bank!
Why this is especially important in muniland is that these U.S. banks write a lot of credit-default swaps insuring the debt of these large states, which seems upside-down given that credit markets view the banks as weaker than the states they insure. This raises questions about the validity of the whole muni CDS market. I’ll dig around on this issue a little more.
Heavy political support for ending municipal-bond tax exemption
Bloomberg writes about several strong political forces in favor of ending the tax exclusion from municipal bond interest payments. I still haven’t seen a definitive cost analysis of the change though. Maybe the President’s proposal to reduce the tax exclusion on muni bonds for those earning over $200,000 is a signal to Republicans that the administration is willing to negotiate the issue. From Bloomberg:
The idea of phasing out the exemption has been raised in the past year as part of a broader discussion of ways to reduce the federal deficit. A proposal in the House of Representatives by Budget Committee Chairman Paul Ryan, a Wisconsin Republican, sought an end to the break.
A report from Obama’s 2010 fiscal commission led by Alan Simpson and Erskin Bowles, which the president embraced in December, also proposed ending the exemption, Bank of New York Mellon Corp. said in a report.
The break is the 11th biggest for taxpayers after such deductible expenses as employer-provided health coverage and home-mortgage interest, according to White House data.
Yet limiting the exemption for higher-income taxpayers may produce relatively little benefit, Harvey said in a statement released with the report. In 2008, for instance, “more than half of the 5.5 million tax returns reporting receipt of tax- exempt interest were filed by taxpayers with adjusted gross income below $100,000.”
“Maybe the President’s proposal to reduce the tax exclusion on muni bonds for those earning over $200,000 is a signal to Republicans that the administration is willing to negotiate the issue.”
Isn’t that part of the “unless you do something, this will happen because you insisted on it” part of the (real) American Jobs Act? That’s not “open to negotiation,” that’s “this will get them to do something.”
Does a downgrade cost anything?
The debt of the United States was downgraded by Standard & Poor’s several weeks ago, but the price of U.S. Treasuries have skyrocketed since then. This confuses many people because a baseline relationship in the fixed-income markets is that lower-rated, less-creditworthy bonds will be relatively cheap and investors will demand higher interest rates to compensate for additional risk.
To see this bond market truism, it’s much more instructive to look at the downgrade of the debt of New Jersey. Fitch lowered the state’s credit rating Wednesday citing heavy debt and benefit obligations. This followed downgrades by Moody’s and S&P earlier in the year. Municipal bond and credit default swap markets didn’t like this third downgrade and did what you would expect them to do: they required more yield in the case of cash bonds and more payment in the case of credit default swaps.
The graph above charts muni CDS prices for New Jersey (data supplied by Markit). You can see the move up in CDS prices began in June when Governor Christie and the state legislature made the final run to their agreement on the fiscal 2011 budget, which began on July 1. The uncertainty and contentiousness of the process must have spooked investors and dealers.
The most widely-used measure of credit risk for municipal bonds is the Thomson Reuters Municipal Market Data (MMD) AAA GO Scale. MMD’s Daniel Berger in his daily note talks about how New Jersey bonds got riskier and cheaper ahead of the downgrade and suggests that cash-bond selling started happening ahead of the Fitch downgrade.
New Jersey has approximately $31 bln of appropriation backed-debt and $2.6 bln in GO [general obligation] debt. The spread of New Jersey’s 10yr GO bonds has steadily risen this past week and closed last night at +47bps to MMD’s AAA GO scale. Last Friday this spread was +40bps.
It looks as if traders were anticipating this move by Fitch and this confirms our thesis that spreads are a leading indicator of credit. For the latest 12-month period this spread averaged +56.3bps which ranks sixth highest among the states actively monitored by MMD.
The MMD chart perfectly shows how traders and investors responded to rating downgrades and how they demanded higher interest rates starting September 22, 1010. This was the date that Moody’s put New Jersey on “negative outlook,” the advance notice of a possible downgrade. The long, flat, tabletop-like area is the period of market confusion following Meredith Whitney’s pronouncement of default doom. Markets settled down after that with a decline in New Jersey’s risk profile before turning up again ahead of Fitch’s downgrade.
The swirl of ratings and CDS
The Wall Street Journal ran an odd article yesterday about the unpredictability of sovereign credit ratings that are below the investment-grade cutoff (BB+ and lower). Check out the table from the IMF of S&P’s sovereign ratings.
The WSJ article seemed to air some highly paid bond-fund managers’ whining that ratings were not a useful signal for when they should buy and sell bonds of specific countries. The complaint was also that ratings don’t include certain data sets that are important, such as fund flows in banks, and that they don’t have the agility of credit default swaps.
