What the Oil Chart Isn't Telling You

A capital markets perspective on three weeks that changed the investment landscape — not financial advice.

I’ve been watching financial crises for long enough to know what a real inflection point feels like. February 28 was one.

In the three weeks since US and Israeli forces launched Operation Epic Fury, oil has gone from $71 to over $115 a barrel. The Strait of Hormuz — the narrow passage through which a fifth of the world’s seaborne oil and a significant chunk of global LNG flows daily — has effectively shut. Rate cut expectations built up over the better part of a year have been dismantled in days. And several structural vulnerabilities that were already straining before the first strike — in private credit markets, semiconductor supply chains, and the architecture of the petrodollar system — have been pulled abruptly into the open.

None of what follows is financial advice. I have no view on your portfolio. What I do have is a view on what is actually happening — because I think some of the most consequential dynamics of the past three weeks are getting far less attention than the oil price and the equity volatility that dominate the headlines.

The Shock Was Anticipated. The Cascade Was Not.

Nobody should have been surprised by the outbreak of hostilities. Two carrier strike groups, 50,000 troops, months of public ultimatums — the buildup was the least covert military deployment in recent memory. Sell-side research in February was full of escalation risk scenarios. The question wasn’t if; it was when and how bad.

So when fighting started and the S&P barely moved on Day 1, the market was doing something rational: invoking history. Carson Group strategists had mapped 40 major geopolitical events across 85 years. On average, the S&P 500 lost less than 1% in the first month and recovered 3.4% over six months.[1] That is a real dataset. It is also, in this case, the wrong map for the territory.

The Strait of Hormuz carries roughly 20% of the world’s seaborne oil and significant LNG volumes.[2] Previous geopolitical shock studies were built from events where that chokepoint stayed open. By March 1, traffic through the Strait had collapsed by 80%. Container shipping lines were already routing around the Cape of Good Hope. More than 11,000 flights were cancelled in the first week.[3] The EIA confirmed what markets were slow to price: Brent moved from $71 on February 27 to $94 by March 9 — then kept going, toward $115 and beyond.[4] These weren’t tail events. They were the predictable consequences of a war erupting in the most energy-critical geography on earth.

Three Cascades behind the Headlines

Oil and equity volatility got most of the coverage. Three other dynamics didn’t — and they matter more for understanding where this ends up.

South Korea. The KOSPI fell 7.24% on March 3. Then 12.06% on March 4 — the largest single-day decline in the index’s history, worse than September 11, 2001.[5] Two days, roughly 18% gone. Most commentary called it a geopolitical reaction. It wasn’t, or not primarily. It was a leverage event that an energy shock detonated.

South Korea imports essentially all of its energy. About 70% of oil imports and 30% of LNG come from the Middle East.[6] When Asian LNG prices doubled in a week and the margin calls started hitting a market that had been leveraged to the gills — 32 trillion won in margin loans before the crash — the cascade was mechanical. Margin calls triggered algorithmic selling. Brokerages cut credit lines. Recovery buying dried up. By the time circuit breakers fired on Day 1 alone, over $270 billion in market value had been wiped — with the two-day total reaching approximately $550 billion.

But here is the part that has barely been reported: this isn’t just about Korean investors losing money. Samsung Electronics and SK Hynix together make up roughly half the KOSPI’s index weighting.[7] SK Hynix is not a peripheral player in global tech. It holds 53-57% of the global high-bandwidth memory market — the specialized HBM architecture that sits inside the majority of NVIDIA H100 and B200 GPU on the planet.[7a] In 2025, US-based revenue driven by NVIDIA’s HBM demand was nearly 70% of SK Hynix’s total sales. SK Group’s chairman, at NVIDIA’s GTC conference in March, put it plainly: HBM supply shortages could persist until 2030, with current demand deficits exceeding 20%, and new production capacity takes four to five years to build.[7a]

Then add the helium problem — which is not getting enough attention. Qatar’s Ras Laffan complex doesn’t just produce LNG. It produces helium as a by-product. Helium is critical for semiconductor lithography, the process that transfers circuit patterns onto wafers. When Ras Laffan went offline on March 2 after drone strikes, roughly 30% of global helium supply vanished from the market simultaneously. SK Hynix confirmed it had diversified supply and secured inventory — but Tom’s Hardware, which covers this closely, reported the supply chain was on a two-week clock before production impact.[7b]

The line from Qatari gas fields to South Korean fabs to NVIDIA GPU racks is not a metaphor. It is a physical supply chain. When people say the KOSPI crash is relevant to the AI trade, this is what they mean — not sentiment, not capital flows. Physics.

