30 Oct 2014
Declaration to bankruptcy court by GTAT's chief operating officer shows just how one-sided the doomed sapphire supply deal was.
by Mike Hatcher
Sapphire is one of the hardest substances known to man. And it is very hard indeed to manufacture – seemingly harder than Apple realized when it signed that huge supply deal with GT Advanced Technologies (GTAT) a year ago this week.
This week a declaration by Daniel Squiller - COO at GTAT and a key figure in the company’s recent chapter 11 bankruptcy filing - has laid bare just how much that deal was loaded in Apple’s favor, and how GTAT believes the iPhone maker’s interference played a part in scuppering the deal.
In document #0324 presented in the case, Squiller states under pain of perjury that the key to making the sapphire deal profitable for both sides was to mass-produce 262 kg boules of sapphire crystal meeting the specifications required by Apple.
Although GTAT was not itself a sapphire manufacturer prior to the near-$600 million Apple agreement, it had sold more than 500 furnaces for growing the crystal to Asian customers.
However, none of those furnaces had been designed to produce boules any greater than 115 kg in weight, while most producers using non-GTAT furnaces grow boules weighing less than 100 kg.
What Apple needed was unprecedented scale, coupled with material quality and within a challenging time-frame. If GTAT could meet the requirements, it would – in Squiller’s words – be “revolutionary” for the Merrimack-headquartered business.
What went wrong
Unfortunately, that didn’t happen and GTAT’s bankruptcy filing was the result. “Production of 262 kg boules of sapphire could not be accomplished within the time frames the parties had agreed, and was more expensive than anticipated,” Squiller explains in his written testimony.
So what went wrong? Fundamentally, it was that GTAT could not make the material to Apple’s specifications. But that, argues Squiller, was largely a result of using Apple-selected fabrication equipment that it appears was not fit for purpose.
“GTAT did not select the fabrication equipment. GTAT was unable to negotiate changes to the pricing regime set forth in the Apple [deal], and, therefore, GTAT was selling sapphire material at a substantial loss,” he writes.
Remarkably, it seems that GTAT had no direct contact with the suppliers of cutting and polishing equipment to specify the required tools, not surprisingly meaning that it was ultimately unable to achieve Apple’s desired sapphire fabrication cost and production targets. “Many of the tools did not meet their performance and reliability specifications,” reveals Squiller.
To make matters even worse, the supply agreement called for the price of the finished sapphire material to decrease in 2015 as Apple stepped up its use – even while the original targets could not be met.
$900 million bill
To date, says Squiller, GTAT has run up a staggering $900 million bill in connection with the project – of which $439 million was funded by the Apple’s pre-payment to finance the production scale-up.
Because of the failure to meet targets, Apple’s final pre-payment instalment of $139 million was not received, and unless supply prices could be renegotiated, GTAT would never have realized a profit on the deal. Despite some attempts to do that, no adjustment was forthcoming and GTAT's next stop was the bankruptcy court.
“In light of the inability to negotiate changes to the Apple Agreements necessary for GTAT to operate profitably, GTAT reluctantly commenced these chapter 11 cases to preserve the value of its business by separating itself from the business relationship with Apple,” is how Squiller’s declaration puts it.
By extracting itself from the exclusive deal and going back to selling furnaces (Apple and GTAT announced an agreement October 13), Squiller believes that GTAT should now be able to tap into what he calls “substantial pent-up demand” for sapphire material from other makers of smart phones and watches.
Whether GTAT can turn itself back into a successful equipment vendor remains to be seen, but Squiller’s testimony offers some unprecedented insights into the perils of agreeing a deal with what he calls “the ultimate technology client”.
Initially, the plan had been for Apple to actually buy 2600 sapphire-growing furnaces, to be operated by GTAT. But following months of negotiations, that changed. Instead, Apple would lend money to GTAT to build those furnaces. Crucially however, Apple would have no obligation to purchase them, or indeed any of the sapphire that they produced – substantially reducing its own risk in the venture.
