Why GE Broke Up: The End of America's Model Conglomerate

GE broke up because its structure stopped working. A finance arm funded with short-term money, a power business bought at the top of its cycle and a dividend sized for the best years left the conglomerate unable to survive a downturn on its own terms. After cutting its dividend to a penny in 2018 and running down GE Capital, GE split into three focused companies: GE HealthCare in January 2023, GE Vernova in April 2024, and GE Aerospace, which kept the name.

For much of the twentieth century, General Electric was the company every industrial board studied. On 30 October 2018 it reported a goodwill impairment of about $22 billion on its power business and cut its quarterly dividend to one cent a share. John Flannery had already halved it in November 2017. Four weeks earlier the board had replaced Flannery with Larry Culp, the former chief executive of Danaher. For a company formed in the 1892 merger of Edison General Electric and Thomson-Houston, a penny a share was a public admission that its cash could no longer be spent on shareholders.

Why did GE break up?

The short answer is that GE had become a bet on the fourth factor in the book's production equation: capital matched to the cycle. Its divisions still needed products worth making, process capability and full plants, and several had them. Its jet engines in particular were among the best businesses in industry. What the parent added was an internal capital market and a strong balance sheet that lowered the funding cost of its finance arm. When that financial engineering failed, the structure had nothing else to justify it.

A conglomerate makes four claims. Headquarters allocates capital better than outside investors. A shared management system lifts every unit. Businesses with different cycles smooth group earnings. And a strong industrial balance sheet cheapens funding for a finance arm. GE leaned hardest on the second and fourth. The second worked for a long time. The fourth is what failed.

What went wrong at GE Capital?

From 1981, under Jack Welch, GE was run on the demand that each business be "number one or number two" in its market, a talent system, and a finance arm that grew into one of the largest non-bank lenders in the world. That combination produced two decades of steadily rising reported earnings. It also taught investors and managers that GE's manufacturing businesses could be valued and managed like a portfolio of financial assets.

GE Capital had borrowed short and lent long inside a company whose investors thought they owned an industrial business. It came under funding strain in the 2008–09 financial crisis. In 2015 GE decided to sell most of it. The risk of losing access to short-term funding later made cutting debt urgent, and customers who held long service contracts needed to believe GE would still exist to honor them (company account, reported).

How did the Alstom deal hurt GE?

In the same year it decided to sell most of GE Capital, GE completed its purchase of Alstom's power and grid businesses, on 2 November 2015, for about €9.7 billion net. It bought at the top of the gas-turbine cycle. Orders for new gas turbines collapsed from 2016.

The power division's long-term service agreements made it worse. These were twenty-year contracts whose margins had been booked on estimates, and the estimates turned out to be optimistic. The Securities and Exchange Commission's investigation of GE's disclosures ended in a settlement in December 2020. In the book's capital-cycles chapter, the Alstom purchase is the textbook case of the pro-cyclical pattern that destroys value: buy at the peak with plentiful cash, then cut spending in the trough when assets are cheap.

Why did the penny dividend matter more than the breakup?

Investors treat a dividend as a promise, and cutting it is read as a confession, which is why boards usually cut too little and too late. GE's cut went almost to zero, and the book argues that this, not the later split, was the decisive event.

The logic is easiest to see with round numbers. The book's stylized example, which uses invented figures rather than GE's, takes a manufacturer paying about $4.2 billion a year in dividends that enters a three-year downturn. It must also repay $4 billion of maturing debt each year and put $1 billion a year into its pension plan, with only $3 billion of spare cash.

Stylized example from Chapter 18 of The Production Equation; invented numbers, not GE's.

Halving the dividend still leaves a gap of almost $13 billion, which must be filled by selling businesses into a weak market or borrowing at the rates of a downgraded credit. Cutting to a penny closes most of the gap and turns the remaining asset sales from a forced auction into a choice. It hurt the shareholders, many of them retirees and employees, who relied on the income. But a company that cuts deeply once regains the confidence of lenders and of customers who need it to survive long service contracts.

What did Larry Culp change?