Qatar’s LNG. QatarEnergy declared force majeure on March 4, two days after drone strikes took Ras Laffan offline. Qatar was already behind schedule on its North Field East expansion — a 33 million tonne per annum project that buyers in Europe and Asia were counting on for 2027-2028 supply.[8] That supply gap just got wider and the timeline just got longer. European natural gas prices nearly doubled on the news.

Private credit. This is the story I keep coming back to. Not because it’s the loudest — it isn’t — but because it may be the most consequential.

UBS put out worst-case private credit default rate scenarios of 15% — higher than peak GFC corporate loan defaults.[9] Apollo’s co-president said many private credit firms may recover 20 to 40 cents on the dollar on software loans that default.[10] BDCs are now trading at around 73% of NAV — the lowest multiple in years.[11]

But the part that isn’t being said clearly enough is this: the private credit stress predates the war and isn’t really about the war. Private credit is heavily concentrated in PE-backed software companies. These are the same businesses that generative AI is now eating from the inside — competitive moats eroding, revenue models under pressure, what some analysts have called “SaaS-pocalypse.” Apollo’s 20-40 cent recovery estimate isn’t primarily a war consequence. It’s an AI disruption consequence landing at the worst possible moment. Because overlaid on top of the fundamental revenue problem is approximately $1.35 trillion of non-financial corporate debt maturing in 2026 — much of it in this exact sector — that now needs to be refinanced at rates that have moved sharply higher, not the cuts the market had been banking on all year.[11a] JP Morgan has effectively stopped lending to software companies privately and marked down portfolio valuations. That’s not a disclosure you make when everything is fine.[12]

Private credit is a broad asset class and blanket generalizations are unfair to the many managers running sound books. But three major funds gated in the same month, redemption requests running at double the allowed caps, and J.P. Morgan marking down software portfolios — that pattern has a historical rhyme that experienced credit investors will recognize.

The Rate Outlook Has Fundamentally Shifted

Coming into 2026, markets were pricing multiple Fed cuts. Now they’re pricing one — maybe. December Fed Funds futures have moved nearly 40 basis points since the end of February.[13] That’s a large repricing in a very short window.

The 5-year breakeven inflation rate hit 2.62% on March 17 — a one-year high.[14] Goldman Sachs estimates every $10 move in oil adds roughly 0.3% to US inflation.[15] Do the math on $115 oil versus $70 oil and you see why the Fed is in a difficult position. February payrolls came in at roughly minus 92,000 with unemployment at 4.4% — that’s the growth side of the mandate screaming for support.[16] The energy price side is screaming the opposite. The Fed has no good answer.

Australia hiked on March 17, explicitly naming the conflict’s fuel price impact in its statement.[17] Japan is pricing a meaningful probability of an April hike. Europe is now pricing hikes where cuts were expected. Synchronized global easing is over — replaced by a world where central banks are fighting an external supply shock with tools designed for demand problems.

Add to this the Fed leadership uncertainty. Powell has confirmed he will stay on the Board of Governors — stating at the March 18 press conference that he won’t leave his governor seat until the DOJ investigation into the Fed is “well and truly over,” and separately signalling intent to preserve institutional independence during a period of sustained political pressure. Markets have started pricing what some analysts call a “governance discount” — a risk premium on the credibility of future Fed decisions.[18] It’s a new variable. Probably not large right now. But it wasn’t there six months ago.

