Sunday, May 10

The S&P 500 finished Friday May 8 at 7,399, another all-time high and the index's sixth straight weekly gain, the longest winning streak since 2024. The Nasdaq closed at 26,247, also a record, helped by a 1.7% jump as chip stocks led the day. The trigger underneath both moves was crude oil. Brent, the global benchmark that prices most of the world's seaborne barrels, fell about 13% over the week. West Texas Intermediate, the US benchmark, dropped close to 20%. Both slid because Washington and Tehran appear close to a framework deal to end their roughly month-long shooting war and reopen the Strait of Hormuz, the narrow chokepoint between Iran and Oman through which roughly a fifth of global oil flows.

The military backstory matters because it explains why the rally is so jumpy. Earlier in the spring the US and Iran traded strikes after Iran's navy interdicted commercial tankers transiting Hormuz. Tehran briefly closed the strait to most traffic, oil spiked above $110 a barrel, and US gasoline jumped past $4.50 a gallon (national average), the fastest run-up in three decades. By Wednesday, leaks from mediators in Pakistan suggested Iran had softened its terms. Brent dropped 8% in a single session. Trump paused "Project Freedom," the US naval escort operation in Hormuz, and said the ceasefire was holding. None of this is signed yet, but the tape acted like it was.

The other half of Friday's move was a jobs report that landed almost exactly where the consensus wanted it. April nonfarm payrolls (the monthly Labor Department headline that estimates how many people are on US company payrolls outside agriculture) added 115,000 positions, beating forecasts in the mid-90s. The unemployment rate held at 4.3%. That combination, decent hiring without wage acceleration, gives the Federal Reserve room to keep cutting later this summer without looking like it's chasing a slowdown. Two-year Treasury yields slid, which traders read as a higher chance of a July cut.

Big tech did the heavy lifting on the day. The Magnificent Seven (the loose label for Apple, Microsoft, Alphabet, Amazon, Meta, Nvidia, and Tesla) drove most of the index gains, with semiconductors leading. AI capex spending continues to be the single biggest swing factor in S&P earnings, which is why a 1.7% Nasdaq move dragged the broader index along. FactSet's running tally of Q1 earnings shows S&P 500 firms beating estimates at roughly the long-run pace, with information technology and communications services sectors carrying outsized weight.

There are two reasons to take the celebration with a grain of salt. First, gas prices at the pump tend to lag crude on the way down. Stations that bought wholesale fuel at $3.20 a gallon last month aren't dropping retail prices to $2.80 just because Brent fell. Most analysts think US drivers will keep paying close to wartime prices through June, which keeps headline inflation sticky. The Cleveland Fed's nowcast pegs May CPI at 3.89%, up from 3.3% in March, almost entirely due to the Hormuz spike still working through the system. The Fed cuts rates expecting inflation to fade. If it doesn't fade in time, dovish (a market term for accommodative or rate-cutting) repositioning unwinds fast.

Second, the rally is narrow. Cash holdings at large institutional investors have crept back up to 3.3%, a defensive sign more typical of late-cycle markets than the early innings of a bull run. Value stocks, defensive sectors, and commodities have lagged the index for weeks. Breadth measures (the share of stocks participating in a rally) are weaker than the headline numbers suggest. When the leadership is this concentrated in a handful of mega-cap chip names, a single bad guidance from Nvidia or TSMC can flip the picture inside a session.

Knock-on effects beyond the stock chart are worth tracking. Lower oil hits energy company margins, which is why ExxonMobil and Chevron lagged a green day. It also relieves pressure on emerging-market central banks that had been propping up currencies against a strong dollar; the Indian rupee and Brazilian real both firmed Friday. European industrials, which import nearly all their oil and got squeezed in March, started to claw back. The biggest second-order trade is in airline stocks, where jet fuel is roughly 25% of operating costs. Delta and United both jumped 4% to 6% on the week. If a deal actually gets signed, those moves probably extend.

What to watch next. The Iran framework, if it materializes, is expected before the next G20 meeting in early June; a missed deadline would unwind much of the oil and equity move. Tuesday May 12 brings April CPI in the US; consensus is for headline inflation around 3.6% year over year, and a print materially above that would dent rate-cut hopes. Earnings from Walmart and Target later in the week will tell you whether US consumer spending is holding up against still-elevated gas prices. And Nvidia reports later in May; given how much of the S&P's rise sits on its shoulders, any softening in data-center guidance moves the whole index.

