AI bond issuance suddenly "hit the brakes" in September, Morgan Stanley judges: it may make a comeback in the fourth quarter.

Title context: AI bond issuance suddenly "hit the brakes" in September, Morgan Stanley judges: it may make a comeback in the fourth quarter. Text: In September, global AI-related bond issuance was only about $23 billion, the second-lowest month of the year. AI issuance in the U.S. investment-grade market came to a complete "zero." Morgan Stanley characterized the slowdown as a "pause rather than a retreat," and expects issuance to rebound in the fourth quarter, but not to repeat the explosive growth seen in the first half. As of the end of September, total global AI-related bond issuance this year had reached $466 billion, more than double last year's full-year total of $216 billion. In its latest report, Morgan Stanley analyzed that the September cooling was not due to deteriorating fundamentals or a shortage of capital - rather, it was the result of three overlapping factors: a large front-loading of earlier issuance, regulatory and political resistance to data center construction, and a sharp rise in interest rates forcing project-level terms to be renegotiated. Why did it cool in September? The pace of AI financing this year has been highly uneven. June set the annual peak with $113 billion in single-month issuance, followed by a month-by-month decline. High-quality hyperscale cloud providers (Google, Amazon, Meta, Microsoft) have issued about $132 billion in the investment-grade market this year, up about 25 times year over year; including Oracle and SpaceX, the six major hyperscalers have issued about $254 billion on a full-caliber basis (including non-USD currencies), heavily concentrated in the first half. Morgan Stanley noted, "After a busy summer, given the volume already completed and issuers' efforts to establish an issuance cadence, we expect near-term supply to slow." In September, U.S. investment-grade AI issuance was zero. There was slight activity in non-USD markets, with Amazon issuing a 4.25 billion pound four-tranche deal. In leveraged finance, SoftBank Group issued about $10 billion of high-yield bonds, AI data center operator Crusoe issued about $500 million of loans, and total September leveraged finance volume was about $14 billion. The second factor is regulatory and political constraints. New data center construction faces permits, power supply, and political resistance - Morgan Stanley previously summarized these as the "three Ps": people, power, and politics. These constraints are shifting from potential risks to real restrictions. The third factor is interest rates. The sharp rise in yields means project-level transactions may need to be renegotiated, and lower-rated issuers and project finance are more sensitive to funding costs. Fourth-quarter outlook: hyperscalers will return, but the pace will differ from last year Morgan Stanley expects fourth-quarter issuance to exceed September levels, but not to cluster into an explosive burst like last year - last year, most supply appeared in the fourth quarter. High-quality hyperscale cloud providers are expected to return to the U.S. investment-grade market while continuing to raise funds in non-USD markets. Non-USD hyperscale issuance this year has reached about $72 billion, close to one-third of total global hyperscale issuance, with currencies expanding from last year's single USD and EUR mix to CAD, GBP, CHF, AUD, and JPY. Morgan Stanley expects euro-denominated deals to also enter the market in the fourth quarter. The pace of data center project financing is harder to predict. Morgan Stanley cut its forecast for ABS and CMBS issuance this year to $25-30 billion, implying about $5-10 billion of remaining supply space in the fourth quarter. An important driver of the fourth-quarter rebound is the 2027 capital expenditure outlook. Morgan Stanley forecasts combined 2027 capital expenditure for the six major hyperscalers at about $1.4 trillion, significantly above the market consensus of about $1-1.1 trillion. Third-quarter earnings reports (expected to be released gradually in October) may bring another round of capex upward revisions, directly driving financing demand. Morgan Stanley believes high-quality hyperscalers are "largely insensitive" to high interest rates - the return on invested capital (ROIC) outlook is strong, and debt financing costs remain below equity financing costs. These providers have overall leverage of only 1.3 times (net leverage 0.4 times), a cash/debt ratio of 132%, a median credit rating of AA-, and ample balance sheet capacity. The macro test is greater than the supply shock Morgan Stanley emphasized that for the investment-grade credit market, "the macro environment - not excess supply - is the biggest test before year-end." Investment-grade bonds' total return has fallen 3% year to date, with quarterly returns down 4%, a drawdown that has reached a level that could trigger large-scale redemptions from mutual funds and ETFs. The 10-year U.S. Treasury yield has risen from 1% to above 5%, and multiple interest-rate repricings over the past five years have repeatedly hit credit spreads and fund flows. However, Morgan Stanley believes the credit market as a whole can absorb higher yields, because nominal growth remains above 6%, corporate earnings growth is even stronger, and the rise in rates is supported by solid fundamentals. The bank's economists expect the Federal Reserve to raise rates only twice more, more dovish than market pricing of close to four hikes. On credit differentiation, Morgan Stanley reiterated a preference for secured assets (about 30% of AI-related debt). Spread volatility in secured data center bonds is mainly driven by construction risk - permit delays, power supply, and lease uncertainty - while ABS and CMBS assets already in stable operation are less affected by such shocks. Chip financing's asset characteristics allow faster cash flow generation and amortization, but publicly traded products remain limited at present. Looking beyond the fourth quarter, Morgan Stanley believes most of the hyperscaler spread compression trade may already be over: issuance cadence is becoming more predictable, capital expenditure is shifting toward shorter-duration assets such as chips, and the strong returns on AI investment continue to validate the soundness of the spending rationale. This article is reprinted from "Wall Street See", author: Gao Zhimou; GMTEight editor: Yan Wencai.