5% Is Back: The Yield Shock Playbook

What just happened is a regime break, not a blip

Takeaway: the bond market just told stocks their multiple is too high.

On Wednesday the 2-year jumped 8bp to 4.464%, the 10-year popped 7bp to 5.058% — first time since July 2007 — and the 30-year gained 4bp+ to 5.347%, per CNBC. Trading Economics put the intraday spike toward 5.1% on stronger-than-expected PMI plus hawkish Fed-speak.

S&P Global flash September PMI: services 58.7 vs 56.5 in August, near a five-year high; manufacturing 56.7, a 4+ year high. Chief economist Chris Williamson: barring the post-Covid reopening spike, the best business activity since early 2015. But input costs jumped at the steepest rate in four years on fuel and transport — oil is the accelerant. Fed Governor Michael Barr added fuel: further hikes likely necessary as inflation-target risks have increased.

Context matters. The Fed hiked last week under Chair Kevin Warsh, and CME FedWatch October hike odds rose to 64% from 55% Tuesday, from <10% a month ago. December adds ~48%. This is a hiking cycle being re-priced in real time, not a one-and-done.

Booming growth + rising prices is the worst mix for multiples

Takeaway: good news is bad news until the 10Y stops rising.

Market reaction today is textbook: SPY $768.66 –0.6%, QQQ $740.63 –0.9%, IWM $283.28 –1.37% (small caps bleed most), TLT $80.60 –1.4%, XLU $39.86 –1.65%. Only energy held: XLE $62.72 +1.5%. Bitcoin slipped to $84,133 –2.6% with $280M in 4-hour long liquidations under $84K, per Cointelegraph; ETH $2,660 –3.0%, SOL $114.13 –2.75%. Fear & Greed cooled to 71 Greed from 78 Extreme Greed.

My opinion: do not buy this dip in duration or long-multiple growth. At 5%+ the equity risk premium collapses. A 5% risk-free yield means the Nasdaq-100 at an all-time high of 30,732 (per CNBC on Burry) needs flawless earnings to justify itself — and BTIG already flags narrowing breadth (only 56% above 200-day vs 77% mid-August).

The trade: fade bonds, own the inflation hedge, shrink the tail

Takeaway: short the bounce in bonds, stay long energy, cut leveraged crypto longs.

Why this construction: TLT –1.4% today still leaves room if October odds push toward 80%. Risk is symmetric and defined by one level. Never set size you cannot hold through a 4.90% shakeout — my invalidation is 4.75%, not 4.90%.

Companion positioning (one sentence each): Hold/add XLE on red toward $60–62 for $72 (open EQ-XLE-1 — oil + gas tightness from Iran-war risk is the offset); trim IWM / unprofitable small-cap into any bounce; take 1/3 BTC into $86–87K rips and keep stops at daily <$82K (open CR-BTC-1 — futures-led rally with –180K BTC 30-day spot demand per CryptoQuant cannot survive 5% real competition).

Bull / base / bear: it hinges on oil and the October Fed

Takeaway: base case is higher-for-longer into October; the tails are oil-driven.

My opinion: consensus underprices the bear tail because it treats oil as transitory. Diesel at $6.31 all-time high and gas-market pricing of prolonged Iran tightness say otherwise — energy inflation feeds directly into PMI input costs the Fed is watching.

What everyone misses: Muse, Burry, and breadth are all rate stories

Takeaway: today's three loudest narratives get louder above 5%.

1. Meta's Muse vs brokerages — Schwab fell 6% Tuesday to early-July lows, down 8%+ since Muse, LPL –7%, RJF –3.5%, XLF –6% from highs, while META hit $754.38 +2.4% and Robinhood made YTD highs, per CNBC. Options: Schwab volume 5× average, 2:1 puts over calls, Jan 90-strike puts most bought. My read: Muse is real disruption of fee-friction, but higher rates amplify it — Piper Sandler's April warning that Schwab's low-yield sweep cash is in the crosshairs bites harder at 5%. Do not bottom-fish SCHW $100 until 10Y stabilizes.

2. Burry's chip short — Burry added to shorts in MU, NBIS, SOXX and PLTR, calling memory shortage temporary with China capacity rising (Acer CEO Chen; CSIS paper), Nasdaq-100 historically overvalued/top-heavy, per CNBC. MU $1,075.87 –1.85% and SNDK $1,829.51 –3.05% today support his timing into MU's Sep 30 earnings. My read: Burry is early on fundamentals but right on discount rates — memory into a hiking cycle with Citi warning prices may peak early is a terrible long into earnings. This reinforces open EQ-SNDK-1 discipline: no chase above $1,910, wait $1,650–1,780.

3. Breadth + crypto leverage — NDX flat since mid-August with breadth collapsing + $1.7B of 2-day BTC ETF inflows now underwater (ETF cost basis ~$86K) + $16B quarterly options expiry Friday with a call-heavy book = forced-selling risk if 10Y holds >5%. BTC $82K is the line Rekt Capital flags to avoid falling back into the $60–80K range.

What would prove me wrong

Takeaway: I will flip on prints, not punditry.

• 10Y daily close <4.75% — cover TLT short, add duration and QQQ.
• September payrolls show hiring stall AND core CPI cools — hiking cycle dies, my base case dies with it.
• Brent breaks <$88 daily with Hormuz reopening headlines — energy hedge off, rate pressure off.
• MU Sep 30 guides up with HBM sold out through 2027 and holds >$1,100 — chip-short thesis (and multiple-compression spillover) weakens.

What to do

Takeaway: sit in T-bills, hedge with energy, trade small until 4.75% or 5.25% picks the direction.

For most readers: no new duration, no new leveraged length. Keep dry powder in T-bills/short TIPS (consistent with open MACRO-DURATION neutral-short stance). If trading: small TLT fade $80–82, target $75, kill >$84. Hold XLE, trail QCOM/PLTR/INTC winners, respect BTC $82K. Next catalysts: Fed speakers into October FOMC pricing, jobs/CPI, MU Sep 30, Trump-Xi Thursday.

Appendix

Contents: A. Data snapshot · B. Model · C. Sources

A. Data snapshot (intraday Sep 23, 2026 ~16:00 UTC, financial-data handler unless noted)

B. Model — why 5.25% / 4.75% are the lines

Assumptions: (1) term premium stays positive while Fed hikes + oil elevated; (2) equity earnings yield must clear 10Y + ~150bp ERP — at SPY ~$769 (~22× forward), earnings yield ~4.5% leaves no cushion at 5.058%; (3) TLT duration ~15y, so +25bp ≈ –3.5–4% price. Load-bearing input: oil stays firm on Iran risk. If wrong (diplomacy + Brent <$88), input-cost channel collapses, 10Y falls through 4.75% and TLT target flips to $86+ — which is exactly my invalidation.

C. Sources

• CNBC — 10-year leaps to fresh 19-year high after hot readings
• Trading Economics — 10Y back above 5%
• CNBC — Burry adds to shorts as Nasdaq-100 hits high
• CNBC — Meta's Muse reignites AI disruption, Schwab targeted
• Cointelegraph — $280M longs as BTC dips under $84K