State Bank of India, the country's largest lender by assets and the de facto bellwether for Indian banking, reported its January-March quarter on Friday May 8. Net profit came in at ₹19,684 crore, up 5.6% from the same quarter a year earlier. By the standards of most banking systems that would be a clean beat. By the standards of the Indian street, which had been modeling something closer to ₹20,300 crore, it was a miss. The stock fell 6.62% intraday to close at ₹1,019.55 on the BSE, the sharpest single-day drop for SBI in nearly two years and enough to drag the Bank Nifty index down with it.

Loan growth held at the high single digits. The miss came from two specific lines on the income statement that had been doing heavy lifting through 2024 and 2025. The first is net interest margin, or NIM. NIM is the difference between what a bank earns on loans and what it pays on deposits, expressed as a percentage of interest-earning assets. SBI's NIM slipped to 2.93% in Q4, down 18 basis points (a basis point is one one-hundredth of a percentage point) from the prior quarter and below the 3% psychological line. That matters because most of SBI's deposit base is repriced quickly when the Reserve Bank of India cuts the repo rate (the policy rate at which banks borrow short-term from the RBI), but the loan book repricing lags. The RBI has cut the repo rate twice since November to support growth. SBI has been on the wrong side of that timing, paying old deposit rates while collecting fresh, lower loan rates.

The second line is non-interest income, the catch-all bucket that includes treasury gains (profit from buying and selling government bonds), fee income, and recoveries from written-off loans. That bucket fell 29% year over year to ₹17,314 crore. The bulk of the damage came from a ₹1,471 crore loss on the sale of investments, against a ₹6,879 crore gain in the same quarter a year ago. Indian government bond yields drifted up through the quarter on heavy state-government borrowing and a spike in oil prices, and SBI was caught with too much duration on its trading book. Operating profit dropped 11% year on year to ₹27,704 crore as a result.

Asset quality, which is what wrecks Indian banks when it goes wrong, actually improved on the headline numbers. Gross non-performing assets fell to 1.82% of loans from 2.24% a year earlier, and the slippage ratio (new bad loans as a share of the loan book) stayed contained. The board declared a ₹17.35 dividend for the year, which is a modest hike. Provision coverage stayed comfortable. The market clearly looked past those positives and priced the earnings power question instead: if NIM has structurally compressed and treasury gains are gone, what is the right multiple for a state-owned bank growing earnings in the high single digits?

The pattern is not unique to SBI. ICICI and HDFC both flagged similar margin pressure earlier in the season. Private banks have a small structural advantage on deposit mix (more current and savings account balances, which are cheaper) and have absorbed the RBI cuts better. The market reaction reflected that gap: HDFC Bank actually closed up on Friday despite the Bank Nifty's 1.4% drop. Foreign portfolio investors have been quietly rotating from public sector banks into private peers since March, a reversal of the 2023-24 trade where SBI was the consensus value pick. SBI's valuation at roughly 1.4x book is still cheap by global comparison, but the re-rating story (the multiple expansion that powered the stock from ₹500 to ₹1,200 over two years) needs fresh fuel.

The skeptical aside worth flagging is that public sector banks in India often beat on credit costs and miss on operating leverage, then make it back later in the cycle when private capex picks up. SBI is the country's largest lender to infrastructure and large-corporate borrowers. If the government's planned ₹11.1 trillion capital spending push for FY27 lands as advertised, and if private capex finally turns (the missing ingredient in the Indian investment story for almost a decade), SBI's loan book grows faster than the system. Margins might still bleed, but absolute earnings could surprise on the upside.

Knock-on effects across the wider market. The Bank Nifty (the index of India's twelve biggest banks) lost 1.4% on the day and is now flat for the year, dragging the Sensex with it because financials carry a 35% weight in the broader index. Bond yields ticked up because a softer banking quarter is one more reason for the RBI to hold rates rather than cut further. The rupee weakened slightly on foreign portfolio outflows. Mid-cap and small-cap PSU banks (smaller state-owned lenders like Bank of Baroda and Canara Bank), which trade at a discount to SBI, fell harder, between 3% and 5%.

What to watch. Bank of Baroda reports next Monday, and Punjab National Bank later in the week; both will tell you whether SBI's margin compression is a one-off or a system-wide pattern. The RBI's June 6 monetary policy meeting will set the tone for full-year NIM trajectory. And the FY27 budget speech in February will signal whether the capex push survives a stretched fiscal deficit. SBI is roughly 2.6% of the Nifty 50 and 11% of the Bank Nifty; how its stock prices over the next three weeks will set the read on Indian financials for the rest of the calendar year.

