Back to index

Update: Oil Re-Spike Invalidates Key Thesis Components

April 13, 2026 11:35 AM ET

The memo below was written based on April 10 data showing oil at $78.5. As of April 13, WTI has re-spiked to $103.7 (+32% from the April 10 low), invalidating several core claims:

  1. "The oil trade is over" (Part 2) — Wrong. Oil re-spiked despite the Hormuz bypass infrastructure.
  2. "Oil normalization Fed unlock is leading" — Downgrade required. Oil at $103 removes Fed cover for June cut.
  3. "Energy-CPI stagflation trap is permanently dead" — Reactivate. April/May CPI will not deflate mechanically with oil at $103.
  4. VIX compression to 22-23 — Partially reversed. VIX now at 21.3 (still compressed but not as much as claimed).

What hasn't changed:

Revised assessment: The oil normalization was temporary, not structural. The Hormuz bypass infrastructure did not prevent the re-spike. The Fed June cut is now off the table unless oil collapses again. The hyperscaler capex credibility test remains the dominant near-term risk, but the macro backdrop is less supportive than assessed on April 10.

The Nine-Day Window: Why the Medium-Term Equity Outlook Reduces to a Single Earnings Season

This memo synthesizes the US equity / tariff / AI journal, two new validated theories (hyperscaler capex credibility test, tariff exemption front-run), updated evaluations of oil-normalization-fed-unlock and trump-ceasefire-playbook-buy-dip, and the current factor snapshot (April 10, 2026).

Part 1: The Institutional View

There is a temptation, when you look at the current market, to see complexity everywhere. A Supreme Court ruling that rewrote the tariff regime overnight. A ceasefire that may or may not hold. A CPI print that says 3.3% on the headline and 2.6% on the core. A semiconductor index at an all-time high while the VIX sits at 27. Gold near $4,800 and oil at $78 after touching $113 five days ago. It feels like you need a PhD in geopolitics, a law degree, and a Bloomberg terminal just to figure out whether to be long or short.

But I think the complexity is an illusion. When you strip away the noise, the medium-term equity outlook — call it the next four to eight weeks — reduces to a single question: will the four largest hyperscalers reaffirm their $650-700 billion capex guidance on their Q1 earnings calls between April 23 and May 1?

Everything else is either already resolved, already priced, or dependent on the answer to that question.

What's already resolved:

The tariff regime has structurally de-escalated. The Supreme Court's 6-3 ruling in Learning Resources v. Trump (February 20) struck down the IEEPA tariffs that had pushed the effective rate to 47%. The replacement — a 10% Section 122 global tariff expiring July 24 — is a fraction of what came before. The April 12 electronics exemption removed reciprocal tariffs on smartphones, computers, and semiconductors entirely, even if temporarily. The effective tariff rate is 13.7%. This is not the tariff environment of April 2025.

The ceasefire playbook has worked. Our trump-ceasefire-playbook-buy-dip theory has been upgraded to leading. SPY recovered from ~629 to 679.46 (as of April 13). VIX compressed from 30+ to 21.3. UPDATE April 13: Oil crashed from $113 to $78.5 on April 10, but has since re-spiked to $103.7. The pattern — escalate, deal, buy the dip — played out as the historical analogs (Soleimani 2020, trade war 2018-2020, April 2025 tariff shock) predicted, but the oil normalization was temporary, not permanent.

The CPI scare is a lagging indicator. March CPI at 3.3% headline looks alarming until you decompose it: energy surged 10.9% month-over-month (gasoline +21.2%), the largest monthly energy increase in nearly two decades. But core CPI came in at 2.6% — below the 2.7% consensus. The headline spike reflects oil at $100+ during the conflict period. UPDATE April 13: Oil has re-spiked to $103.7, removing the mechanical deflation thesis. April and May CPI will likely remain elevated. The factor snapshot from April 10 showed commodity_pressure as NEUTRAL when oil was at $78.5, but that assessment is now stale. The Fed no longer has cover to look through the headline.

