LONDON/NEW YORK, Nov 20 (IFR) - The US$5.6bn debt sale backing Carlyle Group’s buyout of Veritas collapsed in spectacular fashion this week, leaving the eight underwriting banks hurting from a deal gone horribly wrong.
The US$3.3bn loan and US$2.275bn bond package was in line to be a marquee transaction, funding the largest LBO of the year - Carlyle’s US$8bn purchase of Veritas from Symantec.
But in the face of severe market headwinds - and thanks to a couple of clumsy missteps by the banks - the deal unravelled quickly before the eyes of an astonished market.
The banks, led by Bank of America Merrill Lynch and Morgan Stanley, thought all was rosy when they decided in August to underwrite the acquisition.
Symantec was spinning off Veritas, its data storage unit and a top name in the hot cloud-computing sector, and Carlyle was teaming with GIC, a Singapore sovereign wealth fund, to buy it.
GIC and the US private equity giant had full confidence in the deal and showed that confidence in the best way possible - with a US$2.65bn equity contribution.
“It’s an US$850m Ebitda business that is number one in what it does, and what it does is mission-critical to clients,” one banker on the deal told IFR earlier this month.
“I see why it’s Carlyle’s biggest equity cheque ever.”
From the outset, though, timing was a problem.
With the acquisition announced in August and expected to close by year’s end, the banks had to bring to market the financing package - which was split across dollar and euro tranches - at an especially difficult period.
August began a brutal sell-off in US high-yield, especially at the lowest end of the credit spectrum, and the Veritas package included US$1.775bn-equivalent of Triple C rated bonds.
By the time the deal was finally presented to investors in the first week of November, Triple C paper had gapped out to nearly four-year wides on the main US indices.
Moreover, the banks had set the cap rates on the bonds - widely believed to be 7.5% and 10.5% - at the start of August before the market turned sharply south.
When the leads were trying to salvage the deal by offering around 8.5%-8.75% and 11.5%-11.75%, some 175bp wide of initial whispers, investors knew the banks were on the hook.
And there were already plenty of reasons not to like the bonds.
Among other things, the covenants allowed the company to incur secured debt up to 4.75x Ebitda but the language used hid some additional capacity to go past that limit, according to Scott Josefsberg of Covenant Review, an independent research firm.
Other provisions allowed the private equity sponsors to pay themselves a dividend to the tune of at least US$275m on day one, said Josefsberg.
As one high-yield portfolio manager told IFR: “The covenants on this deal were awful.”
In addition, he said, there was confusion in explaining the Veritas business - and difficulty convincing the buyside that its outlook was as positive as Carlyle believed.
“I couldn’t get comfortable that the core business was insulated from declines elsewhere,” he said.
A second portfolio manager echoed those concerns.
“The private equity sponsor articulated that it took them six months to completely understand the company. And they expect us to make a decision in 24 hours.”
To make it all seem even more questionable, Symantec were selling Veritas for only US$8bn - or US$5.5bn less than they paid to buy the business in 2005.
GAME OF CHICKEN
But the biggest weakness may have been in the loan - where timing again turned out to be an issue.
The B2/B rated US$3.3bn term loan was in the market while chipmaker Avago Technologies was selling better-rated tech paper - a Ba1/BB+ rated term loan that was upsized to US$9.75bn.
Moreover, investors realised that with the end of the year approaching, banks would have to absorb even more pain if they were forced to hold the loans over into 2016.
“You’re trying to sell a deal with really high leverage coming into Thanksgiving and Christmas,” said the head of high-yield at one asset management firm.
“Everyone knows that banks close their year around the end of November. So it turned into a game of chicken.”
And the banks blinked first.
They offered to cut the size of the US Term Loan B to US$1.5bn from US$2.45bn - moving US$250m to the bonds and taking the extreme measure of retaining US$700m of it.
It was, said a leveraged finance banker, a violation of the “cardinal rule” to only announce the specific amount you are holding if you know the rest will be sold.
“But it turned out they had a whole bucket of nothing,” the banker said. “When you have a whole bucket of nothing, you’re not supposed to announce all those crazy iterations.”
He said cutting the size of the loan was “an admission that you absolutely do not have a deal together”.
Moreover, it emerged that the banks were only planning to hold the US$700m for four months, according to a loan investor.
He said this was a major issue for investors because the company would not publish its first set of audited numbers for another six months.
From there the deal unwound fairly rapidly.
Underwriters sweetened the loan’s terms to include a higher spread of 500bp over a 1% Libor floor - and a steeper original issue discount of 95 cents to the dollar.
When that still wasn’t enough to get the buyside on board, they were heard bumping up the discount as low as 90 - all to no avail. The financing was finally pulled on Tuesday.
While the Veritas acquisition will still go ahead - the banks will have to fund the bridge loans themselves if unable to offload the debt before year end - market participants are shedding few tears for them.
“It’s good news this was actually pulled,” said the high-yield manager. “It was a step too far in both leverage and terms.”
A London-based portfolio manager called the collapse “a reality check for the banks”.
“Aggressive Triple C deals are not what the market wants right now,” he said.
“There’s plenty of supply looking to come. But unfortunately none of it is in the shape of what people want to buy.”