Argentina's country-risk index, the JP Morgan EMBI gauge that measures the spread (extra yield) Argentine sovereign bonds pay over US Treasuries of equivalent maturity, dropped 28 basis points on Friday May 8 to close the week at 510 basis points. That is the lowest reading since February 18, and it caps a steady three-month grind down from the 700-plus levels that defined late summer. Country-risk numbers translate roughly into the cost the government would pay to issue new dollar debt; at 510 basis points and a US 10-year Treasury near 4.3%, an Argentine 10-year would price around 9.4%. Still expensive, but the kind of expensive that lets you actually issue.

The trigger Friday was a one-notch sovereign rating upgrade from Fitch, the third major rating agency, which lifted Argentina from CCC+ to B-. The CCC+ tier is a category that ratings shorthand calls "substantial credit risk." The B- tier is described as "highly speculative" but materially better; it is the band where most large emerging market issuers sit when they are repairing credibility. Fitch cited the steady disinflation path, the primary fiscal surplus the Milei government has held since early 2024, and progress on the gross international reserves target with the IMF. Moody's and S&P had already nudged up earlier this spring, and Friday's move closed the gap.

Bonds responded immediately. Globales, the dollar-denominated Argentine bonds issued under New York law, rose 2.8% on the day. Bonares, which trade under Argentine law and historically carry a credibility discount, gained 1.4%. The currency moves were quieter. The official dollar held at $1,420 (the rate the central bank lets through under the crawling peg system), the MEP dollar (a parallel rate sourced from buying bonds in pesos and selling them for dollars) at $1,429.68, the cash-with-settlement dollar (a similar maneuver routed through New York) at $1,487.40, and the blue dollar, the informal street rate, at $1,400. The blue trading below the official rate is unusual and reflects how much demand for dollars has cooled relative to peso interest rates of roughly 28% annual.

The broader story is that Argentina is angling to come back to international debt markets later this year. The roughly $9 billion in external debt service due over the next twelve months would normally be paid out of reserves, which keeps the central bank in a defensive crouch. If Buenos Aires can issue fresh dollar bonds at 9% or 10%, it pushes the maturity wall (the cluster of upcoming repayments) further out and frees reserves for currency management. The World Bank is reportedly working on a guarantee facility of up to $2 billion that would credit-enhance a portion of any new issue, lowering the coupon by another 100 to 150 basis points. None of that is committed yet, but the Friday rally suggests bondholders are pricing the option as live.

The skeptical reading is that Argentina's risk premium is still by a wide margin the highest in Latin America. Paraguay sits at 102 basis points, Panama at 113, Peru at 119, Guatemala at 120. Even Brazil, with all its fiscal noise, trades around 172. The Latin American average is 256. Argentina at 510 is closer to Ecuador and El Salvador than to its larger neighbors. The path from 510 to a normal investment-adjacent 250 requires continued fiscal discipline. It also requires a structural overhaul of the capital controls (the cepo, which limits dollar purchases for individuals and dividend remittances for companies) that the IMF wants gone before the next disbursement window in September. Removing the cepo without spooking the peso is the technical puzzle that has tripped up every Argentine government for fifteen years.

Knock-on effects worth watching. Argentine equities listed in New York, the ADRs (American Depositary Receipts, which are dollar-traded shares of foreign companies), did not participate in the rally Friday. Mercado Libre, the regional e-commerce giant headquartered in Uruguay but historically read as an Argentine proxy, fell roughly 12% on a soft quarterly print. Banks like Galicia and BBVA Argentina, which usually move with country risk, were flat to down because investors are watching the second-quarter loan growth data more than the bond curve. YPF, the state oil company, gave back early gains as Brent crude fell on the Iran ceasefire news. The peso appreciation against the blue rate also matters for inflation: imported goods get cheaper, which speeds the disinflation path the IMF and Fitch are tracking.

What to watch. INDEC, the national statistics agency, releases April CPI on Wednesday May 13; consensus is for monthly inflation around 1.8%, the lowest since 2021. The July IMF review is the next hard checkpoint, and a clean pass unlocks the next tranche of disbursements. Congressional negotiations over the 2027 budget begin in late May; the size of any tax cut included will tell you how committed the government is to the primary surplus that anchors the rating story. And the World Bank board is expected to consider the partial credit guarantee proposal at its June meeting; approval there is what would actually let Argentina test the international bond market in a serious way before year-end.

