A little more than a year ago, precisely at the time people were scrambling to protect the value of their municipal bonds, hedging to protect the value of munis became untenable.
In more placid times, brokers, dealers, underwriters, and anyone else sitting on an inventory of municipal paper had an array of tools at their disposal to cocoon it from the capricious jerks, sputters, and starts of the bond market.
The irony of the credit crisis was that the very turbulence that sent people looking for shelters for their municipal bonds also destroyed the security of the shelters.
Hedging is an integral facet of the notoriously illiquid muni bond market.
Imagine a broker holding, for example, $1 million of single-A 10-year munis for a month.
A lot can happen in a month, so if the broker wants to reduce the risk that he will lose money if the bonds decline in value while he is holding them, the broker might employ products in the taxable interest rate market as a hedge. Most of these hedges are derivatives of either Treasuries or the London Interbank Offered Rate.
For instance, the broker could buy a one-month put option, which is the right but not the obligation to sell at a predetermined strike price, a 10-year Treasury.
Whatever losses his municipals sustained would be offset by comparable gains on the Treasury derivative. If established correctly, a 5% decline in the dollar value of his municipal holdings would be met by a 5% increase in the dollar value of the derivative.
As long as Treasuries and munis move in the same direction by the same magnitude, the hedge works.
For most of this decade, it did. The weekly changes in the 10-year single-A muni yield based on the Municipal Market Data scale mirrored the weekly changes in the 10-year Treasury yield at a correlation better than 0.6 over more than 85% of the 10-week periods from 2000 to 2006.
Correlation measures the magnitude by which two sets of numbers move in tandem. A correlation of 1 is perfect mimicry, zero is randomness, and a negative correlation means they move in opposite directions.
Correlation is the ballast of a good hedge. The whole idea is to complement all losses with similar gains, and all gains with similar losses. A high correlation ensures the comparability of the losses on the municipals and the gains on the hedge.
Michael Bouscaren, a municipal analyst with Thomson Reuters, said the primary risk in hedging is “basis risk,” or the chance that the fluctuations in hedge instrument and the underlying bonds do not match.
Measuring basis risk is not science, Bouscaren said. It requires predicting whether relationships between assets will remain stable, he said, which is a technical and subjective judgment.
When basis risk strikes, the hedge does not cover all the losses on the underlying bonds. In some cases, such as when a positive correlation turns negative, it can even tack on additional losses.
And that is what has happened. With the municipal market last year plagued by the wholesale downgrading of the bond insurers, the Lehman bankruptcy, and severe illiquidity, the previously dependable consonance between Treasuries and munis dissolved into dissonance.
The daily changes in municipals and Treasuries bore no resemblance to each other. In the fourth quarter, according to the Federal Reserve, more than $560 billion in funds seeking a sheltered perch flooded the Treasury note market seeking a safe haven.
The 10-year Treasury shed an astounding 160 basis points in the fourth quarter of 2008. Anyone who tried to hedge muni inventory by shorting Treasury bonds could have lost more on the hedge than on the municipals.
“You went through a period where some days the two moved in opposite directions, purely for reasons that affected one market and not another,” said Jed McCarthy, managing director at 1861 Capital Management.
The correlation in the changes in 10-year single-A munis and 10-year Treasuries in the 10 weeks ending Jan. 9 was negative 0.29. Later in the year it was even worse. In the 10 weeks ending April 3, the correlation was negative 0.53. That means a broker would actually have been better off taking the opposite position of his hedge.
“Obviously the hedges proved disastrous from August of ’07 through the end of last year,” said George Friedlander, muni strategist at Morgan Stanley Smith Barney.
One of the main reasons the correlations broke down was the implosion of the municipal arbitrage industry, according to Friedlander.
Municipal arbitrage employed a strategy pitting long-term muni yields against short-term yields. Until mid-2007, this industry dominated the daily trading in municipals, Friedlander said.
The arbitrage strategy was largely implemented by tender-option bond programs, which took short positions on short-term munis and long positions on long-term munis, and hedged both positions with taxable rates, typically Libor.
Because the TOB industry by design bought or sold municipals when their yields deviated from historical patterns, Friedlander said they ensured the fluctuations in municipal rates did not veer too far from the fluctuations in taxable rates.
“The market got overly comfortable with the idea that the correlations would be close to perfect, because the TOBs made them close to perfect,” he said. “As we got deeper and deeper into the period where TOBs were dominating the market, any little blip in the yield relationship created a response from the TOBs in the opposite direction.”
The liquidation of the TOBs allowed more deviant correlations to persist, he said.
Friedlander said dealers are once again using Treasuries to hedge municipals as the relationship has stabilized. The correlations are not what they were when TOBs ruled the market, Friedlander said, but they are similar to what they were before that, when munis and Treasuries still tended to move in the same direction.
“You’re more back to 2004 than 2007,” he said, referring to the period before TOBs arbitraged brief meanderings in muni rates back into line.
In other words, the credit crisis did not reveal that Treasury rates and muni rates are not actually correlated, he said. The crisis simply showed that markets can stop working properly during a panic.
Muni rates and Treasury rates still move in tandem most of the time for good fundamental reasons, Friedlander said.
Those fundamental reasons form the underpinning of the assumption that muni investments can be hedged with taxable instruments in the first place, Friedlander said. The rates move in the same directions based on Fed policy, inflation expectations, and the general level of borrowing costs.
Friedlander cited another reason dealers are hedging munis with Treasuries now — a lot of people think if municipal rates bump up, it will be because they are too much lower than Treasuries.
With that in mind, he said, it makes sense to play one against the other.