Culp treated GE's problem as operational and financial rather than strategic: too much inventory, poor cash conversion, late deliveries, and a culture that hit the plan rather than fixed the process. He sold the biopharma business to Danaher for about $21.4 billion, closed in March 2020. He shrank GE Capital to a run-off, moved decisions into the businesses and made industrial free cash flow, not adjusted earnings, the headline number.

He also went to the factory floor. Culp brought Danaher-style lean daily management to GE's plants, and GE Aerospace now calls its version FLIGHT DECK, "our proprietary lean operating model" (company). Earlier generations at GE had run Six Sigma company-wide from the mid-1990s, with Green Belts, Black Belts and tracked project savings. The shift from counted savings to daily flow on the shop floor is the shift the explainer on the Toyota Production System describes.

The pandemic of 2020 emptied the skies in the middle of the plan, and the plan survived it. Only then did GE announce the split.

How did the GE split work?

GE announced on 9 November 2021 that it would separate into three companies. GE HealthCare borrowed about $10 billion, paid most of it to its parent, and was distributed on 3 January 2023, one share for every three GE shares, trading from 4 January. GE kept 19.9% and had sold all of it by November 2024. GE HealthCare is profiled in this series' healthcare book.

GE Vernova, the energy business, was spun off on 2 April 2024, one share for every four, with little debt. What remained was GE Aerospace. The pattern is typical of US spin-offs, which are usually structured to be tax-free under section 355 of the Internal Revenue Code. The parent takes cash out through debt placed on the child, and the separation agreements decide who carries pensions and liabilities.

The case for splitting rests on the conglomerate discount: a diversified company is often worth less than the sum of focused parts. Investors specialize, analysts apply blended multiples, and the businesses need different capital policies. A medical-device business wants research and acquisitions. A gas-turbine business wants a balance sheet strong enough for twenty-year service contracts. An aircraft-engine business wants to fund decades of aftermarket. One dividend policy fits none of them.

Did the GE breakup work?

GE Aerospace is the clearest beneficiary. It delivered 1,802 LEAP engines in 2025, up 28%, and in the third quarter of 2026 reported orders of $19.0 billion, up 49%. Its backlog of more than $210 billion at mid-2026 included about $170 billion of commercial services (company). That is the engine business model described in the book's section on aftermarket services and the installed base: sell the engine cheaply and earn the return over thirty years of parts and overhauls.

The honest caveat is that GE Vernova and GE Aerospace did well after 2024 partly because their markets turned up. Air travel recovered, and demand for power equipment rose. The share of the gain due to separation itself cannot be isolated. Much of the rise in share price from any split usually arrives when it is announced, and what happens afterward depends on operations.

How GE compares with other industrial breakups

GE was not alone. Almost all of the large American industrial conglomerates split between 2015 and 2026.

Honeywell announced a three-way split on 6 February 2025, after Elliott Management disclosed a stake of more than $5 billion in November 2024. Solstice Advanced Materials was distributed on 30 October 2025 and Honeywell Aerospace on 29 June 2026. 3M spun off its health-care business as Solventum on 1 April 2024, separating a cleaner business from a parent facing settlements over PFAS and earplugs. Danaher has spun off Fortive, Envista and Veralto while continuing to buy. Illinois Tool Works separated by pruning, consolidating about 40 operating segments into seven.

Two lessons recur across these cases. Where the parent's operating system was real, as at Danaher and ITW, the company kept acquiring and separating and its margins kept rising; the system, not the diversification, was the product. Where the structure was mainly financial, as at GE, the breakup came only after the capital factor had failed. The guide to how manufacturing companies make money explains why capital is the factor that turns a cyclical downturn into a crisis, and the list of the biggest manufacturing failures shows what happens when it fails completely.

The takeaway

GE's collapse and recovery turned on capital matched to the cycle. Its jet engines were always a product worth making, and its factories in time regained their process capability, but a finance arm funded short, a power acquisition at the top of the turbine cycle and a dividend sized for the best years left it unable to survive a downturn. Culp rebuilt the capital factor first, with the penny dividend, debt reduction and the run-off of GE Capital, and only then separated the businesses, each with a capital structure for its own cycle. The full story, with the worked example on dividend cuts, is in Chapter 18 of The Production Equation, which you can preview on the book page.