What Has Not Changed

Here’s the thing that keeps getting lost in three weeks of grim macro news: Jensen Huang stood on stage in San Jose on March 16 and doubled his demand forecast. He now sees at least $1 trillion in AI computing demand through 2027 — up from the $500 billion he projected a year ago at the same conference — and said he was certain actual demand would be higher.[19] The four largest hyperscalers are on course to spend over $650 billion combined in capex this year, roughly 70% more than 2025.[19] Data centers are now consuming an estimated 5-7% of US power demand and that number moves in one direction.[20]

Wars don’t cancel the need for compute. If anything, the military and intelligence applications being stress-tested in this conflict validate the AI thesis rather than undermine it. The demand signal is as strong as it has ever been. What the war has done is create a timing disruption layered on top of a structural trend — and timing disruptions, however painful, are not the same as structural breaks.

The demand signal is intact. The supply chain is fragile. Those two facts coexist — and the tension between them will define AI infrastructure investing for years.

The Energy Thesis — Structural, Not Just Cyclical

Most people covering the energy angle are writing about a war trade. Buy energy stocks, sell when it ends. That’s probably right for the near term. But it misses the bigger story.

CSIS’s energy team ran the numbers: if Qatar and UAE supply stays offline for more than a month, the LNG oversupply the market had been counting on for 2026 disappears entirely.[8] They were blunt: “There is simply no way to replace lost volumes from a supplier as large as Qatar.” QatarEnergy declared force majeure on March 4. CSIS estimates full production restoration could take weeks or months even after the fighting stops.[8]

So here’s what that means for the longer term. European buyers were counting on Qatari North Field East expansion in 2027-2028 to ease their post-Russian supply gap. That timeline just got longer. Those buyers will be competing with Asian importers for whatever supply exists. US LNG export infrastructure — which has been building for two decades to exploit exactly the gap between cheap domestic gas and expensive international gas — just became substantially more strategically valuable. Not because of the war. Despite the war, the structural dynamics were already there. The war just compressed a multi-year repricing into three weeks and made the thesis undeniable.

This doesn’t go away when the Strait reopens. Berlin, Tokyo, and Seoul’s energy security planners are writing long-term supply strategies right now, and US LNG is at the top of every one of them.

Rotations Within Rotations

Energy is the obvious winner — both from the oil shock and from the LNG thesis above. But what’s happening in the rest of the market is more interesting than “buy energy, sell risk.”

Consumer staples and healthcare are getting hit — not because of Iran, but because their own structural problems never went away. GLP-1 drugs are beginning to reshape snack food demand. Managed care is being squeezed between government cost pressure and rising medical utilization. These headwinds were there before February 28; the war just removed the AI narrative that had been crowding them out of investors’ attention.

Meanwhile, mega-cap tech is holding up better than the underlying macro would suggest. The dynamic isn’t hard to understand once you think about it: the war displaced the previous dominant narrative — AI disruption, white-collar displacement, SaaS competitive moats eroding — with a simpler one. And in that simpler narrative, the hyperscalers look clean. No oil exposure. No complex Gulf supply chains. Balance sheets that could absorb a 40-basis-point rate reprice without blinking. In a market suddenly full of things that could go badly wrong, that matters.

That relative safety may not last if the rate repricing continues. But for now, the flight to quality inside tech is rational — it’s just not the whole picture.

The Variable Markets Aren’t Pricing

On March 14, CNN reported that a senior Iranian official confirmed Iran is considering allowing limited oil tanker passage through the Strait of Hormuz — on the condition that cargo is priced in Chinese yuan, not US dollars.[22]

I want to be careful here, because this is easy to overstate. Dollar dominance is not going to collapse from one war. China’s own analysts, quoted in the South China Morning Post, urged caution and cited operational feasibility concerns.[22b] The proposal may not be implemented at scale.

But here is why it matters anyway. Previous de-dollarization discussions were theoretical — they lacked an operational mechanism. This one doesn’t. It comes with a chokepoint that Iran physically controls, a shadow fleet already moving Iranian crude to China outside the dollar system (11.7 million barrels in the first two weeks alone, per TankerTrackers.com satellite data[22]), and a payment infrastructure in China’s CIPS system that has been quietly expanding for years. The International Group of P&I Clubs withdrew insurance cover for Hormuz transit on March 5 — removing the dollar-based shipping system’s ability to operate there at any commercially sustainable price.[22a] As a result, tanker traffic through the strait has collapsed 94% for dollar-linked vessels.

