GS Lev Fin Weekend Reading and Observations the divides of summer

Just about everything worked last week. Constructive geopolitical headlines, strong earnings, and continued hyperscaler strength carried the S&P 500 to its 26th record close of the year, while a weak jobs report pulled rate expectations lower13. Both the S&P and Nasdaq registered their largest weekly gains since April, and credit came along for the ride – a welcome turn after a July in which HY posted its first monthly loss since March and IG logged its worst July in nearly a quarter-century11,12.

IG finished marginally tighter – an impressive feat given the $80.2bn that priced across 29 deals – the third-largest weekly volume YTD10. More importantly, the tone fueled outperformance in hyperscaler debt with our 30y Hyperscaler basket wrapping Friday at +159, in from a peak of +172 on July 28.

As we enter the most popular vacation week of the summer, everything working at the index level is the easy part. Underneath it, leveraged finance investors face three divides – and the way they resolve will likely dictate the tone into the fall.

HY posted its best week in nearly three months as muted primary activity – just 5 borrowers printed deals totaling $3.3bn, none of it AI – pushed spreads back near their local tights4.

That reprieve didn’t change the picture underneath the surface. 16 of the 23 data center JV deals still trade wide of their new-issue yields, and only 3 of this year’s tranches trade tighter than where they initially came (h/t GIR’s Shamshad Ali).

The 16 HY deals, each carrying at least one BB rating, currently trade at an average spread of +377 vs. +154 for the median HY BB. AI now accounts for 18.4% of USD HY issuance YTD, and with another $10-15bn potentially coming pre-Labor Day, the question is whether the rest of the asset class reprices toward the new, wider cohort – or whether the cohort comes to it.

While HY struggled in July, the loan market quietly delivered its best month since April, rising 0.79%15. Previously shunned sectors – notably software and insurance – rebounded, tapping into demand for discounted risk and an A&E market that offered an outlet for short- dated paper. 9 of the 10 biggest gainers on the month were software companies wrestling with $40bn of 2028 maturities15.

Heading into September, a ~$65bn forward calendar – nearly 70% of which is initially indicated in loan form – suggests a pickup in supply could mute the floating-rate space’s near-term prospects21. Consider the other side, though:

• Of the AI leveraged finance issuance this year, more than 3x as much has come in HY ($36.6bn) vs in loan-form ($10.65bn);

• Measurable loan demand has exceeded net supply by $65.1bn YTD15; and

• Triple-C loans continue to lag with negative market-value returns for three straight months, but they make up just 6% of the broader index15

Taken together, the case for favoring floating-rate paper into early fall is a credible one, but…

DIVIDE 3: “good” loans vs. “bad”

As is always the case, the devil is in the details. Yes, triple-C loans are just 6% of the market, yet 68% of deeply distressed sub-IG tranches (defined as yields > 20%) are…loans16. The concern isn’t only triple-Cs but rather incremental liability management exercises (LMEs): more than a third of companies that have executed distressed exchanges have had to restructure again7.