Lights Out: Pride, Delusion, and the Fall of General Electric

by Thomas Gryta

as technological innovation shrank the size and cost of digital sensors, they would be embedded in more and more machinery of every conceivable type. Huge streams of data could then be harvested, not just from smartphones and WiFi-enabled thermostats, but from the massive machines at the center of the developed economies around the world:

GE adapted the tech industry principles in Ries’s The Lean Startup to its industrial business with FastWorks, an internal program that would enable a conglomerate full of painstaking engineers to replicate the failure-friendly iterative research approach of major software companies.

When Immelt explained any given strategy or projected the company’s future performance, he continued to lean on athletic ideas of conquest. His competitive instincts in business drove him to such thinking. Like a football coach intent on picking up yards on the field, he liked taking market share, and one of his favorite tropes was promising that GE was poised to “win.” Buying Alstom would drive the other teams backward as GE claimed more and more of the field.

Alstom had been worth examining because it was financially imperiled; the value of pursuing it in the first place had been to capture its assets at a bargain, strip what GE wanted, and then sell the rest, like an old car, for the parts. Now the company was talking about making sure Bouygues saw a return on its investment in a company that, absent a buyer, seemed to be headed for bankruptcy or a possible government bailout. Going into the deal with that kind of logic, Smith thought, GE would be seriously overpaying on almost any price it offered.

It all came back to the way the negotiation had started. “The valuation was artificial,” a former GE executive said. “It was fixed based on Bouygues.” Once the company leaders, from Immelt and Steve Bolze on down, had decided that they needed to make the French investor whole in order to get the deal done, the secret work being done by bankers and lawyers was an exercise not in calculation but in justification: they were coming up with an explanation for the amount the company had decided they needed to pay.

GE was running out of places to cut costs in the oil and gas unit to preserve what it could of its profit margins. The cuts were getting close to the bone, and soon enough, the only remaining option would be to report losses on the business. There was only so much fat to be cut.

GE’s knowledge of the machines it made and the needs of its customers was unmatched, but that didn’t make it the most obvious candidate to construct the scaffolding of code on which the “industrial internet” would rest.

The Owen ads spoke to some of the truth of GE Digital. Notwithstanding some claims that GE was making to investors about its success in luring young programmers away from jobs at Facebook or Apple to work in San Ramon, in fact, recruiters said, the response in Silicon Valley had been not unlike Owen’s fictional brunch table. To the characters in the ads, it was much more enticing to a young programmer to work at something like the fictional “Zazzies,” putting cassava melons on the heads of kittens—or the real-world equivalent thereof—than to enter GE Digital, where the challenge was not just the coding but integrating a software operation into a wary, if not skeptical, company full of engineers.

GE Digital was on track to spend $5 billion by 2016—a massive sum even by GE standards, equivalent to about half the R&D budget for a new, clean-sheet jet engine.

Even GE and Predix had to obey laws of physics that weren’t susceptible to marketing. MRI machines (and their cousins) produce a staggering amount of data in a small amount of time. The files proved to be so big that the feats of real-time data analysis, file transfers, and machine learning that GE had envisioned happening seamlessly among hospitals and clinicians were too cumbersome to achieve quickly—if and when they could be performed at all.

Software businesses usually turn a profit by selling uniform programs that can be sold many times over once they are built, not by designing unique programs, customer by customer.

GE didn’t just pour money into Predix—it smothered the project with cash. But without a coherent strategy and well-thought-out processes, the product development path was a wasteful one. GE’s plan to move fast, produce a viable product, and then perfect it in the field got bogged down partly because of the size of the effort. GE hired armies of new employees and gave them all the resources they wanted to build its vision. It was like an auto company building an assembly plant, hiring workers, and leaving them standing on the production line, waiting for the vehicle to be designed.