What's already priced:

The SOX at 8,926 — an all-time closing high on April 10 — has priced in the structural AI demand thesis. UPDATE April 13: SOXX now at 386.26, down slightly intraday but still near all-time highs. $650-700 billion in hyperscaler capex (+71% YoY). NVIDIA at $188.63 (up from $182.08 on April 11), FY2026 revenue of $215.94 billion (+65% YoY). BofA's forecast of $1 trillion in annual semiconductor sales. The HBM oligopoly with SK Hynix at 50-62% share and Micron printing 75% gross margins. All of this is in the price. What is NOT in the price is any scenario where that capex number comes down.

The nine-day window:

Microsoft reports April 23. Alphabet reports April 24. Meta reports April 30. Amazon reports May 1. Four companies. Nine days. Between them, they account for the vast majority of the GPU orders that underpin NVIDIA's revenue, which in turn underpins the SOX at all-time highs, which in turn underpins the "AI industrialization" narrative that has driven the market's recovery from the Iran conflict.

Our hyperscaler capex credibility test theory frames this as a binary with severe asymmetry. SOXX sits near all-time highs (386.26 as of April 13). Weekly RSI is 71.64 — overbought. The weekly MACD histogram has turned negative while price is at highs — a bearish divergence. There is zero technical buffer for disappointment.

The historical precedent is clear. When Meta hedged its capex guidance in Q4 2022, NVIDIA fell 18% and the SOX fell 12% within one week. When AWS growth decelerated in Q3 2022, cloud infrastructure stocks sold off 15-25% over three weeks. The mechanism is established and the current setup — SOX at ATH, VIX still elevated at 27.4, credit stress rising (HYG/LQD -1.8%), defensive rotation active (XLY/XLP -2.1%) — amplifies the downside.

But here is where I want to be careful about what I don't know. The base case is reaffirmation. No hyperscaler has pre-announced a capex cut. Microsoft has already de-rated 25.6% over 100 days (P/E compressed from 36x to 23x) — some caution is already in the price for the largest spender. The April 12 tariff exemption removes one rationale for hedging capex guidance. And the structural demand signals — TSMC Q1 revenue +35% YoY, DDR4 prices +1,360% in Q1, HBM sold out through 2027 — are not the kind of data that makes CFOs cut capex.

The probability-weighted outcome is probably 65-70% reaffirmation, 20-25% hedge language from one company, 5-10% actual cut. But the payoff structure is inverted: reaffirmation is partially priced (SOX at ATH), while a hedge is not priced at all. This is the kind of asymmetry that makes long volatility the right expression — not because you think capex gets cut, but because the market hasn't priced the possibility.

The secondary trade:

The tariff exemption front-run is a real phenomenon but a weaker trade. The 30-60 day exemption window creates an incentive to stockpile semiconductor imports before sector-specific tariffs replace it. TSMC's Q1 +35% YoY revenue beat and the extreme DRAM price spikes (DDR4 +1,360%, DDR5 +400%) are consistent with pull-forward behavior already underway. But the collapse half of this thesis is structurally capped. HBM supply is sold out through 2027 via binding contracts. Even if the exemption expires and import costs rise, the demand floor is real. This is a calendar spread trade, not a directional short.

The factor regime:

The factor snapshot tells a story of a market in transition. UPDATE April 13: This snapshot is now stale. Based on current data: Rates pressure is NEUTRAL (10Y at 4.35%). Commodity pressure is ELEVATED (oil re-spiked to $103.7 from $78.5). Volatility is MODERATELY ELEVATED (VIX 21.3, down from 27.4 but still above 20). Credit stress and equity duration pressure remain ELEVATED. Defensive rotation is active. Only 41% of stocks are above their 200-day moving average.

This is a market that has recovered on the index level but remains stressed underneath. The recovery is narrow — semis and industrials leading, everything else lagging. It means the rally is concentrated in exactly the names that are most exposed to the hyperscaler capex binary.

The oil-normalization-fed-unlock theory is evolving toward partial activation. UPDATE April 13: This theory must be downgraded. Oil re-spiked to $103.7, removing the mechanical CPI deflation thesis. Core CPI at 2.6% gives the Fed some cover, but with oil back above $100, the June cut is off the table unless oil collapses again. If Fed speakers acknowledge the oil normalization and oil stays below $85, this theory could upgrade to leading and become the structural tailwind that supports the recovery beyond the earnings window. But as of April 13, oil at $103.7 invalidates the activation condition.