The guys quoted in the WSJ might be right on the data sets that raters use and they are certainly right about CDS being more agile than ratings. But ratings are not intended to mirror market sentiment like CDS does. They are supposed to stand above market panics and routs and give a 30,000 foot view of an issuer’s credit condition.
Raters should incorporate every type of information available to them. Markets want hot, accurate information. Raters create reputations for themselves for good analysis, and reputations rise and fall on good predictive ability. Information is money, and markets want to make money.
It’s a well known phenomenon in fixed-income markets that once a bond falls into the junk or speculative category that its rating becomes murkier for predicting default. So maybe credit ratings are more useful for investment-grade bonds and CDS are more useful for junk-grade securities? If there are any academic studies on the subject that you know of, please add them in the comments.
An interesting thing about muniland is that all states are rated investment grade. California has the lowest rating for a state at A-. See it on the last line in the great chart above from Stateline. S&P has adjusted California’s rating 7 times in the last ten years, which is pretty active for rating change on a major issuer. But the CDS on California is moving every day!
Regulator wants to require “fair dealing”
Regulator wants to require fair dealing
In a far-reaching proposal, the Municipal Securities Rulemaking Board (MSRB) has asked the Securities and Exchange Commission for permission to impose new rules to protect municipalities. These rules would vastly expand the disclosures that dealer underwriters are required to give their municipal clients who issue bonds.
MSRB’s executive director Lynnette Kelly Hotchkiss said in a statement:
Dodd-Frank explicitly requires the MSRB to protect municipal entities. This gives us the ability to establish detailed requirements for underwriters and make important information more readily available to state and local governments that sell bonds.
The rules would require disclosure of “conflicts of interest” to municipalities before they enter into contracts to issue bonds. Specifically the new rules would require banks to:
- disclose all “material risks” associated with bond financings
- disclose when floating-rate securities are coupled with interest-rate swaps
- disclose potential conflicts of interest
- disclose incentives paid to recommend transactions
- disclose payments they may get from other parties in a deal
- disclose if they have derivative contracts that only pay off if the borrower defaults
The blog Dodd-Frank.com points out the simple effect of the proposed rules:
Markets hold the whip, but are they rational?
There has been a lot of discussion over the past few days about whether the United States deserves a triple-A rating. The weak and meandering attempts of the Congressional leadership and President Obama to reach a consensus on raising the debt ceiling has prompted this storm of confusion. The political theater is painful.
Most of the talk about ratings revolves around whether the level should be lowered one or more notches. But in The Telegraph today Ambrose Evans-Pritchard goes further and says it’s not really that important whether the United States retains a triple-A because the credit rating agencies don’t have the credibility to strip the rating to the world’s largest sovereign debt issuer (emphasis mine):
Yes, the US may be stripped of its AAA by Standard & Poor’s. A nice one-day story, but otherwise irrelevant. Global bond vigilantes are quite able to make their own judgement on the substantive default risk of the US. The rating agencies are out of their league on this one.
Evans-Pritchard’s statement implies that the qualitative judgments of rating agencies about default risk are less useful than the collective insight of bond-market participants. But is the market rational? The bond markets assess their view of the likelihood of default through a quantitative measures like credit-default swaps. CDS are bought and sold between institutional investors and represent a sort of wager on whether an issuer like the United States or the state of Illinois will default on its bonds. They are a kind of insurance policy because if the issuer does default then the holder of the CDS receives payment of principal from the issuer. In other words, it’s an opinion with a whip in its hand — unlike the assessments of credit rating agencies, whose raters suffer nothing if they assigned the wrong rating.
The CDS market data provider, Markit, sent over some price levels on municipal CDS today that included a price for CDS on U.S. Treasuries. I thought it might be interesting to compare the credit rating of the US (AAA/Aaa) and the corresponding credit-default swaps (in basis points) against some heavily indebted states (see chart above).
What we see is that although the federal government has over six times the amount of debt relative to what they collect in tax revenues, they receive a AAA rating and 58 basis points on 5-year CDS. This is quite low when compared to Massachusetts, which has a AA/Aa rating (two notches below AAA) and 107 basis points on their CDS.
This beneficial treatment for the debt of the U.S. government has persisted for a number of years. The difference between the very low cost of insurance for the United States and the higher costs for the states is always explained away by saying the federal government can simply print more money; it can’t default because it has an endless money supply.
-
Resident Muniland expert
-