The petrodollar system was built on mutual interest and the absence of a credible alternative. Iran’s yuan condition is testing, in real time, whether those conditions still hold. The answer will take years to fully know. But the structural conditions for an accelerated challenge now exist in a way they didn’t before February 28. For institutional investors with exposure to dollar-denominated assets over multi-year horizons, that’s a variable worth tracking — not panicking over, but tracking.

The Question That Matters Most

Not “when does this end?” — though that obviously matters. The question is: what does the world look like when it does?

The most credible research from Brookings, the Atlantic Council, and the Stimson Center converges on a negotiated pause as the most likely resolution within a defined timeframe.[21] When that happens, oil falls, shipping normalizes, risk assets rally hard. Investors who built disciplined buying lists and sat on their cash through the volatility will be well positioned.

But the post-conflict world is structurally different from the pre-conflict one. Energy security has been repriced as a first-order strategic asset, and the AI infrastructure buildout ensures it stays there. LNG infrastructure has been revealed as critical national security infrastructure, not just an investment theme. Private credit is going to emerge from this period more regulated, more conservatively structured, and more honestly marked. The AI supply chain concentration risk that everyone knew about abstractly has now been demonstrated concretely. And the dollar’s role in global energy settlement is a contested variable in a way it wasn’t three weeks ago.

The Strait of Hormuz will reopen. The question isn’t how to trade the day it does. The question is what the portfolio looks like for the years after.

The AI Paradox This Conflict Has Revealed

This needs to be said directly, because most commentary treats these as separate stories.

The conflict is simultaneously the strongest real-world validation of AI’s strategic value that has yet occurred — and a direct physical threat to the supply chain that makes AI infrastructure possible. Both are true at the same time.

On the validation side: Palantir generated $1.9 billion in US government revenue in 2025, growing 55% year over year, with a $10 billion Army contract and a NATO partnership.[23] Rosenblatt analysts have argued that Palantir’s Foundry and Gotham platforms have unique advantages in exactly the kind of multi-domain, time-sensitive intelligence operations this conflict demands. The US military has publicly confirmed AI tool usage in Operation Epic Fury. This is not a future application. It’s live.

On the supply chain side: SK Hynix’s fabs need energy and helium, both of which are disrupted. Samsung’s Taylor, Texas facility was already delayed to 2027. TSMC sits in a geography where US military overstretch in the Middle East changes adversary calculations. The $650 billion in hyperscaler capex committing to build Jensen Huang’s $1 trillion AI demand requires chips manufactured in a supply chain that is now demonstrably exposed to geopolitical concentration risk.

These aren’t contradictory. They define what AI infrastructure investing looks like for the next several years: strong demand, validated thesis, fragile supply chain that needs to be diversified and hardened. The Chips Act’s urgency just got a live demonstration. The policy and capital allocation response — domestic manufacturing, supply chain resilience, energy security for data centers — will be a significant investment theme in its own right.

A Note on Community Financial Institutions

One story that gets almost no coverage in macro market commentary: the credit unions and community banks serving the 135 million Americans who are credit union members and the small businesses that community banks finance.

Counterintuitively, these institutions are reasonably well-positioned for what comes next. America’s Credit Unions’ chief economist projected 2026 savings growth of around 6.5% for the sector — supported, in part, by members seeking stable deposit relationships as equity markets get rough.[24] S&P Global Market Intelligence’s annual ranking of best-performing community banks and credit unions, published March 18, highlighted institutions that had “successfully balanced growth with prudent risk management” — the kind of capital discipline that looks boring in a bull market and looks brilliant when conditions tighten.[24a]

As large banks absorb the pressure from private credit stress, leveraged loan deterioration, and potential regulatory scrutiny following the fund gating episodes of March 2026, the credit union and community bank model has something the alternatives don’t: member alignment. These institutions lend conservatively, know their local economies, and don’t have $82 billion flagship private credit funds to protect. The K-shaped economy doesn’t get easier in a stagflationary environment — but the institutions built around genuine member relationships, transparent lending standards, and community accountability are structurally positioned to deepen their relevance precisely when the institutional alternatives are under stress.

The technology challenge — how these institutions keep pace in an AI-driven financial services landscape — is real and separate. But the macro environment is, unexpectedly, one that plays to their core strengths.