GE also planned to build its own data centers, from poured concrete foundations to roofs. The idea was to have a GE-owned and ‑operated cloud for its customers’ data, but building such an operation from the ground up would be hopelessly slow to achieve and wildly expensive. Besides, companies like Amazon and Microsoft were already pouring billions into providing just such services to other businesses. Why would GE, arriving late to the business, try to duplicate their offerings?

the deal concessions GE was being asked to make had now grown so large that GE’s original logic for the transaction no longer made sense. The cost to close the deal seemed to exceed the overall benefit to the company. And that, they realized, would trigger a breakup provision in the sale agreement, allowing GE to back out of what they increasingly viewed as an albatross of a deal.

We might be able to pull the plug on this thing. The other executive was firm: “This is Jeff’s deal. We’re not backing away.”

the absence of robust opposition also pointed to the broader problem, long cultivated and growing into a quiet crisis within the company, of real candor and self-awareness. For all its vaunted levels of management and procedures of accountability, when it had come time for lower levels of management to stand up to the ultimate boss, Immelt, and tell him that his legacy play wasn’t going to work—and in fact, had been a clumsy mistake all along—no one was willing to do so.

Trian’s endorsement was the stamp of approval that Immelt thought would help others realize the full legitimacy of GE’s expected turnaround.

GE was selling upgrades to existing customers on the promise that it meant less maintenance and more efficiency. The problem was that the future maintenance would also be GE business, so in effect the company was making a sale today that would eliminate work to be done (and billed) tomorrow. In finance parlance, GE was pulling sales forward. It was agreeing to pay Tuesday for a hamburger today.

Unlike Immelt, Kaeser wasn’t sold on the conglomerate model and was moving Siemens in the opposite direction and breaking up the company, although he would never describe it in that way. But the outcome was the same. Siemens was slowly but surely separating its businesses and keeping a stake in the new entities under a strategy Kaeser called “Fleet of Ships.” When it came to the power market that Bolze saw as ripe for harvesting, Kaeser conceded that there would be growth in the very large natural gas turbines that Bolze was touting. But he gave an ominous warning. “So there will be growth . . . But at a cost, because every order is very, very competitive,” he said.

When Immelt’s disappointment soured the mood, the tone changed. “Your people,” Immelt would say, “don’t want it bad enough.” Executives would stretch far to avoid that tone. And that is exactly what they did in Power.

Throughout 2016, Power teams combed through the division’s portfolio of service contracts, each representing payments from power generators to maintain the turbines that GE had sold them. One strategy they often used was to offer discounted turbine upgrades to customers in exchange for extending the length of their contracts as far out as 2050. Such long contracts allowed for huge payoffs—paper profits—once the assumptions were adjusted. Executives also scoured existing contracts for ways to change underlying assumptions. The entire book of GE’s service contracts was reviewed at least once a year. These reviews produced profits that GE could use to hit targets for Wall Street, but they were really future profits, produced by accounting adjustments alone. There was no actual new cash coming in. Such gaps between reported income and cash flow usually close pretty quickly, because they represent a simple timing issue—a bill wasn’t paid, or cycles didn’t match up to quarters. When such gaps are frequent and not closed quickly, it can be a red flag to investors. They see a telltale sign of aggressive accounting, if not fraud, when a company records a lot of profits but no cash to back them up. Anyone can run up a credit card, but someone has to pay that bill eventually. While contract accounting is complicated, investors still expect to see real money coming in the door.

Garden made it clear that if GE’s performance didn’t improve, Trian would ask for a seat on the board.

company lifers like Bornstein were sure to their bones that they knew better how to manage a complex beast like GE than any shortsighted activist. Bornstein was also deeply wary of the tactics that Garden and Peltz wanted GE to use to goose the stock price, especially taking on even more debt to buy back more shares. “If they want to do that,” he muttered at one point, “they’ll have to find a new CFO.” But Trian was willing to stage a public battle, which GE wanted to avoid. Even if Bornstein was up for the challenge, Immelt wouldn’t have it. An activist fight could send Immelt into retirement, leaving him on the outside while the shoots of GE’s rebirth were just breaking through the soil.

