Micron's entry into the $1 trillion club is not just a stock milestone — it is a signal that the AI investment cycle is deepening and broadening in ways that were not fully priced into the market even six months ago. The traditional AI trade was straightforward: buy Nvidia, buy cloud providers, buy AI software. Micron's surge signals that investors are now moving down the supply chain, recognising that every AI workload — every large language model, every inference engine, every data centre — requires enormous quantities of high-bandwidth memory running alongside the GPUs.

UBS's price target revision from $535 to $1,625 — tripling its previous estimate — is extraordinary in its boldness. The key argument: AI has structurally changed the memory chip market permanently, not cyclically. In traditional semiconductor cycles, memory prices surge and then collapse as new supply comes online. UBS argues that the HBM market is different because the lead time to build HBM-capable fabs is 3–4 years, meaning the current supply shortage will persist regardless of how aggressively competitors invest today.

Micron's 2026 HBM supply being entirely sold out confirms this thesis. The company is not just selling chips — it is selling committed capacity years in advance to hyperscalers like Microsoft, Google, Meta, and Amazon who are locked into multi-year AI infrastructure build-outs. About 2,440 institutions disclosed new positions in Micron in Q1 2026 alone, including Rockefeller Capital Management and Schroders — signalling this is no longer a retail momentum trade but a global institutional thesis.

The risk case is real: SK Hynix currently leads in HBM technology with its HBM3E product, and Samsung is investing aggressively to close the gap. If Micron's HBM4 ramp encounters yield problems, or if AI spending growth disappoints, the stock at 8.42× expected earnings looks reasonable — but at the UBS $1,625 target it would be priced for the absolute best outcome. The margin of safety at current levels is thin. This is a stock for investors who believe the AI infrastructure build-out will continue for 5+ years.

Salesforce's 32% decline in 2026 is one of the most debated stories in technology investing. The company is operationally healthy — trailing twelve-month revenue of $42 billion, operating profits of $8.9 billion, net income of $7.5 billion, and a remaining performance obligation (RPO) of $72.4 billion representing locked-in future revenue. These are not the metrics of a business in crisis.

The decline is entirely driven by a single strategic question: can Salesforce successfully transition from a seat-based pricing model to an AI agent-based model before traditional CRM revenue structurally declines? Bank of America, in its May 18 Underperform note with a $160 target, argued the transition will be painful and margin-dilutive. The argument: every enterprise customer that deploys AI agents to handle customer service, sales follow-up, or data entry work is a customer that needs fewer human seats — and fewer seats means less traditional CRM revenue for Salesforce.

CEO Marc Benioff's counter-thesis is Agentforce: charge per Autonomous Work Unit (AWU) completed rather than per licence. If a customer deploys 10,000 AI agents that each complete 1,000 work units per month, Salesforce charges for 10 million AWUs. In theory, this model generates far more revenue per customer than a per-seat arrangement. In Q4 FY26, Benioff reported 2.4 billion AWUs delivered and 19 trillion tokens processed — large numbers, but the monetisation of those units is what tonight's call must clarify.

What analysts will specifically be listening for on the 5 PM ET call: First, Agentforce Work Unit revenue — not just units delivered, but actual billings. Second, Data Cloud ARR growth (was 200% YoY last quarter — can it sustain?). Third, the current RPO trend — if new bookings are accelerating, it signals customers believe in the Agentforce story. Fourth, any guidance raise for FY27 beyond the $45.8–$46.2 billion already guided. And fifth — critically — any commentary on win rates against Microsoft Copilot, the primary competitor in enterprise AI.

The $50 billion buyback authorisation announced last quarter is the financial backstop. Even if revenue growth disappoints tonight, $50 billion in buybacks at a $168 billion market cap represents 30% of the company being repurchased over time — a powerful floor for the stock. Benioff's $300 million commitment to Anthropic AI token usage signals he believes the AI infrastructure he is building on is real and worth paying for. This is not a company hedging. It is a company making an existential bet.