Anthropic, the AI lab behind the Claude family of models, signed a deal this week committing roughly $200 billion to Google Cloud over five years in exchange for compute capacity and chips, according to a report by The Information picked up across the tech press. The agreement, which begins in 2027 and runs through 2031, secures Anthropic about 5 gigawatts of compute, the kind of number that until recently was used to describe national-scale data center buildouts rather than a single customer's procurement. Alphabet, Google's parent, has agreed to invest at least $10 billion in Anthropic alongside the contract, with the figure rising to as much as $40 billion if Anthropic hits performance milestones inside the contract.

The chips at the center of the deal are not Nvidia GPUs, the graphics processing units that have powered most generative AI training to date. They are Google's own TPUs, or tensor processing units, custom silicon Google has been developing since 2015 specifically for the matrix math that neural networks run on. The newest TPU generation, designed in partnership with Broadcom, is purpose-built for the transformer architecture (the model design that underpins modern large language models including Claude and Gemini). Google says TPU performance per dollar runs roughly 40% to 50% lower than comparable Nvidia setups for transformer training workloads. For Anthropic, which is on track for $20 billion in server costs in 2026 alone, that discount is the kind of thing that decides whether the company is gross-margin profitable on inference (the cost of running a model to answer a user query) or not.

The strategic move underneath the headline number is Anthropic finally diversifying its compute base. Through 2024 and most of 2025 the company was effectively monogamous with Amazon Web Services and its Trainium chips, with a smaller TPU footprint at Google. The new commitment flips that ratio. Anthropic is now in the position of being the second-biggest TPU customer in the world after Google itself. Amazon stays in the picture, and the previously announced $50 billion AWS commitment continues, but the center of gravity for new Claude training runs is moving to Google's data centers in Ohio and Iowa. The dual-cloud setup also gives Anthropic leverage in the next round of AWS pricing negotiations, something OpenAI never had with its near-exclusive Microsoft Azure relationship.

For Google, the contract is a counterweight to a long-running narrative that the company is losing the cloud AI race to Microsoft and AWS. Alphabet's cloud backlog (the dollar value of contracts signed but not yet recognized as revenue) doubled in the most recent quarter to over $460 billion. The Anthropic deal alone accounts for more than 40% of that figure. It also validates the TPU economics in a way that internal Google use never could. If Anthropic builds the next Claude generation on TPUs and the resulting model is competitive with whatever OpenAI ships on Azure GPUs, the rest of the frontier-model market starts to take TPU procurement seriously, which is exactly the proof point Google needs to unlock new enterprise sales.

The skeptical read is about concentration risk on both sides. Anthropic is now committed to spending what is roughly five years of its current revenue on a single supplier's roadmap. If Google's next-generation TPU slips, or if a competing architecture (custom chips from Microsoft, Meta, or Amazon's Trainium 3) leapfrogs the TPU on price-performance, Anthropic is locked in. On the Google side, $200 billion of contracted revenue from a customer that is itself burning cash and dependent on continued venture funding rounds is a real counterparty exposure. Anthropic raised at a roughly $250 billion valuation earlier this year, but if the AI capex cycle cools the way some analysts now expect by late 2027, both ends of the deal get re-examined.

The butterfly effects ripple in three directions. Nvidia, which has been the default winner of every AI infrastructure announcement for two years, did not lose much on the news because its order book is already booked through 2027, but the long-tail story (the question of who will be buying H300s and Blackwells in 2028) got slightly worse. Broadcom, which manufactures the TPU silicon for Google, picked up roughly 6% on the week and is now valued north of $2 trillion. Power utilities in the regions where the new data centers are landing, particularly American Electric Power in Ohio, saw a quiet bid. The 5 gigawatts of compute capacity in the contract is roughly the electricity load of a mid-sized US state, and the grid interconnection queue is already the binding constraint on AI buildouts.

What to watch. Anthropic is expected to announce a new Claude generation, internally referenced as Claude Mythos, sometime in the third quarter; that release will be the first proof point of whether the TPU-trained models hit the performance bar Anthropic is targeting. Google's next earnings release on July 24 will give the first official commentary on backlog conversion and capex pacing. The US Center for AI Standards and Innovation announced this week that it will run pre-release evaluations on models from Google, Microsoft, and xAI; Anthropic is already in the program, and the speed at which CAISI clears Mythos for public release is now also a function of the Google relationship. And Nvidia's August quarterly print will be the first read on whether the broader AI infrastructure market is still growing or starting to digest.