Immelt embraced the situational pragmatism of the high-powered. His priorities were the company’s—deregulatory and laissez-faire—but ultimately mutable, in response to the needs of the moment, and contested only when the bottom line was truly threatened. That could lead to unexpected divergence from stereotype. For instance, when President George W. Bush rolled out his tax cuts, Immelt told one investor that the plan didn’t make sense. “Why cut my taxes?” he said, referring to his wealth. “Why not let the poorer people have it and spend it?”

Amid the turmoil, and despite the pressures to perform, GE was still hunting for big deals. A team in the Aviation division had worked with bankers to put together a proposal to buy aerospace rival Rockwell Collins in late 2016. The deal pitch, worth more than $15 billion, reached Immelt in early 2017. He scuttled it, to the frustration of those who saw such a deal as being far more lucrative than Alstom. But Immelt knew he was being watched and was determined not to drop the ball. He couldn’t afford Trian, or others, seeing yet another example of him seeming to bite off more than he could chew. Instead, GE kept repurchasing stock, spending more than $3 billion in the first four months of 2017.

For much of his career, the famously globetrotting chief executive often had an empty business jet follow his GE-owned Bombardier or Gulfstream to far-flung destinations, just in case there was a mechanical issue that could lead to delays. The spare planes started flying early in Immelt’s tenure and continued to follow him in his travels into the months before his resignation. It was an unprecedented and highly unusual practice. Even heads of state didn’t get such treatment, not because they were less important, but because it just wasn’t practical. In case of emergency, there would always be other jets available on the ground.

Bolze and his team had let inventory pile up, hoping the market would turn around and leave GE in a prime position to sell scores of turbines. But the power market was going in the other direction, and GE now had a lot of cash tied up in that inventory. The biggest problem was that a miscalculation on inventory couldn’t be easily or quickly reversed. GE, like its customers and its suppliers, planned its manufacturing out months in advance. Even now, long after the folly of Bolze’s investments was clear, it would take even longer to unwind the massive oversupply of turbines and parts that Power had accumulated.

A picture of the company started to come together. The operations weren’t in the shape they needed to be. Few had doubted the decision to jettison GE Capital, but the plan to replace its earnings had fallen flat. Industrial growth wasn’t coming fast enough, and the company had blown billions on share repurchases that now seemed worthless. During Immelt’s tenure, GE spent well over $100 billion buying up its own stock, much of it at a price far above where it traded now. Without adjusting its sacred and expensive dividend, the company had gutted its ability to generate cash.

Next came Jeff Bornstein, who many had once thought would be CEO someday. He had been a counterbalance to Immelt and was generally respected by Wall Street. Many in the room felt comforted that he had committed to GE and to helping Flannery navigate the job. Now Bornstein launched into an exhortation: Run the company like you own it. Be the leaders General Electric bred you to be. You should all be accountable for every prediction made and every target missed. “I love this company,” he said. Then he stopped and took a breath, deep and halting. He started up, and stopped again. Jeff Bornstein, the muscled, nicotine-gum-chomping, weightlifting, hard-driving CFO, was . . . crying. The reaction in the room was confused and inconsistent. Some thanked Bornstein for his passion, while others left the bizarre scene wary. If this guy was fighting back tears, something must be seriously wrong. The group shuffled out in shock and confusion. Flannery and Bornstein had hammered away at some of GE’s intractable bad habits, but there had been no bombshells revealed. Still, stress and strain had reduced one of their toughest tough guys to tears. The moment would linger in the minds of GE people for years—a revelation, in retrospect. That was the moment many of them realized GE was taking on water, and that no one knew how bad it was going to get.

