Numbers don't lie. But people transcribe them wrong.
A JPMorgan note lands on my screen. SanDisk. Overweight rating. Price target: $2250.
Immediately, my internal alarm triggers. Not because I doubt JPMorgan's analysts. But because I've seen what happens when a single digit gets mis-keyed in a report that moves markets. In crypto, a misplaced decimal in a smart contract drains a pool. In traditional finance, a misplaced zero in a target price creates a narrative that smart money exploits.
Let me walk through the data integrity check before we even touch the semiconductor thesis.
Context: The SanDisk Structure
SanDisk is a NAND flash pure-play. It was spun off from Western Digital. The company operates as an IDM (Integrated Device Manufacturer) in the memory space, partnering heavily with Kioxia on joint R&D and fabrication. Its core product is 3D NAND, used in everything from consumer SSDs to enterprise data center storage.
JPMorgan's call is based on a broad optimism for "memory demand." But here's the first structural problem: SanDisk doesn't make DRAM or HBM. The memory boom everyone is talking about is driven by AI's insatiable appetite for high-bandwidth memory. SanDisk's NAND business benefits from AI indirectly (storage for training data), but it's not the same demand curve. The bank's note likely conflates "memory" with "NAND" — a sloppy categorization that should bother any infrastructure-conscious trader.
Core: The Target Price Anomaly
Let's do the math. SanDisk's post-spin-off share count is estimated between 630 million and 650 million shares. A $2250 target price implies a market capitalization of roughly $1.3 to $1.5 trillion.
For perspective, that valuation exceeds the entire market cap of Micron, SK Hynix, and Samsung Semiconductor combined. It would place SanDisk among the top 10 most valuable companies globally, sitting alongside Apple and Microsoft. The implied P/E ratio, assuming 2025 earnings per share of $5 to $10, would be between 225x and 450x.
This is not a target. This is a data entry error.
The most likely correction is $225 per share, or a target market cap of $225 billion. The discrepancy is a 10x factor. In my experience auditing crypto project financials, I've seen similar errors when analysts confuse total addressable market with revenue projections. The signal here is not about SanDisk's value. The signal is about the quality of the source material.
Data over drama. When a headline number fails a basic sanity check against known market caps, the entire thesis becomes suspect. This is why I never trade on unlocked analyst reports. I verify the math myself.
Contrarian: The Bull Case Is Still Real, But It's Not About the Target
The contrarian angle is that the underlying thesis — memory cyclical recovery — is valid. The NAND market has been in a deep correction. Supply cuts from major players are starting to bite. Prices are stabilizing. SanDisk, as a pure-play NAND maker with Kioxia's technology backing, is structurally leveraged to that recovery.
But the retail narrative will latch onto the $2250 number. Social media will pump it. Chat groups will celebrate it. The smart money will quietly sell into that enthusiasm, knowing the target is fundamentally unachievable. The divergence between narrative and reality creates a liquidity trap. Volume-driven exit strategists recognize this pattern: when the price approaches the headline target, it's time to exit, not to accumulate.
Takeaway: Trade the Structure, Not the Headline
Calculate. Execute. Repeat.
The SanDisk $2250 target is a case study in why I treat every piece of financial data with the same skepticism I apply to a new DeFi protocol's smart contract. Verify the inputs. Stress-test the assumptions. If the math doesn't hold at the most basic level, the entire thesis is compromised.
The real question is not whether SanDisk is a buy. It's whether you have the discipline to ignore a headline that doesn't pass the sniff test. The market will eventually correct the mispricing. The question is whether you'll be on the right side of that correction.