To view the digital version of this report, please click here.
An excess of anything can have some unsightly consequences, even in the world of bonds.
US corporates have been feasting on record low funding costs for more than a year now, thanks to the Fed’s stimulus.
A steady stream of funds out of low-yielding Treasuries and into corporate bonds has also helped issuers such as IBM achieve some of the lowest coupons ever seen in the history of the US capital markets.
But that has left many issuers with a string of ugly high coupons and high dollar-priced bonds at certain parts of their credit curve, which in turn distort their ability to price new deals.
IBM is the first example this year of a company embarking on a much-needed makeover at the long end of its curve.
In May IBM asked bondholders to tender US$322.1m of 7.125% 100-year bonds due in 2096, as well as US$186.67m of 8% 2038s and US$800m of the outstanding US$1.54bn 5.6% 2039s, in exchange for a new 30-year bond with a 4% coupon.
It had received an overwhelming response at the early tender date of June 4, with almost US$847m of 5.6% 2039s and about US$104m of 8% 2038s tendered. The 30-year did not have an early tender date.
At first glance, it is hard to understand why investors would want to give up IBM bonds with coupons that an average junk-rated company pays these days.
However, with a record number of investment-grade bonds now trading well above par and suffering extreme illiquidity as a result, this year could see investors more willing to give up their coveted high coupon investments.
According to Barclays, about 91.9% of all of the bonds in the Barclays US Corporate Investment Grade Index now trade above par, and more than 48% of them trade at more than US$110. The average price of the index is now at US$111.30.
For the corporate index, 541 bonds, or 14.24% of the market value of the index, trade above US$125.00
Some, like the 8% 2038s IBM is tendering for, trade as high as US$165.00. Its 5.6% 2039s trade around US$127–$128 and the 7.125% 2096s trade around US$150–$156.
Banks are also potential candidates for taking out high coupon debt.
“There were a handful of banks that issued high coupon debt around the crisis, that have since improved from a credit ratings perspective and are now keen to take those bonds out, now that their financing costs have come down due to a combination of low overall rates and tight credit spreads,” said Saurabh Monga, a director in Deutsche Bank’s capital markets and treasury services group. “A lot of banks are at the margin trying to improve their capital structure and reduce the cost of funding in order to help earnings.” In April, Discover Financial Services, for instance, swapped US$400m of 10.25% 2019s that were trading around US$135.00 for new 2022 5.2% bonds at par.
The old bonds have soared in price precisely because they offer such high coupons.
Union Electric’s 8.45% 2039s, for instance, trade around US$162; MetLife 10.75% 2039s at US$138 and Altria’s 10.2% 2039s around US$158.
While they love the coupons, liquidity has also become a high priority for portfolio managers. When an issuer like IBM is willing to pay a premium over the market value of the bond as well as a liquid on-the-run security in return, investors can be enticed to loosen their hold on high coupon debt.
“Exchange offers have more appeal to investors now than they did even a year ago, because secondary market liquidity has deteriorated and investors have limited appetite for very high dollar price bonds,” said Pamela Au, head of liability management at Barclays. “These bonds have underperformed due to their lower duration and every dollar above par translates into additional recovery risk for investors.”
High dollar price bonds naturally become more illiquid, because in a default scenario, a bondholder can only claim up to the par amount.
“Every dollar a bond trades above par is a dollar you will not get back in a default scenario,” said Perry Piazza, director of investment strategies at Contango Capital Advisors.
“Exchange offers have more appeal to investors now than they did even a year ago, because secondary market liquidity has deteriorated and investors have limited appetite for very high dollar price bonds”
Also, high coupon and high dollar price bonds have much less convexity than par bonds, which means that the rate of change in their price as yields change is lower than what’s seen in a par bond.
That impinges on their appeal to investors who trade actively in the bond market. “Convexity is an attribute that is helpful to traders so as a result bonds with less convexity don’t trade as well,” said one banker.
Some insurance companies are also restricted from buying very high dollar priced bonds.
For most companies, having high dollar bonds is a high class problem. But for a blue chip issuer like IBM, which has consistently broken the record book in the past year for low coupons, its stable of high coupon longer-dated bonds are an eye-sore and one that has real pricing consequences for a new deal.
“We often make the case to issuers that a high dollar, high coupon bond can distort their credit curve,” said one head of liability management at a major Wall Street firm. “If we are going to bring a new bond to market for an issuer, the most obvious place for an investor to look for a pricing comparable is that issuer’s existing bonds in the same part of the curve.”
Exchanges make sense for companies that can extract a net present value positive scenario after taking up-front tax benefits of paying off old debt and amortising the cost of doing so over the life of the new bond.
IBM relied heavily on the offer of a liquid benchmark, new 30-year to entice bondholders to tender, especially those holding the 100-year bonds maturing in 2096.
IBM made the exchange terms for the 2039s the most attractive of the three. It capped the exchange on this note to US$800m of principal. The 2039s were created in 2009, as part of a tender and exchange for the 2038s and 2096s at the time.
The company paid less for the 2038s than it did in 2009 and nothing in the way of a premium on the 2096s.
Even so, bankers away from the deal thought IBM generally offered a fair trade.
“We expected pretty good participation, although smaller participation for the 38s and the 96s because they tendered for the residuals left from a previous exchange and so we expected they would be harder to get.”
One banker thought IBM offered about a 20bp pick-up on average for the 39s and 38s, and nothing for the 96s.
The 2039s were quoted at around US$127–$129 to yield about 3.84% at the higher price, in the week of the exchange announcement. IBM was offering an extra US$40 for every US$1,000 denominated bond if the bonds were tendered by an early exchange date of June 4. That equates to a US$1,155.07 principal amount of new notes for every old bond tendered. Bondholders also received US$150 in cash per old bond.
The 2038s were around US$165.00 to yield about 3.82%. IBM offered an early exchange premium of US$40 principal amount of new notes, but no cash. That equates to US$1,694.8 principal amount of new notes for every old bond.
IBM did not, however, offer anything in the way of a premium on the 2096s. Each old bond was exchanged for US$1,369.69 of new notes, plus US$200 cash. Those bonds were trading around US$150–$156 when the exchange was announced, to yield around 4.6%.
“The tender for the 2096s was really about the liquidity they were offered in the new bond,” said one banker not involved in the deal. “The 2096s were trading at US$125 in 2009 and during the exchange period they were trading at US$156.”
Exchanges rarely get full participation and one that takes out at least half to 75% of the bonds eligible for exchange is considered a success.