S&T SELL

GS Privorotsky GS Basics De Rating Invoice Not a Crisis Raising WFE Forecasts 25 Aug 2026

Aug 25, 20263 pages

From the report报告摘录Tech Valuation Pressure: SOX forward P/E down to 15x (from 21-22) amid NVDA/Broadcom CDS widening, semi inventory builds, and policy pushback on data centers.

Inside the report报告内文 Verbatim from the original PDF — first pages原版 PDF 开篇原文 · 逐字摘录

GS Privorotsky - GS Basics De Rating Invoice Not a Crisis Raising WFE Forecasts 25 Aug 2026 Rich Privorotsky · Goldman Sachs · Managing Director, FICC & Equities Tue 25 Aug 2026, 3:13am ET

Tech: Rebounding a bit after another challenging session for Nasdaq and the broader AI trade. Hard to isolate one catalyst, but I think there are a few things going on: First, at least in pockets credit is starting to ask questions that equity largely ignored. NVDA/Broadcom CDS has widened materially as the market focuses on AI financing backstops and off balance sheet commitments, feeding into the broader derating of semis. Second, Street dialogue has started shifting toward inventory builds across parts of the semi chain. Third, power bottlenecks and growing policy pushback around data centers are adding uncertainty around the speed and ultimate capacity of the build. None of these kill the AI story, but the range of outcomes is clearly getting wider = lower multiples (SOX on 24m forward P/E has gone from 21-22 to 15ish).

Oil: Oil lower yesterday because, despite the "economic D Day" rhetoric on Iran, the actual package came with enough boundary conditions and enough time that it wasn't an immediate change. "The Treasury Department did announce new sanctions on 60 individuals, entities and vessels, but the list did not include any of the Chinese financial institutions suspected of facilitating Iran's oil trade." (Reuters) More importantly, Bessent basically told you the reflexivity himself. Asked why he stopped short of actually imposing penalties on Iran and declined to identify the countries that would be targeted, Bessent responded, "Why would I want to blow up the global financial system?" (Reuters) All roads ultimately lead back to trying to keep the oil price as contained as possible.

Rates: Longer duration bonds have stabilized a bit and, short term, duration extension, month end and the change in the rate of increase in oil can keep them underpinned. But the bigger point is captured well in the WSJ op-ed Let the Bond Market Speak. "If the 30-year must trade at 5.5% to clear, that isn't a crisis. It is an invoice." Big picture, there is no real cure for long rates without accepting the trade off somewhere else. My guess remains weaker currency rather than austerity. Nobody takes the pain voluntarily. Europe tried austerity and voters rejected it. Eventually you accept a weaker currency, somewhat higher inflation and more intervention in the back end of the curve. Gold and real assets remain the clean expression of that.

Ex-AI Winners vs Hardware: To that point, financials are trading well, materials are trading well and industrials (ex the AI levered ones) are trading well. Throw healthcare in the same category, think AI spend is getting more useful for lower cost and the winners are those who spend the most on human capital who are currently spending the least on AI capex. The hardware complex is de-rating on the other side but at some point will be floored by earnings/revision. On AI specifically, there are new models arriving at an almost weekly cadence. SSI is particularly interesting given the rumors around continual or self learning, but nothing has been officially released or demonstrated yet. If continual learning really

works, it could be quite disruptive to the existing pretraining paradigm and reduce the need to continuously rerun ever larger training clusters. I am much less convinced it means less compute overall….if anything, better models that continuously learn probably push substantially more activity into inference. So for hardware, potentially a shift in where compute gets consumed rather than outright demand destruction. Mixed to positive if it arrives soon.

Risk: It has felt challenging because people are still long the obvious semi trades and they simply haven't worked. The other interesting feature is that US cyclicals have had a pretty serious relative pullback, broadly coincident with the rise in oil. Call it a stagflationary impulse from the highs. The cleanest resolution is simply lower oil. On AI, I think the debate will ebb and flow, but ultimately there is just a huge amount of vested interest in…

Read the full report + PDF阅读全文与 PDF

The full summary (3 key points) and the original S&T PDF are for MastermindX Pro members. 完整摘要(3 个要点)与 S&T 原始 PDF 为 MastermindX Pro 会员专享。

Read on MastermindX前往 MastermindX 阅读

Related institutional research相关机构研报

Not investment advice. MastermindX hosts third-party institutional research for reference and education; ratings and views are the authors', not ours. Browse the full Research Vault → 非投资建议。MastermindX 仅收录第三方机构研究,供参考与学习;其中评级与观点均属作者本人。浏览完整研报库 →