Flex Ltd.
Analysis
Flex is splitting into an AI-infrastructure company and a manufacturing company, and each is forecast to grow faster apart than the two did together. The bill arrives first: capital spending more than doubles this year while the sales it pays for land later. How much of the FY2027 capex is already committed versus still to be spent?
- The spin-off is the story: two companies, each guided to grow faster alone.
Cloud and Power Infrastructure, the power and cooling arm for data centers, lists separately in early 2027, guided to grow 65-75% while the rest of Flex grows single digits, because customers no longer want to buy those pieces separately. What margins does the remaining business earn once the data center unit is gone?
- The non-data-center half is growing too, while consumer products shrink.
Networking gear that sits alongside data centers grew 20% last quarter and lifted guidance, and industrial is strong on warehouse automation. Consumer devices and lifestyle stay soft, and management keeps walking away from that work.
- The build is being paid for long before the sales arrive.
Capital spending more than doubles this year to fund factories customers fill only later, so Flex cut its cash target from 60% to about 40% and barely broke even on free cash in Q1. The build is now paid for by the business itself. Which segments absorb most of the capex increase this year?
- Next year's data center revenue is largely already booked. With orders locked, the job is finishing factories on time rather than finding buyers, work Flex does at scale already.
- Money already spent starts earning again next year. Management expects to win back the profit rate it gave up building capacity, so earnings grow faster than sales.
- Industrial and health care keep the manufacturing half growing. Steadier work like medical devices and warehouse robots gives the smaller company a base that does not need AI to grow.
- Customer deployments slipping would leave the new factories underused. The factories are paid for either way, so a delay leaves the company paying for capacity it cannot yet sell.
- Compute work earns less than the power products do. If the lower-profit part grows fastest, the whole company's profitability climbs more slowly than its sales.
- The remaining Flex business is guided to grow only slowly. If the steady half stays in the single digits, all the growth and all the risk sit inside one spun-off company.
| Politician | Transaction | Amount | Traded | Filed |
|---|---|---|---|---|
Donald Trump Republican · Executive | Purchase | ≥$15K | Jul 8 | Sep 22 |
Donald Trump Republican · Executive | Sale | ≥$1K | Feb 12 | May 13 |
Donald Trump Republican · Executive | Purchase | ≥$1K | Jan 12 | May 13 |
Ro Khanna Democratic · House · CA-17 | Purchase | ≥$1K | Aug 3 | Sep 4 |
Gilbert Ray Cisneros, Jr. Democratic · House · CA-31 | Sale | ≥$1K | May 12 | Jun 8 |
Congressional trading data provided by QuiverQuant
How Flex Ltd. turned $25.8B of revenue into $1.2B of operating income.
| Revenue | ||
|---|---|---|
| Flex Agility Solutions (FAS) | $14.1B | ▲ 1.08% YoY |
| Flex Reliability Solutions (FRS) | $11.7B | ▼ 6.03% YoY |
| Total revenue | $25.8B | |
| Gross profit | $2.2B | 8.4% |
| Operating income | $1.2B | 4.5% |