Resurrecting Iridium

Continuing my report on Eccentric Orbits: The Iridium Story, one of the Wall Street Journal “20 books that defined our year” for 2016.

The central figure in the book is Dan Colussy, a former Coast Guard officer who ran Pan Am for 10 years and then did a rollup of government contractors called United Nuclear Corporation, sold to GE in 1997.

The central villain is Motorola. They had insurance against their satellites falling out of the sky and hitting people but it wasn’t of unlimited duration. When Iridium went bankrupt, after paying out most of the funds raised to Motorola itself, the Motorola executives desperately wanted to de-orbit the $6 billion worth of satellites, thereby insuring that their investors and partners (the gateway owners, the handset manufacturer Kyocera, et al.) would be left with nothing.

Who was partying with the bondholders’ cash during the Chapter 11 proceedings that started in August 1999?

The legal profession was very happy to assist in the demise, restructuring, and rebirth of Iridium, and by one estimate the lawyers were burning through $4 million a month in fees approved by the bankruptcy court. You had lawyers for Iridium, lawyers for the creditors, lawyers for Motorola, lawyers for Boeing, and lawyers for every gateway. You had lawyers on staff and outside lawyers for specialized matters like licensing and insurance. The White House counsel was involved, as were the employees of the biggest lawyer of them all, Attorney General Janet Reno. There were lawyers from the Departments of Commerce, Defense, State, and Treasury, including one—David Cohen—whose title was Senior Counsel to General Counsel in the Office of the General Counsel. Colussy hired lawyers to deal with the FCC, the various aviation insurance companies in Paris and London, and the International Telecommunication Union in Geneva.

What was the risk? The good news for fans of climate change fans is that the “science was settled.” After all, the formulae for orbital mechanics go back to Isaac Newton, 1684-7. Estimate #1:

[NASA] said the chances of being hit by falling debris would be 249 to 1, with the chance of an actual death at 40,000 to 1. The NASA analyst believed that, in addition to the titanium fuel tank, five other parts would survive reentry: the battery, three structural brackets, and the electronic control panel.

Estimate #2:

Risk assessments carried out by aviation insurance experts hired by Boeing estimated the probability of personal injury from de-orbiting the Iridium satellite system at 1 in 1.063 trillion.

Nobody ever figured out how to reconcile these apparently conflicting results.

No technology is so beautiful that it couldn’t be ruined by software developers:

Each country billed its Iridium phone service in local currency, and each customer had a choice of four different service plans. That meant that a call could originate in a country licensed to use Zambian kwachas, go through a gateway priced in Brazilian reals, and terminate over landlines priced in Peruvian nuevo soles. If the phone had been purchased in, say, Oslo, then all of this would have to be fed into the Reston computers, where it was converted into South African rands for the originating gateway and Saudi riyals for the gateway owners, translated into American dollars for Iridium, and summarized in a monthly bill denominated in Norwegian kroner. The diciest part of the whole system was the terminating leg of the call—the part that continued on land after the satellites passed it down. These “tail charges” could be as low as 4 cents a minute or as high as $2.50 a minute, depending on where the call ended up, and that was on top of the already exorbitant Iridium per-minute charges—anywhere from $3 to $7. Iridium had hired Andersen Consulting to program ten million lines of computer code for the billing system, but the system was unstable, functioning so erratically that Leo Mondale eventually developed an in-house Excel system that worked just as well. Naturally the customers were annoyed by the unpredictability of it all, and one of the first decisions Colussy made was to change all billing to one price and one plan. Iridium would get 80 American cents per minute, period. Service providers could charge whatever they wanted—most charged around $1.50—but the wholesale price was 80 cents. (Ten years later, that was still the Iridium price.)

The U.S. government was an important future customer, but because hadn’t yet invaded Afghanistan or Iraq, people in Washington and at the Pentagon didn’t feel a huge urgency. The U.S. government was also critical for dealing with Motorola, which insisted on a total release of liability above and beyond what any commercial insurance policy might cover.