The honest assessment:

I think the market is probably right that the worst is behind us. The tariff regime has structurally de-escalated. The ceasefire playbook worked. UPDATE April 13: Oil normalized temporarily but has re-spiked to $103.7. Core inflation is stable at 2.6%, but headline CPI will remain elevated. The AI demand thesis is real and structural.

But "probably right" at SOXX all-time highs with VIX at 21.3 and credit stress rising is not the same as "safely right." The market has priced in the good news and has not priced in the possibility that even one of four hyperscalers uses the word "monitoring" on an earnings call. That asymmetry is the trade.

Part 2: The Asymmetric View

The institutional analysis above is correct about the facts and too cautious about the trade. Let me explain why.

The Marks voice says "long volatility into the earnings window" because the asymmetry favors it. Fine. But let's do the actual math that the institutional view was too polite to spell out.

NVDA: the per-share economics nobody's running

NVDA at $188.63 (April 13). FY2026 revenue $215.94B. FY2027 consensus is $280B (Blackwell + Rubin ramp). At 75% gross margin and 65% operating margin, that's $182B operating income. Tax at 12% (NVDA's effective rate with international structure), you get $160B net income. On 24.5B diluted shares, that's $6.53 EPS.

At $188.63, you're paying 29x FY2027 earnings for a company growing revenue 30%+ with 75% gross margins in a supply-constrained oligopoly. The S&P 500 trades at 22x forward earnings growing 8%. NVDA's PEG ratio is under 1.0.

If all four hyperscalers reaffirm and FY2027 consensus moves to $300B (which it will if Azure, AWS, and GCP all guide up), the EPS goes to $7.35. At 30x, that's $220. At 35x (which the market will pay in a risk-on regime with VIX below 20), that's $257.

The institutional view says "long volatility." The asymmetric view says: the straddle is the hedge, but the directional bet is long NVDA into the reaffirmation. You're paying 29x for 30%+ growth. That's not expensive. That's the cheapest mega-cap growth stock relative to its growth rate.

The tariff exemption math everyone's ignoring

The April 12 exemption covers semiconductors from reciprocal tariffs. The Section 232 tariff (25% on advanced chips) remains. But here's what nobody's calculating: the Section 232 tariff exempts chips for data centers, R&D, and domestic manufacturing buildout. That means the hyperscalers — the ones buying $650-700B worth of infrastructure — pay ZERO tariff on the chips they actually need. The 25% tariff hits re-export and non-US-destined chips. It's an onshoring incentive disguised as a tariff.

So when the institutional view says "tariff uncertainty could cause hyperscalers to hedge capex," the actual answer is: the tariff structure is designed to encourage exactly the capex they're already doing. The exemption isn't temporary relief — it's the permanent structure with a temporary label.

The oil trade that's already over

UPDATE April 13: This section is now incorrect. Oil re-spiked to $103.7 from the April 10 low of $78.5. The institutional view correctly noted oil at $78.5 and said the stagflation trap is invalidated. But it didn't go far enough — and events proved it went too far. Read the external analysis from April 12: KSA's east-west pipeline is now carrying 7 million bpd (up from 5 million in March). The UAE opened a 1.8 million bpd line to Fujairah. That's 4 million barrels per day of new bypass capacity that didn't exist six weeks ago.

The Strait of Hormuz bypass infrastructure exists, but it did not prevent the oil re-spike. The ceasefire may be fragile, or demand dynamics may be driving the move. Either way, the claim that "the Strait will be worth little more than a memory" was premature. The oil risk premium is back.

This means the oil risk premium is structurally lower than the market thinks. UPDATE April 13: This claim is now invalidated. Oil at $103.7 shows the risk premium is NOT structurally lower. Not because geopolitical risk has decreased, but because the chokepoint that created the risk premium is being engineered out of existence. The bypass infrastructure exists but did not prevent the re-spike. Bessent has already sold 53 million barrels at an average of $105 and can buy the Cal 27 strip at $73 — that's $5 billion in profit for the US Treasury. The government is literally making money on the crisis.

