#分享美股交易赢英伟达股票 Sharing US stock trading and winning NVIDIA shares

At the beginning of 2023, when NVIDIA's stock price was still hovering around $200, I carefully analyzed its business structure. Although many people saw it as "a company that sells graphics cards," its data center business had already surpassed gaming to become the primary source of revenue. The AI wave was just beginning, and training large models heavily depended on the CUDA ecosystem—this is not only hardware but also a software moat.

So I built positions in batches, with an average cost of about $240. The story that followed is well known: the large model arms race kicked off, H100s were in short supply, and the stock price broke through $500, $800, and $900. The entire process tested my holding mentality: every pullback was met with voices of "top," but as long as the demand for AI infrastructure didn't slow down, the trend remained intact.

My strategy was to set dynamic take-profit lines and not raise stop-loss points. After the stock price broke through $800 last year, I partially sold to recover the principal, letting the remaining position run with the profits. By early this year, I gradually sold in batches between $950 and $1,000.

The biggest takeaway from this trade wasn't the doubled returns but learning to distinguish between "volatility" and "risk." True risk isn't the stock price falling but your buy-in logic being disproven. As long as NVIDIA's GPUs remain the "shovels" of the AI era, short-term fluctuations are just noise.

Ultimately, this round of NVIDIA trading achieved an approximate total return of 280%. The biggest regret is that I didn't dare to leverage at the time—but I also feel fortunate I didn't, as a mid-term pullback could have been hard to hold.

The charm of US stocks is that you don't need to catch every opportunity; as long as you dare to hold heavy positions on the right targets and can hold on, one good trade can change your entire year's performance curve.