Carveouts Are Where Private Equity Can Still Find the Move From 5 to 12

What's New

The operational improvement now required to generate private equity returns is most accessible in corporate carveouts, because the businesses have been systematically underinvested and the opportunities are identifiable to a buyer who has done it repeatedly. Anuj Ranjan, CEO of Brookfield's Private Equity Group, argues this in a conversation on Alt Goes Mainstream. Brookfield reviews around 400 companies a year and buys four, selecting for businesses that are already market leaders and price setters but leaving money on the table. Ranjan names the constraint that defeats inexperienced buyers: a carveout requires standing up an entirely new management team rather than backing an existing one.

Why It Matters

Ranjan's framing of the return requirement, that 12 is the new 5 in earnings growth terms, puts every buyer of cleanly packaged auction assets on the other side of the argument. His claim is that a gift wrapped business sold to the highest bidder offers no obvious route to that growth, while a carveout does, provided the buyer can absorb the work. Brookfield's positioning supports the stance, with roughly 25% of every fund coming from its own balance sheet, which means the operational burden is being taken on with the firm's own capital exposed alongside clients.

Big Picture Drivers

• Selection for undeservingly unloved assets: Ranjan targets businesses that are genuinely good but have not received attention inside their parent, sometimes because they were not core and sometimes because capital expenditure was prioritised elsewhere.

• Pricing as the recurring untapped lever: Where only two companies make a product, customers have no reference point, and the incumbent often prices on cost plus inflation rather than on what a customer facing a plant shutdown would pay.

• Pattern repetition as the underwriting edge: Having done this around 300 times, Ranjan describes roughly 80% of each situation as familiar and 20% as new, which is what allows the operational opportunities to be identified before purchase.

• Patience enforced at the institutional level: Brookfield spent five years in India without transacting and entered the Middle East in 1997 without deploying until 2015, and Ranjan was promoted twice during the period he did no deals.

• Industrial businesses as the least digitised sector: Many of these companies have not implemented enterprise resource planning or customer relationship management systems and have not collected the operational data that would support dynamic pricing or predictive maintenance.

• Platform information as a sourcing and diligence input: Data from 400 portfolio companies feeds private equity decisions, and the energy platform's view on nuclear underwrote the Westinghouse acquisition when the sector was in bankruptcy and out of favour.

By The Numbers

• 400 to four: Companies reviewed each year against companies acquired, the selectivity the carveout focus requires.

• 12 is the new 5: Ranjan's shorthand for the earnings growth now needed to produce the returns a lower rate environment once delivered.

• Roughly 25%: Share of every Brookfield fund coming from the firm's own balance sheet.

• $35 billion: Brookfield's invested position in India, built after five years of watching the market without transacting, alongside $18 billion of deals in the UAE and Saudi Arabia since 2015.

• Around $80 billion: The scale of Westinghouse's recent contract with the United States government, for a business acquired out of Chapter 11 when nuclear was out of favour.

• 8 to 9 times: The earnings multiple at which Ranjan would prefer to buy a business whose management team can deploy AI without becoming the bottleneck.

Key Trends to Watch

• AI applied to industrial operations rather than technology assets: Ranjan argues the largest opportunities sit in manufacturing businesses that are not themselves subject to disruption but have invested least in technology, starting with data collection through connected devices and edge computing.

• Predictive maintenance as the first measurable application: Detecting abnormal vibration on a specific machine before failure protects uptime and defers capital expenditure, which affects both cost and revenue lines.

• Physical AI approaching an inflection without having reached it: Ranjan describes Brookfield employees training models in factories wearing goggles and gloves to digitise a physical world that is not represented on the internet, and expects rapid acceleration once the capability clicks.

• Deployment capability determining AI winners: Brookfield partnered with OpenAI on a deployment company on the argument that adoption speed, rather than model quality, separates outcomes.

Memorable Quotes

• "the biggest thing that people miss in a carveout is actually the people" Ranjan on why the operational thesis fails without a management team built from scratch.

• "we want undeservingly unloved businesses" The selection criterion, distinguishing neglect from genuine impairment.

• "we look at 400 companies a year, we only buy four" The filter rate behind the strategy.

• "AI has moved from innovation to execution" His reasoning for investing in deployment capability rather than model development.

The Wrap

The approach succeeds if the operational levers Ranjan describes, pricing power, footprint rationalisation and technology adoption, produce the earnings growth required, and if Brookfield's repetition advantage holds as more buyers move toward complex transactions for the same reason. It fails if competition for carveouts compresses entry multiples to the point that the improvement is already priced, or if management team construction proves harder to scale across four deals a year than across one. The evidence sits in realised outcomes rather than marks. The vintages being bought now will report in five to seven years, which is when the 12 from 5 claim becomes checkable.