GE named Trian’s Ed Garden to its board, giving the activist a direct say in decisions and access to its financial details. The move, stemming from GE’s failure to hit the agreed cost-cutting and performance targets, was the same deal hatched in Bornstein’s Back Bay townhouse months before. GE’s failure to hit its targets had finally let a wolf in the door.

If the shares fall, buying back shares is little different than simply setting the cash on fire. Paying a dollar for something that will be worth 80 cents in the future isn’t a great way to spend someone else’s money. Jeff Immelt will not be remembered for wisely deciding how to spend GE’s cash. Buybacks were a regular fixture under Immelt, who spent more than $108 billion on them after 2004, when the SEC required companies to disclose their practices. At the end of 2018, GE’s entire market value was $67 billion. In Immelt’s last eighteen months as CEO, he spent almost $26 billion in cash on repurchases even as the stock fought to stay near $30 during that period. Just fifteen months later, it had dropped below $10.

Flannery insisted on transparency and honesty and tried to communicate that to his employees in his weekly iPhone-created video message, when he openly answered their questions. His tone and outlook on management were pragmatic. GE had the tools and would fix itself, he told employees, but it wouldn’t be easy or enjoyable. Under Flannery, the message reverberating around the headquarters and the entire company was: no more success theater. In other words, no more pretending that broken software was working. No more making deals at any cost. No more pretending that vague concepts and metrics had value. No more sticking to the marketing story. No more infallibility. No more mythology.

As he tells it, Immelt asked for a review of the holding and the potential need to increase reserves as the portfolio showed that premiums were going to cover expected payouts. The preliminary estimate of the needed reserve fund, he said, was $2 billion to $3 billion. That was a lot of money, but not in Immelt’s GE, with its persistent view that such small developments were almost irrelevant because of GE’s size and ability to offset such pressures. Remember, $1 billion can be either a great deal of money or immaterial depending on whether it’s a sales contract or a noncash adjustment. In actuality, no one had looked closely at the policy details for years and no one knew what assumptions were used. When auditors took them apart and rebuilt models with more conservative assumptions, the outcome was disastrous. GE needed $15 billion in cash to cover its liability.

The core problem was that GE had made some bad decisions in reinsuring the long-term care policies. Once it was clear that the company had amassed a collection of some of the worst insurance assets possible, there was no chance it could find a buyer. A Justice Department investigation of the holdings was also making the asset impossible to sell. Meanwhile, as former employees described it, lax oversight and buried risks would keep the company from realizing the worsening results over the years and booking bigger reserves. GE has denied allegations of fraud. As of early 2018, GE was reinsuring about 300,000 long-term care policies, about 4 percent of all such policies in the industry. Incoming premiums still weren’t covering payouts.

GE’s board of directors was unquestionably weakened from having the CEO as the chairman of the board. Many companies still cling to this structure, but it openly violates basic corporate governance principles. The CEO shouldn’t also be in charge of the committee that employs him. Some GE directors would challenge Immelt, but that usually was fruitless and could land them in the penalty box, or sometimes off the board entirely.

In naming Culp to the board, Flannery was fully aware of the possible future that appointment might bring about. Before Culp joined the board in April, an adviser had warned Flannery that Culp would be the man to replace him atop GE if things soured. He didn’t care, Flannery said. He just needed the best people to help him right the listing ship.

Immelt often lamented the hand he was dealt, the external catastrophes he had to navigate, but there had been one solid truth in his repetitive invocation of the past—GE had weathered a lot of storms with a business model that many thought was obsolete.

The board had had enough. They decided to hold Flannery accountable for the latest raft of bad news. After sixteen years of Immelt, Flannery thought that he had more time to turn the ship around, but when he looked for support from the board, there was none there. By the end of the weekend, Flannery was out, fired after fourteen months, the shortest stay at the top in GE’s long history. The new boss was Larry Culp. Flannery was reeling. He had told the board that the turnaround would take time and be painful. And the board had pledged their support, then given up on him. The board in turn worried that Flannery was in over his head and knew they couldn’t risk stumbling on the execution.

