Why So Many Giant Companies Are Breaking Themselves Apart

The Era of the Giant Conglomerate Is Ending

June 29 20262024. Warner Bros. Discovery has announced its own separation. The era of sprawling do-it-all companies is giving way to something sharper and more focused and the reason comes down to a single idea in corporate strategy

The Conglomerate Discount

A conglomerate is a company that owns several unrelated businesses under one roof. For decades the argument was that diversification would smooth out the hurdles that a strong aerospace year would cover a weak chemicals year and that a talented core team could allocate capital across the portfolio. The problem is that the market often values the combined company for less than its parts would be worth on their own. That gap is called a conglomerate discount and it's real money. In the case of HoneywellAnalysts estimate that standalone companies could generate operating multiples 25 to 30 percent higher than what the conglomerate performs as a single company

Why the Discount Exists

Several forces create this. Investors who want pure exposure to aerospace don't want it diluted into an unrelated automation business so they pay less for the package. Analysts struggle to cover a company that spans five industries and complexity that's hard to understand tends to trade at a discount. Capital can get trapped and cash from a strong division quietly funds a weak division that should have been sold or fixed. And accountability becomes blurred because a struggling unit can hide inside a state.large and diversified results

A Worked Example: Running the Sum of the Parts

The entire breakup argument is based on a calculation so it's worth doing it once with numbers instead of adjectives. Here's a simplified conglomerate of two segments with round figures chosen for clarity

Step one value each segment as if it were independent. The aerospace division earns 8 billion EBITDA. Exclusive aerospace suppliers trade at 18 times EV/EBITDA so the segment is worth 8 times 18 or 144 billion. The automation division earns 4 billion EBITDA and its pure automation peers trade at 14 times so it is worth 4 times 14 or 56 billion

Step two add them up and it becomes part of the capital again. The total value of the company based on the sum of the parts is 144 plus 56 which is equivalent to 200 billion. We subtract the net debt from 25 billion and the parts imply equity worth 175 billion

Step three: compare with what the market really pays. Let's say the conglomerate's shares are worth 140 billion. If we add the 25 billion of net debt the market values ​​the entire company at 165 billion. Against 12 billion of combined EBITDA this is a combined multiple of 13.75 times even below the cheapest of its two segments

The fourth step and this is the step activists tend to skip subtract what the breakup costs. Two independent companies need two boards of directors two financial organizations two legal teams and two audit relationships. Let's call the permanent duplicate overheads 400 million a year. Capitalized at the same time 15 times that is 6 billion of value destroyed forever. Add the one-time costs of separation systems advisors and tax work which amounts to 2 billion

The net value unlocked is 35 billion discount less 6 billion ongoing dissynergies less 2 billion one-time costs leaving 27 billion against an equity value of 140 billion which is about 19 percent

That 19 percent is why this happens. It's also why it doesn't happen everywhere. Change pair multiples in two shifts or increase doubled overhead to a billion per year and the whole case evaporates. The breakout argument is a non-starter. It's a subtraction and the answer depends heavily on data that reasonable people will question

The GE Proof

The clearest proof is General Electric. When it split into GE Aerospace GE Vernova and GE HealthCare the combined market value of all three pieces grew about four times what the individual company was worth at its 2022 low. The same assets the same people the same products and a hell of a lot more value simply by allowing each business to stand on its own two feet and be judged on its own merits. That result is exactly what has been haunting every breakup since

The magnitude of the decline that preceded it is worth remembering because it explains why GE became the model. Under Jack Welch GE was the most valuable company in the world and the model that all conglomerates imitated. Then GE Capital the financial arm that had quietly become the profit engine nearly brought down the entire company in the 2008 crisis. Dividend cuts asset sales and restructurings followed for a decade. In June2018 GE was removed from the Dow Jones Industrial Average after more than a century as a member the last original component to disappear

The split itself spanned from a November 2021 announcement to the separation of GE HealthCare in early 2023 and the separation of GE Vernova in 2024 leaving GE Aerospace as the remaining company. Three focused businesses each with its own investor base its own comparables and nowhere to hide

The Role of Activist Investors

Many of these breakups are not management's idea. They are driven by activist investors who buy a stake do the sum of the parts and publicly campaign for a split to unlock trapped value. Honeywell's move was a consequence of pressures of exactly this kind. Activists have turned the conglomerate's discount into a repeatable playbook: find a growing company trading below the value of its parts take a stand and campaign for a split until the board relents

It pays to be clear about the incentive. Typically an activist stays for a defined period and is rewarded by re-rating which comes at the time of the announcement. Duplicate overheads and loss of the domestic capital market come over the next decade by which time the position is usually already sold. That doesn't mean the analysis is wrong. It means the person presenting it is not indifferent to the answer

