Why Sandisk Stock Dropped Today

Broadcom reported Q2 fiscal 2026 EPS of $2.44 on $22.2B in sales, beating expectations, and guided Q3 revenue to $29.4B versus the $28.5B consensus. However, its AI chip sales outlook of $16B for Q3 fell short of the $17.2B Street estimate, triggering a 15% sell-off in Broadcom and a 1.7% pullback in Sandisk. The article argues the broader memory demand thesis remains intact despite near-term volatility.

Analysis

The market is treating Broadcom’s guide as a read-through on AI demand, but the more important signal is that the AI supply chain is shifting from scarcity in compute to scarcity in memory bandwidth and packaging. That tends to be a second-order positive for SNDK because memory pricing is driven less by one company’s quarterly beat and more by multi-quarter capacity discipline; short-term wobble in an AI bellwether does not meaningfully change the structural shortage. The fastest reaction is likely in the group’s factor exposure rather than fundamentals: momentum traders are selling “AI infrastructure” on any hint of deceleration, which creates dislocations in the memory names that are more sentiment-linked than demand-linked.

The contrarian takeaway is that AVGO’s “miss” on AI is still growth that is compounding at a rate most semi names would envy. If the street is disappointed by a modest guide-down while revenue is still exploding, expectations for adjacent winners may have been lifted too far rather than broken. In that setup, the better read-through is not that AI capex is stalling, but that hyperscaler spend is rotating toward more memory-intensive architectures, which should keep pricing power intact for flash and HBM-adjacent suppliers even if compute silicon growth becomes less linear.

The risk is time horizon mismatch: memory equities can correct 5-10% on a single data point even when the underlying supply/demand balance changes only over 2-3 quarters. If Broadcom’s softer AI guide is echoed by NVDA or other infra vendors over the next month, the current narrative could morph into a broader multiple compression trade on AI beneficiaries. Conversely, if next memory pricing checks confirm continued tightness, this selloff should reverse quickly because the bear case depends on a demand air pocket that is not yet visible in actual pricing or channel data.

Key Decisions for Investors

• Add to SNDK on weakness over the next 1-2 sessions; use the post-Broadcom flush as an entry point for a 4-8 week trade targeting a re-rating back toward analyst-driven highs if memory pricing remains firm.
• Pair long SNDK / short AVGO for the next 1-2 months: long the name with direct pricing leverage and short the one most exposed to multiple compression from any AI capex skepticism; structure sized small because both can move on the same macro tape.
• If already long AI semis, hedge with short-dated NVDA puts into the next data-point cluster; broad AI sentiment is fragile and one more guide-down could trigger a 1-2 week de-risking cascade.
• Avoid chasing the post-earnings selloff in AVGO until the market digests whether this is a true demand inflection or just a high-expectation reset; the risk/reward is poor for immediate dip-buying after a 15% gap lower.
• For higher-conviction investors, consider a call spread in SNDK with 1-2 quarter tenor to express a recovery in memory pricing without taking full delta risk against broader AI factor volatility.