Signal & Noise Filter
Triangulating the Macro Case for a Right Tail
Leopold’s liquidation cleared the immediate leverage risk from the semiconductor complex, and the Citadel purchase put a floor to the forced selling.
But none of this really answers the fundamental question heading into Q3, which is whether the forward growth, inflation, and earnings pictures are strong enough to justify holding risk into a market still working through an unresolved 20%+ semi correction, which these pages still view as live and in-play.
Growth
Skip Headline GDP from last week.
Q2 real GDP printed 1.5%, down from 2.1% in Q1.
Real final sales to private domestic purchasers, which is the cleaner read that strips out trade, inventories, and government, came in at 3.9%, the strongest print since Q1-2023.
Consumer spending accelerated to 3.2% annualized from 0.5% in Q1.
Nonresidential fixed investment ran at 8.4%.
Data center construction rose by 7% in June.
Every dollar-denominated measure of private demand in this economy got stronger last quarter.
Why did the headline number break bad? Net exports subtracted a full percentage point, on an 11.5% surge in imports.
However, look at what’s being imported: servers, networking equipment, cooling systems. These are all physical inputs for the data center buildout.
The import component of the GDP print is the growth-accounting confirmation of Pinebrook’s thesis that the hyperscaler capex is large enough to be a genuine investment-demand shock, and it’s now showing up everywhere we look — real yields, import data, capital goods orders.
This is the same mechanism the July 28th Signal & Noise Filter identified in the rates complex. RRFNR, the real, duration-adjusted neutral rate, round-tripped from a January trough of 1.14% back to 1.72%, essentially matching the level from the day the Fed started cutting in September 2024, and the move has been driven by real yields, not term premium or inflation expectations.
Thus, the Q2 GDP report did not point towards a demand problem. The weak Q2 print reflects an accounting identity that does not distinguish between weak demand and strong demand for things we don’t make. In short, hyperscaler capex puts the boom-boom in booming, and will likely push Q3 GDP to a high-two or low-three handle for GDP percentage growth.