The Washington bureaucracy fully deserved its reputation, and any agency with the word “license” or “permit” in its charter was bound to be staffed by an army of people whose deliberately measured passive-aggressive voices threatened to put you to sleep at any moment. The goal of the process seemed to be the collection of bulky brown folders full of data.

Colussy hires the right lobbyists, though

… senior partner at the vast Sidley Austin firm, to carry his case before the FCC, the insurance companies, and any other regulatory agencies that cropped up and to ensure that all the licenses were transferred. In the rarefied world of “regulatory enforcement law,” Carter Phillips was king, having been an Assistant Solicitor General before going into private practice, and currently holding second place on the all-time list of cases argued before the Supreme Court (in excess of fifty)—cases touching such obscure and complex parts of the federal bureaucracy that, if placed end to end in a bound volume, would render the most dedicated law student comatose.

Eventually he gets to the right person in the Pentagon:

In many ways a meeting with the Deputy Secretary had an air of seriousness that meetings with the SecDef lacked, mainly because you knew that, at the end of the meeting, the decision was likely to be made one way or the other. The SecDef was the person you talked to at the beginning of a process. The DepSec was the last person you talked to.

“I have a message for your CEO,” said [Rudy] deLeon. “Tell Mr. Galvin that if I hear one more fucking threat to bring these satellites down, then his corporation is going to have a really hard time doing business with the Pentagon in the future.” The room went still. “I will give him that message,” said Schaffner. “All right,” said deLeon, “we should all now work together to get the constellation indemnified so that we can move forward without destroying it.”

More than a year after the bankruptcy filing, the Pentagon wakes up a bit:

in October [2000], USS Cole, a guided-missile destroyer out of Norfolk, Virginia, was going through a routine refueling stop at the port of Aden, in Yemen, after coming through the Suez Canal. It had pulled up next to a “dolphin” platform—the equivalent of an offshore gas station—… Forty-eight minutes after refueling began, an inflatable small craft manned by Ibrahim al-Thawr and Abdullah al-Misawa was seen pulling alongside the Cole by the duty officer. Al-Thawr and al-Misawa both stood at military attention and saluted as they approached. What was impossible to see from the deck was that the two men were standing on top of 250 kilograms of composition C-4 plastic explosives, and once they pulled alongside, they used a detonator to blow a sixteen-hundred-square-foot gash in the side of the Cole, directly opposite the mess hall, while most of the crew was at lunch. The two al-Qaeda bombers died instantly, and the Navy casualty toll would end up being seventeen dead and thirty-nine injured. … But the communications center aboard the ship had been disabled by the explosion, so his only contact with the Cole was through a single phone, borrowed from the U.S. embassy—an Iridium phone.

Colussy was trying to get a contract from the Feds for $72 million for 20,000 phones, to be used worldwide, and unlimited minutes over a two-year period. (That’s $150 per month per user.) The military’s proposed alternative was going to cost at least $10 billion to construct. The repair of the Cole alone cost $250 million (source). Despite the comparatively miniscule cost of the Iridium contract, it seemed that the government had endless file cabinets full of bureaucrats who needed to look at it. This gave competitors time to poke around and lobby as well (it seems that senators were cheaper to buy than satellites):

Unfortunately, Senator Ted Stevens and Representative Jerry Lewis were not going away. Keith Bane, newly energized by the prospect of McCaw owning Iridium, had Motorola lobbyists in Washington working on Stevens, and even volunteered to make a call to the senator himself. The grassroots efforts in Stevens’ home state were starting to work as well. Sam Romey, a rough-and-tumble mountain man and service provider in northern Alaska, was so infuriated by the Stevens-Lewis letter that he had rounded up a contingent of Alaska State Troopers to talk to Stevens. But Bernie Schwartz wasn’t defeated so easily. One of the most powerful Democrats in the Senate—Ted Kennedy of Massachusetts—was now lobbying for Globalstar, calling the Pentagon to ask questions, but even that was not as surprising as the fact that one of the most powerful Republicans in the Senate had jumped onto the Schwartz bandwagon as well—Trent Lott of Mississippi.