The Case for Staying Together

Breakups are not free and the other side deserves a hearing. Splitting means duplicating corporate financial legal and human resources functions in new companies which raises costs. It eliminates the internal capital market that allows a parent company to finance promising projects without turning to outside investors. And a smaller more focused company has nowhere to hide during a crisis in its only industry because diversification actually reduces volatility. The honest question is whether the discount the market applies is greater than the benefit provided by thediversification and investors increasingly decide that this is the case

The internal capital market deserves more credit than it usually gets. A conglomerate can finance a promising but unproven division with the cash flow of another division without explaining it to anyone. An independent company in the same position has to raise money publicly at a price set by people who do not know the business in whatever window the market offers. This is a real cost and is invisible in a sum-of-the-parts table

Where the Breakup Thesis Breaks

I find the logic of the breakup persuasive and I think it is oversold. Four objections that I take seriously

A sum of the parts is a comparison not a valuation. It says the market values ​​these peers at these multiples. It doesn't say the segment deserves that multiple. If the aerospace division is subscale has weaker contract positions or takes on the group's pension liabilities then 18 times is an incorrect comparable and the discount you calculated is partly just a quality difference in structural disguise

GE's evidence is confusing. This is the objection I push the most because the quadruple number is cited everywhere. GE's 2022 bottom was a low point for the commercial aerospace industry specifically as air travel is still recovering from the pandemic. Much of what followed was the changing of the aerospace cycle and GE Aerospace would also have substantially rerated within the old structure. Attributing the entire gain to the spinoff requires assuming that thehypothetical conglomerate would have remained stable during an aviation recovery something no one has proven

Breakups pile up at the top. Splits are announced when stock markets are receptive and multiples are high because that's when the sum-of-the-parts gap appears widest and the spun entity can float well. By definition that's also when the comparables you're comparing against are most generous. The playbook is procyclical and procyclical strategies produce excellent results until they don't

Dow and DuPont went the other way and came back. The two merged in 2017 into DowDuPont explicitly under the logic that combining and then reorganizing would create a better set of parts and split into three companies in 2019. Back and forth enormous transaction and integration costs several years of management attention and a result that few would describe as a triumph. The same advisors can create a compelling platform to combine and separate and the fee is charged both ways

My own opinion is that the discount is real that it is less than the sum-of-the-parts headline suggests once dissynergies are honestly subtracted and that the strongest cases are those where the segments actually compete with each other for capital. That's an opinion and many people who do this professionally would put the number higher than I would

How I Would Actually Run a Sum of the Parts

If you gave me a conglomerate and asked me if it should be broken up this is the order I would work in

I would start with segment disclosures in the 10-K rather than multiples of peers because everything after that depends on whether the reported segment EBITDA is real. Corporate overhead allocation is the number to look for. A group that pushes core costs toward one segment makes the other look artificially profitable and a sum of the parts built on that split favors the answer before choosing any multiple

I would then choose comparables and write in one sentence each why the segment deserves that multiple rather than a lower one. If I can't write the sentence I use a lower multiple. Most sum-of-parts analyzes I've read fail exactly on this point and fail handsomely

You would then size the dissynergies explicitly as a permanent annual number and capitalize them with the same multiple used for the segments. Leaving them as a footnote is the most common way these analyzes overstate the premium

Finally I would like to ask the question that the table cannot answer: does the separation change the way the business is actually run? A breakup that produces two management teams with greater incentives and no cross-subsidization is doing something real. One that produces the same operations with two headquarters has only rearranged the ownership certificates

None of this is investment advice. Here's how I would structure the paper if the question landed on my desk

Why It Matters for the Role

A breakup is one of the purest corporate valuation and strategy exercises out there. Doing sum-of-the-parts analysis valuing each business based on the multiple its pure-play peers demand and comparing it to the conglomerate's true price is exactly the kind of work that decides whether a company is worth more as a whole or in parts. It's strategy and finance rolled into a single high-stakes issue

The Bottom Line

Conglomerates are breaking up because the market has decided it will pay more for businesses it can understand separately than for a portfolio it has to unravel. Honeywell in 2026 GE will end in 2024 and Warner Bros. Discovery will be next

The arithmetic is a subtraction that anyone can perform. Segments worth 144 billion and 56 billion in even multiples imply a 200 billion company versus a market paying 165 billion a difference of 35 billion. If you remove 6 billion of capitalized duplicate overhead and 2 billion of separation costs about 27 billion is actually available about 19 percent of thecapital.Move two entries and the case disappears completely

So treat the discount as real and the headline as marketing. GE's result is impressive and in part an aviation recovery in disguise: Dow and DuPont merged and split up in three years and the person showing the sum of the parts usually gets paid at the time of the announcement and not in the following decade