You don’t have to look very far these days for evidence that banks and investors are keen to hedge their exposure to explosive borrowing on the part of mega-cap US tech companies.

“Keen” might not be the best word, though. As discussed here, the burgeoning market for, to use the most obvious example, CDS on Mag7 hyper-scalers isn’t so much a function of “eagerness,” per se, as it is prudence — a necessity in light of these firms’ growing footprint in the high grade corporate debt market.

Between them, Alphabet, Amazon, Meta and Oracle have sold $125 billion of debt since September — more than that if you count Alphabet’s sterling and franc offerings. That’s a lot in such a compressed window, and there’s more AI-related supply to come.

A direct result of that borrowing is an fivefold increase in the net notional outstanding of CDS on hyper-scaler borrowers. That’s a $10 billion market now.

If that’s a little too esoteric for you, the hedging’s also evident in a “simpler” metric: Short interest on the largest investment grade credit ETF.

There’s the chart. It’s from JPMorgan’s Nikolaos Panigirtzoglou. The red annotation’s mine. Note that the y-axes are different to account for generally higher short interest on HYG (the junk ETF).

What’s remarkable beyond the rapidity of the upward inflection in LQD short interest (clearly a function of the same hyper-scaler borrowing that’s created a market for previously non-existent CDS on names like Meta) is that short interest for the investment grade retail credit product is now nearly on par with short interest for the high yield ETF (22% versus 29%).

“The aversion to US AI stocks [has] spilled over to credit, with short interest on LQD rising sharply YTD [and] by more than HYG,” Panigirtzoglou remarked. “This is because much of the newly-issued AI debt that caused indigestion in credit markets belongs to the high-grade universe.”

On the bright side, elevated short interest can be a contrarian indicator in the event it gets squeezed.