Back in Washington, Dave Oliver called Colussy to say, “I have neutralized the Democratic side in Congress”—but now Globalstar had a new ally in the form of the Office of Management and Budget (OMB). For weeks Oliver and deLeon had been coordinating all the approvals Iridium would need through a group of twelve agencies and departments, and there had been dozens of e-mails, memos, and PowerPoint presentations as they got ready for the showdown meeting in the Situation Room. (Dannie Stamp called it “organizing the circle jerk.”) But that debate had occurred without involving OMB. Now OMB had announced its intention to oppose the Iridium deal. Bringing OMB up to speed was like starting the whole process all over again, with every policy wonk in the city now alerted to a unique situation that read like a case study at the Kennedy School.

And then, as if the story couldn’t get any stranger, word came from Isaac Neuberger that weekend that President Clinton and [Motorola CEO] Chris Galvin were playing golf together in Vietnam. What he didn’t say was that on the golf course earlier that day, Galvin had spent most of the round telling Clinton that it was better to let the satellites be destroyed, because they had a life span of only five years anyway and some of them had already been flying for three years, so it was silly to go through all these machinations in Washington just to keep a constellation alive that couldn’t be replaced as satellites went out of commission. After all, who was going to throw another $6 billion after the $6 billion that had already been spent? Clinton was “spooked” by the remarks, according to Neuberger, and passed them along to Sperling.

Sperling was still jet-lagged from his 3:00 a.m. touchdown from Vietnam when he arrived at the West Wing to chair the Situation Room meeting on November 20. Keith Bane’s de-orbit deadline was two days away, and the battle lines had been clearly drawn. There were twenty-eight people present from nine departments and agencies, but only two mattered. The Office of Management and Budget was determined to change this deal or get rid of it entirely, and the Pentagon was determined to make it happen.

All of the alternatives to Iridium were technically inferior or still on the drawing board, so no matter how many politicians they bought the competitors couldn’t quite kill Iridium, e.g., Ted Stevens had to back off when Alaska State Troopers and natives told him that Iridium was essential to search-and-rescue for his constituents.

Colussy buys the $6.5 billion system for $6.5 million:

But the cash up front wasn’t $25 million; it was actually only $6.5 million. A few days before the hearing Colussy had polled his investors and come up with just $102 million promised, with at least $10 million of that being kind of shaky. That would normally be enough to open for business, but he owed $25 million to the banks and $15 million to Motorola for the property in Tempe. That put him back at $62 million in working capital, give or take, and he really needed $80 million to guarantee he could operate the satellites for a full year while the company started to build up revenue. Time was up. He had to go to the bankruptcy court, and he was $18 million short. Colussy’s solution: he told the banks that, if they wanted the deal to close, they needed to lend him $18.5 million of the $25 million he would owe them at closing. In return, the banks would receive 5 percent of the new company. Since the main goal of the banks was to get control of the $155 million in the Iridium cash drawer, and since the other bidders for Iridium were unlikely to be approved by the bankruptcy judge, what other choice did they have? Ironically, the total cash outlay of the Colussy group—$6.5 million—was the precise amount that Bob Galvin had authorized in 1989 to fund the Iridium brain trust.

It is actually a bit unclear when the company finally emerged from Chapter 11. It seems that one relevant date is December 2000 (nearly 1.5 years after the filing), but the case was kept alive through 2008 while the banks and other investors pursued Motorola to try to recover some of the cash.