“At the same time, GTAT would be precluded from doing business with any other manufacturer in or supplier to the consumer electronics market,” Squiller says, adding that a labyrinthine structure of leasing and lending subsidiaries also had to be set up per Apple’s insistence.
So, under the deal finally agreed, GTAT was to borrow $578 million from Apple’s vast cash pile. That money would be used to assemble some 2036 sapphire furnaces, most of which were to be installed at the new facility in Mesa, Arizona. GTAT would then manufacture sapphire according to Apple’s specifications and repay the loan using either cash or completed sapphire material as an alternative form of currency.
Clearly, for GTAT to make any money on the deal, Apple would have to buy more than $578 million worth of sapphire material. And in theory that could happen – especially if an iPhone featuring a unscratchable, shatter-proof sapphire screen proved popular.
What could possibly go wrong? In his testimony, Squiller highlights many of the issues at play, not least that although GTAT had committed to supply millions of units of sapphire material and could not engage with any other potential customers without express consent, Apple had no obligation to buy any of that sapphire material. At all.
Despite that, GTAT would have to accept and fulfill any purchase order placed on the date selected by Apple. Remarkably, the penalty for a late delivery would be $320,000 per boule of sapphire – compare that with Squiller’s cost estimate for a typical boule of less than $20,000.
There was more: Apple, without compensating GTAT, had the right to cancel a purchase order in whole or reschedule a delivery date - any time it liked.
Having raised the issue of what looked like a pretty one-sided agreement with “the ultimate technology client”, GTAT’s management was simply told that this was what other Apple suppliers had signed up to, and fell into line.
By June 2014, sapphire production at Mesa was supposed to be in full swing. However, the first phase of the build-out was not completed until December 2013. The site even needed some floors to be rebuilt, delaying the sapphire ramp-up by around three months. “This was critical lost time for GTAT,” says Squiller.
Power supplies were another problem. To grow the kind of high-quality crystals needed to meet the Apple specifications, a stable and reliable power infrastructure was critical. But in a decision that GTAT distances itself from, back-up power for the furnaces was deemed too expensive. You can guess what happened:
“On at least three occasions, power interruptions occurred, leading to significant delays and losses of whole production runs of sapphire boules,” says Squiller.
There is plenty more on the same theme in Squiller’s full declaration, and Apple will no doubt dispute some of the claims. Whoever's version is closest to the truth, one is left with two over-riding thoughts:
First, that Apple appears to have simultaneously underestimated the complexities of delivering high-quality sapphire and hampered GTAT’s ability to meet its specifications. In GTAT's view, a combination of top-down meddling and goalpost-moving made an already ambitious time-frame in which to scale an unprecedented sapphire production line a virtually impossible one. An extraordinarily one-sided approach to contracts, which appears designed to squeeze its suppliers until their pips squeak, only made matters worse.
Second, that GTAT’s management agreed to such a one-sided deal in the first place now seems bewildering. Squiller’s description of Apple as the “ultimate technology client” suggests that they may have been blinded to the risks of putting all of GTAT's eggs into one – admittedly attractive – basket.
Happily, Apple and GTAT have at least been able to reach a deal that may see the latter exit its bankruptcy proceedings with a viable business – assuming that the court accepts the agreement in the light of claims from GTAT’s other creditors and the deal holds good amid the public airing of Squiller's testimony. An extended legal wrangle remains a possibility, and there's no prizes for guessing who has the deepest pockets.
Less happily, spare a thought for the losers – the 1000-plus workers hired to produce sapphire at the Mesa facility in Arizona. Twice now, the town has been sold a failed promise of hundreds of quality manufacturing jobs: initially with First Solar, and now GTAT and Apple.
But if you’re interested in buying a furnace capable of making a quarter-ton lump of sapphire, you know where to go…
About the Author
Mike Hatcher is the editor-in-chief of optics.org.