Anomaly Detection: Hard Drive Prices Only Rose
In August last year, a former ByteDance employee bought two Seagate high-capacity hard drives on Pinduoduo for a personal quantitative trading platform. Shortly after delivery, he noticed the same model’s price kept rising—multiple price adjustments within a week, always upward. For a standardized industrial product with ample capacity, such unilateral increases were abnormal. Using price tracking tools like Manmanmai and Keepa, he extracted price curves for Seagate and Western Digital models, confirming the rise was systemic across the entire high-capacity HDD line, not just a single model or short-term promotion.
AI Demand Extends from GPUs to Storage
Further investigation revealed the root cause: AI data centers’ surging demand for mechanical hard drives. While the market focused on AI’s GPU needs, large model training and inference generate massive data requiring long-term, low-cost storage—primarily served by enterprise nearline hard drives. Major cloud providers (Microsoft, Amazon, Google, Meta) were buying aggressively. Seagate’s HAMR technology meets the data center’s need for higher per-drive capacity, but limited production capacity forces vendors to prioritize more profitable enterprise orders, squeezing retail supply—hence the price increases on Pinduoduo. Seagate’s most recent quarterly revenue grew 39% YoY, with record gross margins, and the storage sector began being priced as part of the AI supply chain.
Using 13F Filings to Confirm Institutional Interest
Personal conviction needed validation. The trader turned to 13F filings—quarterly disclosures of U.S. equity holdings by institutions with over $100 million AUM. He observed Seagate’s institutional ownership over multiple quarters: in H2 2024, only ~800 institutions held the stock, with slight declines. In Q2 2025, the trend reversed, and by Q3 it accelerated to over 1,200 institutions, with new entrants increasing each quarter. While part of the market cap growth came from share price appreciation, the breadth indicators (number of holders, new positions) showed sustained, broad-based institutional accumulation. Confident, he bought 500 shares at ~$150 and later added through LEAPS call options on $STX and $SNDK.
Reflection: From Two Drives to a 6x Return
Seagate closed at ~$150 on the day of purchase; today it trades around $965, a gain of over 600%. In 2025 it briefly overtook Palantir as the S&P 500’s top performer. The initial 500 shares represent ~$400,000 in unrealized profit, excluding later additions. He admits the trade originated from two hard drives bought on Pinduoduo, an outcome he never anticipated.
Investment Framework Summary
The trader distilled a four-step methodology: (1) Pay attention to everyday anomalies like price increases or shortages—they often precede news and earnings reports, giving ordinary people earlier signals than professional institutions; (2) Turn price data into a curve to distinguish trends from noise; (3) Ask whether the underlying demand is long-term and structural, then find listed companies directly benefiting at key positions in the supply chain; (4) Use 13F filings to verify institutional sentiment, observing trends over consecutive quarters rather than a single quarter. He stresses the approach doesn’t guarantee success but ensures decisions are based on logic rather than gut feeling.
Risk Warning and Survivorship Bias
The author clearly states this is a successful case. He has also tracked price signals that turned out to be short-term fluctuations—these were not posted to his group, but they are real and illustrate survivorship bias. This article is a personal review, not investment advice. He encourages readers to notice unusual signals in everyday purchases, dig deeper, find the listed beneficiaries, and use 13F data to confirm institutional interest. Next time a product you frequently buy rises in price for no obvious reason, think: who is capturing that profit, and is that company publicly traded?