Colussy streamlines operations:

From his years of acquiring companies at UNC, he was a believer in keeping people and getting rid of real estate, and he followed that formula here. The only building he wanted was the Tempe gateway, which resembled a state park nature center flanked by three giant globes designed like Rubik’s Cube sculptures. (Called “radomes,” these were the actual earth-station antennas that processed the calls.) Motorola had donated the land for the gateway to Arizona State University, then leased it for eighty years, so Colussy would simply take over that lease. The Chandler Lab, on the other hand, wouldn’t be needed. Colussy wanted Motorola to give him everything for free in return for taking the system off its hands, but Ted Schaffner had come down from Chicago and was computing how much each transfer would cost. Motorola was the kind of company that would, on the one hand, say that something was too small to bother with (it was shutting down two plants that could have been of value) and, on the other hand, demand fees and transfer payments and percentages of future revenue for tiny assets that were the equivalent of the albatross computer building in Reston. There turned out to be seven spare satellites, not four, that needed to be launched. (It’s hard to say how you misplace satellites, but apparently Motorola had done so.)

Colussy kept trying to retire, but couldn’t find decent CEOs to replace himself.

By the end of 2005 Iridium had shown its first small profit ($30 million). But that was also the point at which Colussy, with board approval, asked for the resignation of Carmen Lloyd, the company’s fourth CEO, because Lloyd had been at the helm during a disastrous failed merger with Inmarsat that ate up nine months of due diligence time and, in the opinion of some, gave away trade secrets to the enemy. Lloyd had also tried to raise $100 million in a Bank of America bond offering—and failed. In early 2006 Colussy apologized to his wife once again and told her that he was going to spend his seventy-fifth birthday functioning as a full-time CEO. Four months later, Iridium had $210 million in interim financing. Colussy had done a one-man road show and come home with the bridge money the company would need. But Colussy had now interrupted his retirement for six full years, and he wanted out. He wanted to be building his final home in Florida and resuming his golf lessons.

[Note that by moving from Maryland to Florida, Colussy was skipping out on 16 percent estate tax imposed by the state and an income tax of nearly 6 percent.]

Department of Nobody Knows Anything:

Oddly, many of the uses Colussy had predicted in that first meeting in John Castle’s Palm Beach house—the most obvious uses—never worked out. Year after year, Colussy thought the major airlines of the world would adopt Iridium as their first choice in cockpit communications and air traffic control, but time and again they passed on the mere $30,000 per plane that would have been required to employ Honeywell’s AirSat system. To demonstrate just how reliable it was, Honeywell installed it in an F-16 and had the pilot do three- to five-second aileron rolls at Mach 1.6 in an effort to lose the signal. The best the pilot could ever do was a slight fade when the plane was upside down.55 Nevertheless, the airline industry as a whole was continuing to use Inmarsat, despite it being triple the cost of Iridium and despite it being worthless beyond the 65th parallels, as late as 2011. That was the year that Iridium was finally certified by the FAA for aviation safety services, and slowly the airlines started coming around to AirSat. Likewise, the plan to provide village phone service to the most disadvantaged parts of the Third World never panned out, despite concerted efforts by Syncom. Shortly after the new Iridium launched, a special WorldTel division was created for western Africa. Ty Brown made several trips to Ghana, Senegal, and Nigeria in 2001 and 2002 to secure licenses but could never line up any African partners.

“We were threatening British Telecom and France Télécom,” said Brown. “They wanted to develop something by themselves. If we developed an alternative to their land-based trunk lines, we could possibly get bigger contracts later and supplant them. And when you have a battle like that, it’s usually the legacy company that wins out.” In the early days of the project, Brown expected to get support from the World Bank and other funding organizations devoted to the welfare of the developing world, but that never worked out either, primarily because the potential African partners couldn’t get their act together to apply for the funding. Eventually Brown and Wilkins gave up on that particular concept, but the idealistic goal of “global connectivity” for the Third World was still alive eight years later when Google announced it was backing an Internet-in-the-sky project called O3b, supposedly to provide Internet access to “the other three billion” without any connection to the Web. Ten other partners joined with Google to raise $1.2 billion for 8 MEOs and 180 LEOs. Strangely, O3b would cover only 70 percent of the world’s population.