A Final Observation

I’ve found the most instructive thing about the past three weeks isn’t the oil chart or the equity drawdowns. It’s how differently informed people are reading the same information.

Some see the market’s resilience as evidence the shock is being absorbed. Others see it as evidence the shock hasn’t fully arrived. History is genuinely mixed on which read is usually right. What it does suggest — fairly consistently — is that when private credit stress and external economic shocks arrive simultaneously, the timeline tends to be longer than it initially appears.

None of this calls for panic. Panic never works, and certainly not in an environment this complex. But attention — to what the data is actually saying, not what we need it to say, and to the structural shifts happening underneath the day-to-day noise — does work. Always has.

Markets recover from shocks that feel permanent. They also underestimate how long disruptions last when they’re clearly visible in hindsight. The investors who compound through periods like this aren’t the ones with the best predictions. They’re the ones who can hold genuine uncertainty without forcing a resolution before the evidence actually supports one.

That’s the discipline worth cultivating right now.

Sources & References

[1] Carson Group analysis of 40 geopolitical events, S&P 500 returns, cited in CNN Business, March 3, 2026

[2] U.S. Energy Information Administration (EIA), “Strait of Hormuz is a world oil transit chokepoint,” March 2026; CSIS Energy Security and Climate Change Program, March 2026

[3] Cirium aviation data, cited in CNBC live coverage, March 2-3, 2026

[4] EIA Short-Term Energy Outlook, March 10, 2026 — eia.gov/outlooks/steo

[5] Al Jazeera, “South Korea’s stock market suffers biggest drop in history,” March 4, 2026; Bloomberg, “Panic Sweeps South Korea Stocks,” March 4, 2026

[6] U.S. Energy Information Administration, Korea energy import data; CNBC, March 4, 2026

[7] Korea Capital Market Institute; Morningstar equity research; CNBC, “KOSPI: South Korea’s stock market volatility,” March 6, 2026

[7a] The Diplomat, March 2026 (South Korea chip sector / Nvidia export controls coverage) — SK Hynix US revenue nearly 70% of total sales driven by NVIDIA HBM demand; SK Group chairman at GTC 2026 warned HBM shortages could persist to 2030 with current deficits exceeding 20%; Counterpoint Research HBM market share data: SK Hynix 53%, Samsung 35%, Micron 11% (Q3 2025)

[7b] Tom’s Hardware, March 2026 (Qatar helium supply disruption / semiconductor supply chain coverage) — Qatar accounts for roughly 30% of global helium supply; helium essential for semiconductor lithography; SK Hynix confirmed supply diversification. Helium/semiconductor risk also confirmed: Euronews/Frederic Schneider, “Markets may be underestimating the Iran war,” March 16, 2026

[8] CSIS Energy Security and Climate Change Program, “What Does the Iran War Mean for Global Energy Markets?” March 2026

[9] UBS Global Research, private credit default scenario analysis, February 2026

[10] Apollo Global Management co-president remarks, cited in Bret Jensen / Seeking Alpha, March 18, 2026

[11] Forvis Mazars / Breakwave Advisors Weekly Note, “Private Credit: Still quite far from a 2008-event,” March 19, 2026; Reuters, Morningstar BDC data

[11a] Financial Content / MarketMinute, “The 2026 Credit Crunch: Geopolitical Shocks and the ‘Maturity Wall’ Collide,” March 18, 2026 — approximately $1.35 trillion in non-financial corporate debt maturing in 2026 facing refinancing at elevated rates; heavy private credit concentration in PE-backed software companies

[12] Breakwave Advisors, March 19, 2026; Americans for Financial Reform, March 17, 2026; Bloomberg, “Apollo and JPMorgan Impose a Dose of Reality on Private Credit,” March 12, 2026

[13] Mott Capital Management / Seeking Alpha, “A Massive Shift in the Fed’s Rate Outlook,” March 15, 2026; CME FedWatch Tool

[14] Federal Reserve Bank of St. Louis, FRED Series T5YIE, 5-Year Breakeven Inflation Rate, as of March 17, 2026 — fred.stlouisfed.org/series/T5YIE