For the Fed, this is the best possible outcome. UPDATE April 13: This is no longer the best possible outcome. Oil normalizes not because of a fragile ceasefire but because of permanent infrastructure changes. But oil re-spiked to $103.7 anyway. April and May CPI will deflate mechanically. UPDATE: No, they won't. Oil at $103.7 means CPI stays elevated. The June cut is back on the table — not as a hope, but as a near-certainty if oil stays below $85. UPDATE: Oil is at $103.7. The June cut is off the table.

The trade:

The institutional view says "long volatility." Here's what I'd actually do:

  1. Core position: long NVDA into the April 23-May 1 window. 29x FY2027 earnings for 30%+ growth is cheap. If hyperscalers reaffirm, this goes to $220. If one hedges, you take a 15% hit and buy more — because the structural demand is real and one quarter of hedge language doesn't change $650B in committed capex.

  2. Hedge: NVDA put spread ($175/$160, May expiry). Costs ~$4-5 per contract. Caps your downside at $175 if the worst case materializes. You're paying 2.5% of position value for insurance against a 15-20% drawdown.

  3. Calendar spread: long TSM May / short TSM July. Captures the pull-forward beat, hedges the exemption expiry. TSMC's April monthly revenue (released early May) is the confirmation signal.

  4. UPDATE April 13: The oil trade is NOT over. Oil re-spiked to $103.7. Don't short XLE — the KSA/UAE pipeline expansion means Gulf producers are the winners, not the losers. But don't be long oil above $100 either. The re-spike invalidates the "Hormuz bypass makes the risk premium a decaying asset" thesis.

  5. Watch for the Fed unlock. UPDATE April 13: With oil at $103.7, the Fed unlock is off the table. If Waller or Daly says "we can look through the energy component" in the next two weeks, go long TLT. But with oil back above $100, this is unlikely. The June cut repricing from 65% to 85%+ probability is a 3-5% move in long-duration bonds — but it requires oil to collapse again first.

TL;DR:

Addendum: Response to "Market Vibes" (April 12, 2026) — Updated April 13

UPDATE April 13: The analysis below was written when oil was at $78.5 on April 10. Oil has since re-spiked to $103.7, invalidating the core claim that the Hormuz bypass makes oil normalization structural.

The external analysis from April 12 contains an insight that our research framework missed and that materially strengthens the bull case. UPDATE: This insight was correct about the infrastructure but wrong about the outcome.

The key claim: the Strait of Hormuz is being permanently bypassed. KSA's east-west pipeline expanded from 5 to 7 million bpd. UAE opened a 1.8 million bpd line to Fujairah. That's 4 million bpd of new capacity that routes around Hormuz entirely. The analysis argues that even if the ceasefire collapses, the oil chokepoint risk is structurally diminished because the physical infrastructure to bypass it now exists.

UPDATE April 13: The infrastructure exists, but oil re-spiked to $103.7 anyway. This suggests either: (1) the ceasefire is more fragile than assessed, (2) demand dynamics are driving the move, or (3) the bypass capacity is not yet fully operational or sufficient to offset Hormuz risk. The claim that the bypass makes the oil risk premium "a decaying asset" was premature.

This has three implications for our thesis framework:

1. The oil-normalization-fed-unlock theory is stronger than we assessed. UPDATE April 13: No, it's weaker. We evaluated it as "evolving toward partial activation" based on oil at $78.5 and core CPI stability. But if the Hormuz bypass is permanent, the oil normalization isn't dependent on the ceasefire holding. It's structural. The theory should be upgraded to leading regardless of ceasefire outcome. UPDATE: Oil re-spiked to $103.7. The theory must be downgraded, not upgraded. The bypass infrastructure did not prevent the re-spike.

2. The energy-cpi-stagflation-trap is permanently dead, not just temporarily invalidated. UPDATE April 13: No, it's reactivated. We downgraded it because oil fell to $78.5. But the bypass infrastructure means oil cannot re-spike to $95+ on a ceasefire failure — the chokepoint that created the spike no longer constrains supply. The hypothesis's reactivation condition ("oil re-spikes above $95 on ceasefire failure") is now structurally unlikely. UPDATE: Oil re-spiked to $103.7. The reactivation condition has been met. The stagflation trap is back on the table.