Was it a stupid idea to do everything from space? The author says no. What is stupid is cell towers:

The bottom line is everything about a world using satellite phones would have been safer and probably cheaper as well. The average consumer believes that cell service is much more pervasive than it really is. Only about 12 percent of Earth will ever be covered by terrestrial cell towers. The five million towers that exist today were erected at a cost of about $1 trillion, and they periodically have to be replaced, moved, torn down, or altered at a huge continuing cost not just in money but in lives. (Working on cell towers is among the most dangerous construction jobs in the world.)

Maybe the author is correct if the goal is voice service, but can it be true in our world of data? Would it truly be possible to provide LTE-speed Internet service from satellites to 1+ billion handsets? With a cell tower the user is just a short radio hop from the physical fiber network.

If you like Airbus you’ll love the next-generation Iridium and its higher data bandwidth:

Thanks to the enthusiasm of the French aerospace industry—including some of the same men who once tried to sabotage Iridium—the second-generation “Iridium NEXT” system will be built in Toulouse, then launched on Falcon 9 rockets provided by Elon Musk’s SpaceX. The satellites will fly into perpetuity.

A consortium of six French banks, headed by Société Générale, were willing to set up a $1.8 billion credit facility for Iridium’s next generation, and that credit facility was guaranteed by COFACE up to 95 percent. The rest of the eventual $2.9 billion cost of the new eighty-one-satellite system would be covered by Iridium cash flow, which was still growing with each passing year, and by secondary payloads.

You might want to wear a hardhat for the next year or so:

In 2017 Iridium would begin de-orbit of the Chandler satellites—and yes, Public Law 85-804 would still be in effect. The annual cost of running Iridium NEXT was expected to be only about $35 million a year, meaning that finally, after twenty years, it would be a “mature” system. In fact, Iridium after 2018 was expected to be such a cash cow that a third generation around the year 2033 was projected to be funded entirely out of the company’s cash flow.

Pilots take note. The new Iridium satellites will be the basis of air traffic control in much of the world, eliminating the need for radar:

But the most impressive move Desch made wasn’t the rocket but the creation of his own secondary payload in the form of a subsidiary called Aireon. Aireon would be the first truly global air-traffic-control system. Most people don’t realize that airplanes fly out of range of traffic controllers all the time, especially over the heavily traveled North Atlantic, where planes have to be kept five miles apart in all directions until they return to the grid. There are even huge landmasses, notably in Africa, where radar systems are spaced so far apart that airplanes often can’t be seen by controllers. Much of the world was shocked in March 2014 when a Boeing 777 operated by Malaysia Airlines apparently disappeared off the face of the earth, but it wasn’t the first time that had happened since Amelia Earhart’s ill-fated flight in 1937. (The doomed Malaysian airliner was using Inmarsat.) The Aireon system would make it possible for aircraft to be tracked 100 percent of the time, which means planes could fly much closer together, travel times could be decreased, airports could handle more traffic at peak times, search-and-rescue efforts could begin immediately, and billions in fuel costs could be saved.

This time, to make sure the system would be adopted, Desch sold off an equity position in Aireon to NAV CANADA, the world’s busiest air-traffic-control system, covering eighteen countries, and, more to the point, a fully privatized system, meaning CEO John Crichton could work with Iridium without all the interminable delays Desch faced at the FAA. Desch then sold off positions to ENAV (the Italian agency that controls much of Europe’s airspace), the Irish Aviation Authority (which controls much of the North Atlantic), and Naviair (the Danish agency that controls much of the Arctic).