Flannery revealed serious problems, and every decision he made was a hard one, affecting thousands of factory jobs and lonely retirees waiting for a dividend check. In exile, Flannery remains certain that there was no quick fix to GE’s problems, no matter how much investors and the board wanted one. So far, he hasn’t been wrong.

Some GE lifers saw Flannery as a tragic figure, taking the fall as he tried to clean up years’ worth of others’ mistakes. Flannery has told friends that he remains certain that there was no quick fix for the serious problems he unearthed. Slowed down now from the relentless pace of his glory years, Jack Welch continues to privately rage about the man he chose to succeed him. The aging CEO still likes to say he gives himself an A grade for the way he ran GE, and an F for his choice of successor. A conversation with Welch inevitably circles around to the latest news of the company that he managed into the stratosphere. Now he is livid at the collapse of the pivot, at Immelt’s buybacks and spending, at the simple failure to run Power right. Welch scolds the board for hanging on to Immelt too long. They could have replaced him long ago, Welch has argued, and had good reason to. “I’m terribly disappointed. I expected so much more,” Welch noted. “I made the best choice I thought I could make, and it didn’t turn out right.” Jack Welch doesn’t talk about the old GE coming back to life, though he hasn’t given up faith in the conglomerate model. He hopes that Larry Culp can “build a new GE” instead.

Culp has pointed to the GE culture of Welch’s days, when operations were focused and lean manufacturing honed the company’s efficiency. He is a student of the purest form of those processes, which came out of Toyota. The outsider CEO has tried to rally workers by declaring that he is one of them. But even the success of Culp’s efforts would undermine a central tenet of GE’s oldest and most precious belief: that it knew how to manage any business and could teach any of its own to do so. In the end, when General Electric most desperately needed a manager to save what was left of the company, it had to go looking somewhere else.

the path ultimately leads back to Immelt. As chairman, he was also responsible for steering the board. There is no doubt that GE’s size and complexity, which grew exponentially under Immelt, made it difficult or even impossible to manage. The CEO of a company is responsible for its daily functions and for managing its operations, however vast. The chairman guides the board, which is responsible for overseeing management and the CEO. When the board chair and CEO are the same person, the top executive is essentially his own boss. It can only get worse with time if a chairman remakes the board to his own liking. Simply put, it is terrible governance to give so much power to a single person and so little voice to shareholders.

Looking at GE through the lens of the stock price alone was simplistic. The portfolio had been as transformed under Immelt, by some measures, as it had been under Welch. Immelt was trying to make GE great again. Latour didn’t want a cute story about air conditioning. She wanted the Journal to finally write the story that she and GE’s press operation had demanded for years: a sweeping examination of Jeff Immelt’s legacy—the ways in which he had changed the company and thereby changed the trajectory of business in America and around the world.

The GE of the late twentieth century and the fitful first two decades of the twenty-first has stood for more than what it made. It has represented a capitalistic meritocracy, a locus not just of success but of a certain version of virtue—the virtue of targets made, goals surpassed, earnings earned, markets won. And it has stood for a vague but well-marketed notion that, in the unapologetic pursuit of a company’s fortunes, and one’s own, there is a certain uprightness—and a lesson for others. But GE has stood for a well-bred hubris as well. Under Immelt, the company believed that the will to hit a target could supersede the math, even when hundreds of thousands of livelihoods—those of investors, customers, and suppliers, to say nothing of workers, retirees, and their families—hung in the balance. GE became synonymous with the idea that all that stood between a company and an arbitrary $2 earnings target was drive—simply wanting it bad enough.

Where, Jeff Immelt used to ask from behind the dais in the Pit, is the guy I used to know? Your people don’t want it bad enough. This was inspirational leadership—right up until it wasn’t. Reality isn’t a tackle sled. Story isn’t a strategy. Sometimes imagination alone doesn’t work.