Brookfield Business Partners trades legacy assets for AI-powered industrials
A $1.2B capital-recycling program, two new platform buys, and a deeper bet on OpenAI's deployment arm reset the portfolio for a more resilient, tech-inflected second half.
BBU · Earnings Call · 2026-07-31
Strategic pivot: capital recycling and new legs
Brookfield Business Partners' Q2 2026 report is less about the headline numbers and more about the shape of the business being built. The company generated $587 million in adjusted EBITDA, and same-store EBITDA rose roughly 5% to 6% across segments. But the real signal is the capital being recycled: “Multiplex marks one of the last significant legacy assets left on our balance sheet from the spin-out.” — Anuj Ranjan, Chief Executive Officer · 2026-07-31 That sale, at ~$650 million, closes out an era. In its place, management has committed over $300 million to two market-leading businesses—World Freight Company (WFC) and Gregg Distributors—and closed a strategic investment in the OpenAI Deployment Company (DeployCo).
The capital recycling engine is clearly the cornerstone. In the first six months, BBU generated $1.2 billion in proceeds, well on its way to the $2 billion target set at Investor Day. CFO Jaspreet Dehl said, “we generated $1.2 billion in proceeds from asset sales and distributions” — Jaspreet Dehl, Executive · 2026-07-31—and the pace is meant to accelerate. The company is also leaning harder into buybacks: more than $300 million repurchased at ~50% discount to NAV, with an additional $150 million allocated from the latest recycling activity.
Today, our balance sheet is as strong as it has ever been.
The new acquisitions fit the playbook of buying acquire 2 market leading businesses with recurring demand and operational levers. Adrian Letts described WFC as the world's largest general sales and service agent for air freight, with an asset-light model and a fragmented market ripe for consolidation. Gregg Distributors, similarly, is a high-touch MRO distributor. Both are being bought at roughly 9–11x EBITDA, squarely in the range BBU has historically used.
Sagen's loss ratio and CDK churn: normalization is a two-way street
Not everything is getting better. The residential mortgage insurer, Sagen, is seeing its loss ratio climb to 17%—still within the long-term 15–25% range but a clear departure from the single-digit ratios of recent years. Jaspreet Dehl said, “The loss ratios ... have gone up and they're ticking towards the more normalized levels” — Jaspreet Dehl, Executive · 2026-07-31, attributing the move to lower home prices and reduced cure rates. That's a recurring theme from prior calls; in Q1 2026, Bart Dziarski had noted a 12% loss ratio, and Stuart Levings framed it as a return to normality: “Our long-run pricing loss ratio is in the 15% to 20% range” — Stuart Levings, CEO · 2026-05-08. The honesty about cure rates falling is important—it means the company is no longer underwriting in a tailwind. Still, Sagen's core underwriting activity remains strong, and management expects the addressable market to be structurally larger.
CDK, the dealer software business, continues to be the other weak spot. Elevated churn is being offset by cost optimization and annual price increases, but the technology modernization is consuming capital. Analysts pushed for reassurance; management reiterated that liquidity is strong and that CDK remains a stable business. The elevated churn keyword has been prominent for the company across recent quarters, and it's not yet clear when it will break favorably.
AI as a compounding tool, not just a tech story
The most forward-looking piece is DeployCo. Originally committed at $150 million, BBU ended up investing $100 million, syndicating the rest while retaining the strategic partnership. Anuj Ranjan said, “It's going very well so far ... having access to the talent that is actually quite limited out there to actually deploy these solutions in a customized fashion” — Anuj Ranjan, Chief Executive Officer · 2026-07-31. The company is applying AI across its portfolio—sensors and machine learning at Clarios, predictive blending at Chemelex, and back-office automation at WFC and Gregg. This is a AI value creation office in action, and it's a differentiator versus many industrial conglomerates that are still in the PowerPoint stage.
At Clarios, the battery maker, the AI-driven order fulfillment and inventory optimization is already producing millions in savings and a 14% improvement in service performance. The company repaid $500 million of debt in the quarter, freeing up cash for further investment. It's a reminder that BBU's mature industrial businesses are not just steady cash generators—they are becoming laboratories for AI-led margin expansion.
What changed and why it matters
The market is increasingly rewarding resilience, and BBU is positioning itself accordingly. The sale of Multiplex removes a cyclical construction legacy; the acquisitions of WFC and Gregg add mission-critical, recurring-revenue businesses; and the DeployCo partnership brings the capability to apply AI across all of them. The numbers are credible: same-store EBITDA up ~5–6%, adjusted EFO up substantially, and a $2.8 billion pro forma liquidity pool. The buyback at a 50% discount to NAV is a clear signal that management believes the stock is undervalued.
What hasn't changed is the company's willingness to discuss the difficult parts. Sagen's loss ratio and CDK's churn are laid out plainly. That candor, combined with the accelerated recycling and the AI bet, makes this quarter a genuine inflection point. The question now is whether the operational improvements in the new and existing businesses can outrun the normalized loan losses and continued technology spend—and whether the capital recycling engine will keep funding the growth.
This is a company that is deliberately shrinking its legacy complexity while doubling down on its core thesis: buy good businesses, make them better with data and AI, and recycle capital when the valuation makes sense. For now, the evidence suggests the playbook is working.