I think some of the above is incorrect, by the way. As far as I know, Nav Canada handles just Canada. The way that this will work is that the satellites will serve as ADS-B stations, receiving GPS-based position reports from ADS-B transponders in aircraft. ADS-B goes back to the 1990s and was supposed to be fully implemented by 2020 but regulatory hassles seem likely to delay compliance until 2025 or 2028 depending on the type of aircraft. The circuitry costs just a few dollars but the regulatory compliance paperwork can be over $100,000 per airplane.

Young people take note. Nearly all of the Clinton Administration bureaucrats with whom Colussy was dealing became highly paid lobbyists, employees of government contractors, or executives at companies dependent on favorable government regulation. These new jobs were started, in some cases, within a few weeks of Bill Clinton’s departure from the White House:

Rudy deLeon left the Pentagon in 2001 to take a job with Boeing.

Bill Cohen continued to build the Cohen Group into one of the most powerful international lobbying firms in Washington and a first stop for many senior-level undersecretaries, generals, and admirals when leaving government, including retired Lieutenant General Harry D. Raduege Jr., former head of DISA, who turned out to be a three-star, not a two-star. Cohen developed a special expertise in companies doing business in the People’s Republic of China, where he established two branch offices. Janet Langhart Cohen published her memoir, From Rage to Reason: My Life in Two Americas, in 2004, changing her designation from “First Lady of the USO” to “First Lady of the Pentagon.” Three years later the Cohens collaborated on Love in Black and White, their joint memoir, which includes reflections on race and religion. The Cohen Group remains on the Iridium payroll.

Kathy Brown left the FCC to work at the Washington offices of the prestigious law firm Wilmer Cutler Pickering Hale & Dorr, then joined Verizon as Senior Vice President/Public Policy Development and Corporate Responsibility.

Investors take note:

In the Iridium bankruptcy case, Judge Blackshear ruled that the creditors could go after Motorola separately from Iridium because, in his opinion, Motorola controlled much of what Iridium did. Blackshear retired in 2005, before that case was complete, so it was passed on to Judge James M. Peck, who presided over the testimony of fifty-two witnesses during fifty trial days stretched over eight months, admitting 866 exhibits into evidence—and that was only the first of what was expected to be two phases of litigation. Since all parties were exhausted by then, they came to the bargaining table, and Peck approved a final settlement in May 2008. The bondholders received $16 million from Motorola—one penny per dollar of what they had loaned to Iridium.

[regarding the equity shareholders who lost $billions] The most remarkable Motorola case—one for the law books—occurred in Washington, D.C., where eighteen class-action suits brought by plaintiffs who had bought Iridium stocks and bonds were consolidated in the court of Judge Thomas Penfield Jackson. Jackson was best known for calling Microsoft “stubborn mules who should be walloped with a two-by-four” after ruling against Bill Gates in the government’s landmark antitrust case, and he was equally withering toward Motorola. After Jackson’s retirement in 2004, the cases passed to visiting judge Nanette K. Laughrey from Missouri, and she surprisingly denied all Motorola’s motions to dismiss. Cases of this sort almost never reach trial because the standard of proof is so difficult. The plaintiff has to prove that the seller of stock either spoke untruthfully or said things that were technically true that would lead to an untrue conclusion, and that those false or reckless statements directly caused people to buy the stocks or bonds. That’s why only thirteen class-action shareholder cases in history have ever gone to trial. Iridium was one of them. After nine years of litigation, the production of 1.5 million documents, and what one lawyer called a “five-ring circus” of depositions, the parties finally settled in October 2008. Motorola was ordered to pay $20 million. [former CEO] Ed Staiano and [former CFO] Roy Grant were ordered to pay $14.89 million. And five underwriters of Iridium stocks were ordered to pay a total of $8.25 million—Merrill Lynch, Goldman Sachs, NationsBanc, Salomon Smith Barney, and SoundView Technology Group.

It is a little difficult to believe that any company could be as villainous as Motorola is portrayed. But quite a few American enterprises have exceeded our expectations for villainy.

I strongly recommend the book.

More: Read Eccentric Orbits: The Iridium Story,