「ByteDance Employee Quits After Making 30 Million Yuan from Stock Trading」: A Recap – It All Started with Two Hard Drives on Pinduoduo
- Core Thesis: The author identified an anomaly on Pinduoduo – a continuous one-way price increase for large-capacity mechanical hard drives. By tracing this back step by step, he deduced it was caused by a squeeze on retail supply from massive procurement of enterprise-grade hard drives for AI data centers. Acting on this, he bought Seagate stock and ultimately obtained a substantial paper profit.
- Key Elements:
- After buying a hard drive on Pinduoduo, the author noticed its price kept rising consistently in one direction, rather than fluctuating with seasonal promotions. Using price comparison tools, he confirmed this was a widespread phenomenon across the entire product line.
- In-depth research revealed that the massive amounts of data generated by AI model training require long-term storage on low-cost, large-capacity mechanical hard drives (nearline), not SSDs.
- Manufacturers like Seagate prioritize supplying higher-margin data center orders, squeezing retail supply. This was the root cause of the retail hard drive price increases, and their financial reports showed record revenue and gross margins.
- By analyzing 13F reports, the author found that the number of institutions investing in Seagate grew steadily and orderly throughout 2024, confirming a consensus of professional capital on the industry and strengthening his conviction to add to his position.
- The author's summarized trading methodology: start with anomalous price increases in everyday consumer goods, validate the trend with data, trace back up the supply chain to find the core listed company, and finally confirm the direction using institutional holdings data.
Original Title: "Buying Hard Drives on Pinduoduo Led Me to Unexpectedly Invest in Storage | How Ordinary People Can Trade Using 'Information Around Them'"
Original Author: Leto Bao (X: @leto_bao)
All the information below was shared last year in the ByteDance US Stocks group, including trading targets, positions, and profits/losses, all verifiable.
Last August, I originally just wanted to buy two hard drives.
At the time, I was building a small quantitative trading platform and wanted to pull some tick-level market data locally for storage. Dozens of terabytes of data needed a place, so I ordered two large-capacity Seagate drives on Pinduoduo. Looking back, that hard drive purchase was the starting point for my foray into storage investments over the past two years.
Hard Drive Prices Started Changing Daily
After the hard drives arrived, I bookmarked the link, planning to buy more if needed. A few days later, I checked again, and the price had increased; a few more days passed, and it rose again. The same model, the same store, with multiple price adjustments within a week, and it only ever went up, never down.
For a highly standardized industrial product with massive production capacity, the retail price shouldn't keep rising unidirectionally like this. I saw it as an anomaly worth investigating, not just a simple seller price hike.
I used price comparison tools like Manmanmai and Keepa to pull the price curve for that drive over the past few months and compared it with other large-capacity models from Seagate and Western Digital. The conclusion was consistent: this wasn't an issue with a single model. The entire line of large-capacity mechanical hard drives was rising, and it was a sustained, unidirectional increase, not a short-term fluctuation caused by promotions.
At this point, I could basically confirm there was a bigger reason behind it.
Following the Trail Led to AI Competing for Hard Drives
Digging deeper, the logic became clearer.
The market focuses more on AI's demand for GPUs, but its demand for storage is equally enormous. Training and inference of large models generate massive amounts of data that need long-term storage. For long-term, low-cost data storage, the primary medium isn't SSDs but large-capacity mechanical hard drives. The type of drives used in data centers are called nearline enterprise hard drives, and major cloud providers like Microsoft, Amazon, Google, and Meta are purchasing them in large quantities.
Seagate was perfectly positioned to capitalize on this demand wave. Its flagship HAMR technology significantly increases per-drive capacity, directly meeting data center needs. With limited production capacity, manufacturers prioritize supplying higher-margin enterprise orders, thereby squeezing retail supply. The price increases I saw on Pinduoduo essentially reflected AI data center procurement demand trickling down to the consumer end.
I checked Seagate's financial reports at the time: revenue grew 39% year-over-year in the latest quarter, gross margins hit record highs, and the market was beginning to price data storage as a critical component of the AI supply chain.
After confirming the logic, I bought 500 shares at around $150 and posted my rationale, cost basis, and position in the company's internal US stock discussion group.
What Gave Me Confidence to Add to My Position Was the Subsequent 13F Filings
Thinking it through myself wasn't enough; I needed to confirm whether institutional capital was making the same judgment.
The US stock market has a very useful piece of public data: institutions managing over $100 million must disclose their US stock holdings every quarter, known as the 13F filing. This acts as a legal, public record of institutional holdings that anyone can access.
I didn't immediately add to my position. Instead, I wanted to observe the trend over several quarters. A change in a single quarter could be coincidental; consistent movement in the same direction over consecutive quarters is more credible. When the 13F filing for the third quarter came out in November, I plotted Seagate's institutional holdings over the past year, and the direction was clear:
In the second half of 2024, this stock was largely ignored, with the number of holding institutions hovering around just over 800, even declining slightly. There was a clear inflection point in Q2 2025, with further acceleration in Q3: the number of holding institutions increased from over 800 to over 1,200, and the number of new institutions entering each quarter also grew sequentially.
It's worth noting that while the market value of holdings surged to $45.6 billion within a year, a significant portion came from the stock price appreciation itself, not entirely new capital inflows. However, "breadth" indicators like the number of institutions and new positions opened are more telling. Their sequential increase quarter over quarter meant this wasn't a bet by one or two funds, but a wave of professional capital steadily entering the space.
Only after confirming all this did I feel confident enough to significantly increase my position and seriously study the storage sector. Subsequently, I continued adding to my positions in $STX and $SNDK through LEAPS CALLs.
Looking Back Now
On the day I bought the hard drives, Seagate closed at just over $150. Today, it's over $965, a more than sixfold increase. Last year, it briefly surpassed Palantir to become the top gainer in the S&P 500 for the year. Just the initial 500 shares represent an unrealized gain of around $400,000, not to mention the subsequent additions to the position.
That two hard drives ultimately led to such a trade still feels somewhat surprising to me.
Summarizing the Thought Process
It's actually not that complicated:
· Anomalies in everyday items (price hikes, shortages, queues) often precede news reports and financial statements. Ordinary people sometimes get first-hand signals even earlier than professional institutions.
· Don't stop at the impression that "it seems more expensive." Plot the price as a curve; you can basically tell whether it's a trend or noise.
· Then, dig deeper: Is there a long-term, structural demand behind this? Then, find the listed companies that directly benefit and hold key positions in the industrial chain.
· Finally, use 13F filings to verify institutional sentiment, and look at trends over multiple consecutive quarters, not just a single one.
This method isn't foolproof, but it at least ensures that buying decisions are based on logic, not just feelings.
Finally, Let's Be Clear About the Risks
This was a successful case. I've also tracked a price increase signal that ultimately proved to be a short-term fluctuation. Those instances weren't shared in the group, but they were equally real. Survivorship bias is very evident here.
The above is a personal review and does not constitute investment advice. All profits and losses are your own responsibility.
But I do endorse this way of thinking: Pay attention to anomalies in everyday life, ask one more layer of "why," find the listed company behind it, and then check 13F filings to see if professional money agrees. Next time something you regularly buy increases in price for no apparent reason, you might think: Who is making money from this? Is that company already public?