Every PCE release matters — but Thursday's is uniquely consequential because it is the first major inflation data point that lands on Kevin Warsh's desk as the new Federal Reserve Chair. Warsh has not yet given a major public speech. His first FOMC meeting is in June. The market has no clear read on how hawkish or dovish he will be. Thursday's PCE number is likely to be the primary input into whatever signal he chooses to send in his first press conference.

The specific mechanics: core PCE (excluding food and energy) is the Fed's most trusted inflation gauge because it strips out volatile components and gives a cleaner read on underlying price pressures. The March reading came in at 2.6%, above the Fed's 2% target. The April reading — coming Thursday — will capture the first month where oil prices began declining materially from their Iran-war highs. If energy deflation flowed through to core goods and services pricing, the April reading could surprise to the downside.

The complication: services inflation has been the most stubborn component of PCE throughout 2024–2026. Wage growth remains elevated. Healthcare costs are rising. Housing costs are sticky. Even if energy prices fall, services PCE may not follow quickly — which is why many analysts expect the April reading to still come in above 2.4%, not the 2.2% consensus median.

For the Fed's June decision, the decision tree is: above 2.5% core PCE = Warsh almost certainly signals a hike at June meeting or July at the latest. Between 2.3%–2.5% = Warsh likely holds but with a hawkish tone. Below 2.3% = Warsh can credibly hold rates and signal data-dependence without implying imminent cuts. The bond market will price all three scenarios instantly at 8:30 AM ET Thursday. Watch the 10-year Treasury yield in the 60 seconds after the print — that will be the cleanest real-time signal of which scenario markets believe.

For India: the RBI held its last MPC meeting with oil above $100 and the rupee at ₹94. With oil now at $97 and the rupee at ₹93, the immediate crisis has eased. But if PCE comes in hot on Thursday and the Fed signals a June hike, the dollar will strengthen aggressively — reversing the rupee's recent gains and potentially forcing the RBI into an emergency inter-meeting response. Emkay Global's warning of "inter-meeting action if oil prices remain elevated" extends equally to a dollar-surge scenario driven by a hawkish Fed. Thursday at 7 PM IST is the most important single moment for Indian monetary policy this month.

India's equity market is experiencing its strongest momentum since before the Iran war began. The Sensex has recovered approximately 2,000 points from its recent lows near 73,500. The rupee has strengthened from ₹94 to ₹93.20 — a meaningful move that reduces imported inflation across fuel, edible oils, electronics, and fertilisers simultaneously. Brent crude below $100 is the macro catalyst that is unlocking multiple positive chain reactions in the Indian economy at once.

The FPI outflow situation — which saw $22.2 billion leave Indian equities in under three months — is showing early signs of reversal. When the rupee weakens, dollar-denominated returns on Indian stocks erode for foreign investors even when stock prices rise. With the rupee now strengthening, that calculation begins to reverse. Currency-hedged foreign investors who reduced India exposure are now facing a better entry point than at any time since Q4 2025.

The sector rotation is playing out as expected. Domestic economy plays — OMCs, aviation, auto, banking, infrastructure — are leading. Export-oriented sectors — IT and pharma — are lagging as rupee appreciation reduces their dollar revenue in rupee terms. This is the textbook post-oil-shock recovery trade in an import-dependent emerging market.

However, two important caveats. First, no Iran deal has been formally signed. The oil drop below $100 is driven entirely by deal optimism — if talks collapse and oil snaps back to $107+, all of this week's recovery reverses within 48 hours. Second, Thursday's US PCE data at 7 PM IST is a genuine binary risk event. A hot PCE print will trigger dollar strength, rupee weakness, and a reversal of the FPI inflow narrative before it has even fully materialised. Indian investors who have rotated into rate-sensitive domestic plays this week should treat Thursday evening as a risk checkpoint.

The medium-term picture, assuming Iran deal and cool PCE: Emkay Global's bull case of Nifty 82,000–85,000 by December 2026 requires exactly this combination of conditions. Oil below $90, rupee toward ₹90–91, RBI holding or cutting, FPIs returning to India. Those conditions are now within reach for the first time since February. But they require both diplomatic resolution and macro cooperation from the Fed. Thursday will tell us how much of that cooperation is available.