3. The Bessent trade is a signal. The US Treasury sold 53 million barrels at ~$105 average and can buy the Cal 27 strip at $73. That's $5 billion in profit. When the government is actively profiting from the oil normalization, the policy incentive is to keep oil low, not let it re-spike. This aligns the government's financial interest with the Fed's inflation mandate — a rare convergence that makes the June cut more likely, not less. UPDATE April 13: Oil re-spiked to $103.7 despite this incentive. Either the government's ability to suppress oil is limited, or other factors (demand, geopolitical) are overwhelming the policy incentive.

The analysis also raises a geopolitical dimension we hadn't considered: the blockade severs Iran-China financial flows. China loses 2 million bpd of cheap Iranian oil paid in RMB and must replace it with dollar-denominated open-market purchases. This is dollar-positive (supports DXY), gold-negative (reduces RMB-denominated commodity flows), and structurally bearish for the Iran-China economic relationship. It's a financial weapon, not a military one.

For our purposes, the actionable takeaway is: the oil risk premium is a decaying asset. UPDATE April 13: This takeaway is wrong. Don't hedge for an oil re-spike. The infrastructure to prevent one now exists. UPDATE: The infrastructure exists but did not prevent the re-spike. Oil is at $103.7. The risk premium is back.

April 13 Market Open Update (11:30 AM ET)

Peace talks collapsed. The blockade is back on.

Over the weekend, US-Iran negotiations failed to produce a deal. Trump announced a blockade of Iranian shipping, leaving the ceasefire in a fragile state and Hormuz effectively closed to Iranian-linked flows. This is the scenario the memo's April 13 updates flagged as a risk but did not fully price: not just a re-spike, but a structural re-escalation.

Intraday price action as of midday:

Asset Price Change
SPY 679.59 +0.32%
QQQ 611.89 +0.40%
NVDA 188.62 +1.39%
SOXX 387.50 +0.66%
VIX 19.94 -2.21%
XLE 57.29 -0.57%
GLD 432.16 -0.60%
TLT 86.38 -0.07%
Brent ~$102 +7.0%

The session is telling a coherent story, and it is not the one you'd expect. Equities are up. VIX has compressed below 20 for the first time since the conflict began. NVDA is the session leader at +1.39%. SOXX is at 387.50, a new high. The market is not treating the blockade as a crisis — it is treating it as a known condition.

The divergence that matters is XLE -0.57% while Brent is +7%. Energy equities are not following oil higher. This is consistent with the Bessent trade thesis: the US government is long oil via the SPR short and has a policy incentive to keep the conflict contained rather than escalate. The market may be reading the blockade as a pressure tactic, not a prelude to strikes on energy infrastructure. MST Marquee analyst Saul Kavonic flagged the key remaining risk correctly: "if the U.S. renews strikes on Iran, raising the risk of strikes on energy infrastructure across the region." That scenario is not priced. The current session prices a contained blockade.

What this does to the thesis framework:

The Fed unlock is now definitively off the table for June. Money markets are pricing less than 20% probability of any cut in 2026. Trump himself acknowledged on Sunday that oil and gasoline may stay elevated through the November midterms — a rare admission that removes any political pressure on the Fed to act. TLT at 86.38 with 10Y yields at 4.33% (+2bps) confirms the bond market has absorbed this. Long TLT is not the trade until oil breaks below $85 and that now requires either a ceasefire or a demand shock.

The hyperscaler capex binary (April 23–May 1) remains the dominant near-term risk, and the intraday action is consistent with the market front-running a reaffirmation. NVDA leading the session at +1.39% with VIX below 20 is the setup for a squeeze into earnings if Microsoft and Alphabet open with strong capex language. The asymmetry described in the memo — reaffirmation partially priced, hedge language not priced at all — is still the correct framing. The put spread hedge ($175/$160 May expiry) is now cheaper given VIX compression, which makes it more attractive, not less.

One new signal: Goldman Sachs published a "Rule of 10" framework today flagging NVDA and Meta as positioned for a comeback. This is institutional cover for the reaffirmation base case. When Goldman publishes a bull framework two weeks before hyperscaler earnings, it is not analysis — it is positioning. Note it, don't trade it, but recognize it as a sentiment indicator that the institutional consensus is leaning toward reaffirmation.

Revised TL;DR for April 13 open:

Related Research: