It should go without saying that getting into iBed with Apple (NASDAQ:AAPL) cuts both ways. Investors are easily hyped at the prospect of a company becoming an Apple supplier due to the potential upside. At the same time, it's common knowledge that Apple squeezes its suppliers in unforgiving ways.
By far, the most prominent public example is GT Advanced Technology (NASDAQOTH: GTATQ), which announced a shocking Chapter 11 bankruptcy filing in early October. Both Apple and GT have fought to keep the details of its partnership under wraps. But due to the interests of all of GT's public shareholders that woke up to losses in excess of 90%, a bankruptcy judge has now ordered that many documents become unsealed.
One such document is an unedited affidavit from GT Advanced COO Daniel Squiller, shedding unprecedented light into some of the negotiating tactics that Apple utilizes behind closed doors. The most shocking one is an attempt to emasculate GT's senior management.
"Put your big boy pants on and accept the agreement"
Apple structured this deal dramatically in its favor (more details on this later). When GT senior management expressed concerns about the asymmetrical nature of the arrangement, Squiller attests that Apple told management to "put on your big boy pants and accept the agreement."
On top of that, Apple also said GT should "not waste their time" trying to negotiate the arrangement, because Apple simply does not negotiate with its suppliers and that other suppliers face similarly stringent conditions.
These sound like an awfully juvenile tactics when you're trying to convince a potential supplier into transforming its entire business model for your benefit. Well, it worked.
Who's the boss?
Squiller said that Apple exerted "inordinate control over GTAT's liquidity, operations, and decision making." While Apple was approaching GT as a potential customer, the structure of the deal also put Apple in a lender role. Apple even required GT to pay off its Bank of America credit facility so that Apple could have a lien on GT's assets, although it did permit the company to subsequently raise nearly $300 million in capital through debt and equity offerings.
This control extended to the Mesa facility operations, where Apple "embedded itself." Apple employees with no experience in sapphire growth or furnace operation oversaw operations, and at times would occupy 30% of the research and development and manufacturing teams' time. There were even times when Apple employees would instruct GT employees on what to do, and GT had to remind these Apple employees that doing so was inappropriate.
GT advised Apple that having uninterrupted power was "essential" to the facility, which was larger than multiple football fields. Any power disruptions could render sapphire being grown unusable. The company estimated that even a brief power interruption could result in losses of upwards of $30 million, and requested Apple ensure backup generators were available. Apple considered the request, but determined the generators were too expensive and deemed them "non-essential."
There were 3 separate occasions where the facility saw power interruptions, resulting in entire production runs of sapphire boules being lost. GT estimates cumulative losses at $10 million. On top of that, GT lost important data during each of these power interruptions that would have helped it learn and improve the manufacturing process.
The Mac maker had also negotiated deep pricing discounts, allowing it to buy sapphire at "below market value." GT had no way to come out ahead unless Apple opted to buy sapphire in excess of what it owed. The sapphire maker was struggling to produce the material in an economically viable way that met Apple's stringent and constantly changing specifications (which "remain in flux to this day"). These changes resulted in additional losses.
GT tried to negotiate pricing concessions to no avail, and was "forced to sell every unit of sapphire material at a substantial loss." Even worse, the agreements called for price decreases in 2015, which would have only accelerated GT's losses. So far, GT has incurred $900 million in costs related to the project.
GT felt that it had been "forced into the role of a 'captive' supplier," and the Chapter 11 filing was the only way to "extract itself from Apple's control." Even if the project had been successful, GT says "Apple would have obtained a groundbreaking product from GTAT at below-market cost and GTAT would own 2,036 well-used furnaces with limited resale value." Squiller even characterizes the deal as a "risk-free option to acquire millions of highly engineered units of sapphire material."
Deal with the iDevil
We already knew that GT was required to supply a certain minimum quantity of sapphire, but that Apple was not required to purchase any minimum. It's also been publicly disclosed that GT would have to pay $50 million per occurrence for any breaches of its confidentiality agreement, which is partially why it fought to keep the court documents sealed.
But these are just scratching the surface of how one-sided this deal was (insert sapphire scratch resistance pun here). Here's some of what Squiller calls a "best of" collection of the contractual terms.
- If Apple placed an order for sapphire, GT would need to ensure it arrived by specific delivery dates. That could include paying out of pocket for expedited shipping costs if necessary. If GT's delivery was late, it would have to pay Apple $320,000 in damages per sapphire boule. A sapphire boule costs less than $20,000.
- If GT sold sapphire boules to rival OEMs in violation of its exclusivity agreement, it would have to pay Apple $640,000 in damages per boule.
- If GT operated a sapphire furnace for the benefit of a rival OEM in violation of its exclusivity agreement, it would have to pay Apple $650,000 in damages per month of operation. A sapphire furnace costs $200,000 (one-time upfront) to purchase and assemble.
- GT was unable to modify any equipment, manufacturing process, or product specification without Apple's consent. Apple can change any of these at any time and GT must immediately comply.
- Apple initially required GT to set up a wholly owned subsidiary, GT Equipment, that was little more than a shell holding company with no economic purpose. GT Equipment primarily shielded Apple from risk. If Apple were to effectively become a lessee of the furnaces under a Termination Event, Apple and GT would have to pay rent to GT Equipment under a lease agreement. GT would have to pay $9.9 million per month, while Apple would pay $50.
- Sapphire would be sent from GT to two other Apple vendors for further finishing. If any material was deemed defective, a 3-party committee consisting of one of these vendors, Apple, and GT would vote on whether or not GT was at fault and liable. Majority wins.
In effect, Apple had turned GT from a successful supplier of sapphire furnaces into an "experimental research and development venture for Apple funded substantially by GTAT's other stakeholders."
Apple had confirmed in August 2014 that it would make the fourth and final prepayment (which is why GT management was so confident on that fateful August conference call), but quickly changed its mind on product specifications again, allowing it to withhold payment. Weeks before the bankruptcy filing, GT made a presentation to Apple, outlining the higher-than-expected costs and explicitly telling Apple that it would run out of cash within weeks.
Doesn't this sound like a terrible deal?
You might wonder why GT ever agreed to this deal in the first place. Well, Apple initially approached GT with a different proposal. At first, Apple wanted to purchase over 2,600 furnaces from GT but have GT operate them on its behalf, which would have been the largest purchase order in GT's history, but in line with its existing business model.
After months of negotiating, the deal evolved to the current one (which GT considers a "classic bait-and-switch"). Apple wanted to loan GT the funds to assemble 2,000 furnaces (which GT would own), and then GT would produce and sell sapphire to Apple. GT had brief talks with rival OEMs, but after spending months working on "the most significant contract in the company's history," the company felt that it was "out of options" and didn't have time to follow through with the alternatives.
Beyond that, GT management had more to gain personally while shareholders bore the brunt of the risk. Management undoubtedly knew that the deal's announcement would send shares soaring following investor hype, and several executives set up Rule 10b5-1 trading plans following the announcement to cash out shares over time. Much like the partnership itself, there was an asymmetrical risk profile here. Management benefited from the sheer hope that the deal would succeed, while shareholders bore the outsized risk of its failure.
Besides, GT needed to assure Apple that it had indeed put on its big boy pants that morning, and that GT's dad could beat up Apple's dad any day of the week.