[15] Goldman Sachs research note, cited in Barchart, March 19, 2026: “Every $10 Jump in Oil Could Add 0.3% to U.S. Inflation”

[16] OANDA / MarketPulse, “FOMC Meeting Preview,” March 17, 2026; Bureau of Labor Statistics February 2026 employment data

[17] Reserve Bank of Australia, Rate Decision Statement, March 17, 2026

[18] OANDA / MarketPulse, “FOMC Meeting Preview,” March 17, 2026; Atlantic Council MENA Futures Lab (Khalid Azim), “How to Understand the Iran War Market Swings: A Geopolitical Put Option,” March 9, 2026

[19] Jensen Huang, GTC 2026 keynote address, San Jose, March 16, 2026 — Axios, “Nvidia chips to reap $1 trillion, CEO Jensen Huang says at Nvidia GTC,” March 16, 2026; CNBC, “Nvidia GTC 2026: CEO Jensen Huang keynote — Blackwell, Vera Rubin,” March 16, 2026; Yahoo Finance, “Big Tech set to spend $650 billion in 2026 as AI investments soar,” February 2026; Bloomberg/Doubleline/Macrobond hyperscaler capex data cited in Bret Jensen, Seeking Alpha, March 18, 2026

[20] Goldman Sachs Research, “AI to Drive 165% Increase in Data Center Power Demand by 2030” — Goldman estimates data centers accounted for approximately 4% of US power demand in 2023, projected to more than double by 2030; Lawrence Berkeley National Laboratory estimates 6.7%-12% of US electricity by 2030

[21] Brookings Institution, “After the Strike: The Danger of War in Iran,” March 2026; Stimson Center, “Experts React: What the Epic Fury Iran Strikes Signal to the World,” March 2026; Atlantic Council MENA Futures Lab (Khalid Azim), “How to Understand the Iran War Market Swings: A Geopolitical Put Option,” March 9, 2026 (also cited at [18]); Charles Schwab, “Iran Conflict: What It Could Mean to Global Markets,” March 2026

[22] CNN, “Iran may allow tankers through Hormuz only if oil traded in Chinese yuan,” March 14, 2026 (senior Iranian official confirmed); CNBC, “Iran sends millions of oil barrels to China through Strait of Hormuz,” March 11, 2026 — TankerTrackers.com confirmed 11.7 million barrels shipped to China since Feb 28, all settled outside dollar system; Lloyd’s List Intelligence: 77 ships transited Hormuz in first half of March 2026 vs 1,229 in same period 2025 (94% collapse)

[22a] International Group of P&I Clubs withdrawal of Hormuz cover, March 5, 2026, cited in multiple shipping industry and financial press sources

[22b] South China Morning Post, “Does Iran have a yuan-for-Hormuz oil trade plan? Why analysts in China are urging caution,” March 14, 2026

[23] Motley Fool / Globe and Mail, “3 AI Stocks Caught in the Crossfire of the Iran War,” March 13, 2026 — Palantir $1.9bn US government revenue 2025, 55% YoY growth; $10bn Army contract; NATO partnership; Rosenblatt analyst cited; Seeking Alpha, “Palantir: The US-Iran Conflict Validates Its Unstoppable AI Military Moat,” March 2, 2026; US military confirmation of AI tool usage in Operation Epic Fury via Al Jazeera, March 12, 2026

[24] America’s Credit Unions, Senior Economist Dawit Kebede, “What to expect from the economy: 2026 outlook” — credit union savings growth projected at approximately 6.5% in 2026; former Fed Governor Adriana Kugler, remarks at America’s Credit Unions Governmental Affairs Conference 2026

[24a] S&P Global Market Intelligence, “Annual Rankings of Best-Performing US Community Banks, Public Banks and Credit Unions for 2025,” March 18, 2026

Disclaimer: These are personal observations on publicly available market commentary and macroeconomic developments. Nothing in this article constitutes financial advice or a recommendation to buy or sell any security. All market data cited is sourced from publicly available third-party sources as referenced above. All views are the author’s own and do not represent the views of any employer, client, or affiliated entity.

Kunle Fadeyi | CTO & Co-Founder, TAPP Engine | March 20, 2026