WhiteLine Daily｜The real red line for AI investing is credibility, not free cash flow

WhiteLine Daily: brings together the thoughts of the Wu Shuo team to deliver the most valuable information and analysis of the day for readers, capturing the trend shifts of the AI era.

Over the past two years, tech giants have mostly used their own cash flows to build AI infrastructure. Now, Alphabet has begun issuing both bonds and stocks at the same time, and Oracle is also expanding external financing.

This means the AI race is entering a new phase. What the market needs to assess is no longer just how much money a company is willing to invest, but also how high the financing cost of those investments is—and how much revenue they can ultimately generate.

01｜Earnings Preview｜Alphabet & IBM

Company budgets have shifted toward hardware—can cloud providers capture the revenue?

Alphabet and IBM will release their Q2 earnings or final results on July 22, Eastern Time. The two companies represent two different sides of the AI investment chain.

IBM has already delivered an unfavorable early answer. The company preliminarily expects Q2 revenue to be $17.2 billion, up 1% year over year. Software revenue is expected to grow 5%, consulting revenue is roughly flat, and infrastructure revenue is down 7%. IBM said that some customers have shifted quarterly capital expenditures to servers, storage, and memory to lock in scarce hardware supply ahead of time.

This indicates that enterprise AI spending is indeed increasing, but the incremental budget may not flow evenly to all technology companies. Hardware and data centers may receive orders first, while spending on traditional software and consulting could be squeezed.

The question Alphabet needs to answer is different: when enterprise budgets shift to computing power, can Google Cloud convert demand into revenue, profit, and orders. In Q1, Google Cloud revenue grew 63% to $20.0 billion, and operating profit reached about $6.6 billion. Q2 will test whether this growth can be sustained.

WhiteLine view:

The market’s easiest mistake is to equate CapEx growth with confirmation of AI demand. On July 22, the real test is whether investment and revenue can accelerate in sync. If Alphabet continues to increase investment, but cloud business profits and orders do not improve, the market will reinterpret the incremental CapEx as future depreciation and financing pressure.

02｜Core Metric｜Free Cash Flow

Free cash flow is down—this will not force Alphabet to slow down for now

Alphabet’s Q1 operating cash flow was about $45.8 billion, capital expenditures were about $35.7 billion, and free cash flow fell to about $10.1 billion. At the same time, the company still raised its 2026 capital expenditure guidance to $180 billion—$190 billion, and expects 2027 to further grow significantly.

Free cash flow itself equals operating cash flow minus capital expenditures. The faster data center construction proceeds, the lower free cash flow will be in the current period. Therefore, free cash flow can reflect cash consumption, but it cannot, on its own, determine whether a company must stop investing.

Alphabet still has rapidly growing cloud revenue and strong operating cash flow. As long as AI demand continues to exceed available computing power, the company is more likely to expand financing, reduce buybacks, or accept a near-term decline in free cash flow—rather than voluntarily give up the infrastructure competition.

WhiteLine view:

Alphabet will not actively hit the brakes in the short term due to a decline in free cash flow. Only if AI revenue, orders, and profit margins continue to lag behind depreciation and financing costs will cash flow pressure shift from an accounting outcome into a capital allocation issue.

03｜Financing Capability｜Alphabet

AI buildout starts using cash, bonds, and stocks at the same time

Alphabet’s financing approach has already changed.

In February, the company raised about $31.51 billion in the global bond market, including a rare 100-year bond. In June, it also announced an equity financing plan with a total of $84.75 billion. Alphabet has made it clear that AI infrastructure will be financed jointly through operating cash flow, debt, and equity.

This shows that AI investment can no longer rely on current cash flow alone. Tech giants are calling on the entire balance sheet and shifting part of the investment cost to creditors and new shareholders.

Alphabet still has ample credit headroom. S&P expects its credit-adjusted leverage to be about 0.2x in 2026 and about 0.3x in 2027—still clearly below the roughly 1x rating downgrade warning threshold.

However, the rating red line is not the only constraint in the market. Equity financing causes dilution, and debt financing increases interest costs. Even if the company has not approached downgrade levels, the stock and bond markets may demand higher returns in advance.

WhiteLine view:

The first hard brake on AI capital expenditures is not a decline in free cash flow, but net leverage approaching the rating red line, and issuance costs rising significantly. If Alphabet remains far from the 1x leverage level, Q2 is more likely to maintain and even further raise CapEx.

04｜Market Pricing｜Oracle

The valuation will come under pressure before the company even hits the credit red line

Oracle has already provided a stress test.

The company expects FY2027 total capital expenditures to reach up to $95 billion, of which about $70 billion will be covered by itself, with another about $20 billion—$25 billion expected to be reimbursed by customers. At the same time, Oracle plans to raise nearly $40 billion through debt and equity financing.

Although remaining performance obligations rise to $638 billion and orders remain strong, its stock price subsequently fell by as much as 12% at one point. Investors are concerned whether these orders can generate enough profit to cover data center depreciation, interest expense, and the incremental cost of new financing.

This shows that the capital market will not wait until the company truly hits the rating red line to reprice. When capital expenditures, liabilities, and financing costs rise in sync but the pace at which revenue and profits are realized is insufficient, valuation will come under pressure first.

WhiteLine view:

The truly dangerous combination is not “high CapEx, low free cash flow,” but rather CapEx continuing to grow while net leverage, CDS, and credit spreads deteriorate, yet AI revenue and profit fail to accelerate in parallel. If the credit market raises financing prices first, semiconductor and data center order momentum may still persist, but valuations will start to diverge first.

Today’s main thread

The market is moving from pricing “how long AI revenue can keep growing” to computing “how much profit each dollar of CapEx can ultimately generate, and how much the company can borrow to keep investing.” Alphabet is still far from the credit red line, so investment is likely to keep accelerating. The real risk will only appear when net leverage, financing costs, and AI returns deteriorate at the same time.