Being a rare double-A rated issuer out of China, CNOOC Limited as expected attracted a lot of demand for its $2 billion dual-tranche bond, which also tapped into the current investor appetite for long-dated paper. Wanting to take advantage of the latter, but at the same time keep its funding costs down, the Hong Kong-listed offshore oil producer decided to issue $1.5 billion of 10-year bonds due 2021 and $500 million of 30-year bonds due 2041.
The desired split wasn’t disclosed during the marketing, although one banker said the company did have a preference for 10-year paper since it will not swap the proceeds and therefore will have to deal with the outright cost.
When the deal priced early in the morning in Hong Kong on Saturday following a one-week roadshow and two days of bookbuilding, it had accumulated a total order book of $5.7 billion, with $3.3 billion of demand for the 10-year tranche and $2.4 billion for the 30-year. Based on the final split, this left more excess demand from the longer-dated of the two tranches, which was 4.8 times covered, compared with 2.2 times for the 10-year.
The impact of this was evidenced in a relatively tight 10- to 30-year spread, but overall the issuer refrained from pricing the bonds too aggressively – something which had been a concern during marketing among some market watchers who had feared that the company may lean too much on the huge bid from Chinese banks, which typically isn’t price sensitive.
But the concerns proved unwarranted and CNOOC “didn’t sacrifice distribution for pricing”, one source said.
For sure, the pricing was tight, but the fact that both tranches kept trading at around the re-offer price yesterday afternoon, with a slight spread tightening for the 30-year, was evidence that it hadn’t pushed too far.
However, the pricing was tightened quite substantially during the two-day bookbuilding, especially for the 30-year tranche which was supported by a large appetite for yield by US investors in particular. The initial whispers of a yield spread in the low 100s for the 10-year tranche and mid-100s for the 30-year, which were communicated when the bonds were launched early last Thursday, gave way to an initial guidance of US Treasuries plus 100bp to 110bp for the 10-year and Treasuries plus 125bp to 135bp for the 30-year during New York hours later the same day.
On Friday this was firmed up even further to 100bp plus/minus 5bp for the 10-year and 120bp plus/minus 5bp for the 30-year, with both tranches eventually pricing right at that guidance.
The 10-year came with a 4.25% coupon and was reoffered at 98.59 for a yield of 4.426%, or a spread of 100bp over the 10-year US Treasury.
The 30-year was priced with a coupon of 5.75% and was reoffered at 99.448 to yield 5.789%, which translated into a spread over the 30-year US Treasury of 120bp.
Barclays Capital, BOC International, Bank of America Merrill Lynch, Citi, Goldman Sachs and J.P. Morgan were joint bookrunners.
With no obvious comparables bankers used a variety of benchmarks to guide them on pricing, including top-tier triple-A rated names like Temasek, double-A rated overseas oil producers such as Royal Dutch Shell and Statoil, single-A rated names like Conoco-Philips and BP, as well as Chinese issuers like China Development Bank and a few others.
CNOOC, which is rated Aa3 by Moody’s and AA- by Standard & Poor’s, did price wider than most of the double-A rated overseas oil producers. For example, Shell has a 10-year bond that was trading 50bp above Treasuries at the time of CNOOC’s pricing and a 30-year that was at 70bp over. The high-grade Chinese issuers are mostly shorter-dated, but were still trading at spreads above 100bp.
One analyst also noted that double-A rated 10-year industrial issues in the US are on average traded at a spread of Treasuries plus 90bp, which makes CNOOC’s pricing “fair to tight”.
Sinochem’s 10- and 30-year dual-tranche bond that was issued in November was discarded as a good comp because of that company’s lower triple-B credit rating and, indeed, CNOOC’s new bonds came well inside the levels where Sinochem was trading. That said, Sinochem was used as a reference with regard to the 10- to 30-year spread, which was quoted at about 35bp at the time of pricing. Reliance Industries was trading at a 10- to 30-year spread of around 30bp, while other similarly-rated issuers were on average fetching spreads of 25bp-30bp, according to bankers. This made CNOOC’s gap of 20bp seem very tight indeed.
In the late Hong Kong afternoon yesterday, the 30-year tranche was trading at a slightly tighter spread of 118bp/116bp, while the 10-year was quoted at 101bp/100bp.
US investors were particularly keen on the 30-year bond and were allocated 64% of that tranche, while 25% went to Asia-based investors and 11% to Europe. Of the 10-year bond, 43% went to the US, 47% to Asia and 10% to Europe. In all, 184 investors participated in the 10-year tranche and 116 in the 30-year.
In terms of investor type, funds bought 56% of the 10-year, banks 30%, public investors/agencies 10% and others 4%. Of the longer-dated bond, 74% was taken up by funds, 19% by insurance companies and 7% by others.
The deal is the largest US dollar bond by an Asian issuer so far this year, and sources said the successful deal may encourage other high-grade issuers to also look at long-dated issuance.